Fluid Physiology : A Handbook for Anaesthesia and Critical Care Practice [1 ed.] 9781527542020, 9781527540316

Citation preview

Fluid Physiology



Fluid Physiology: A Handbook for Anaesthesia and Critical Care Practice By

Thomas Woodcock

Fluid Physiology: A Handbook for Anaesthesia and Critical Care Practice By Thomas Woodcock This book first published 2019 Cambridge Scholars Publishing Lady Stephenson Library, Newcastle upon Tyne, NE6 2PA, UK British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Copyright © 2019 by Thomas Woodcock All rights for this book reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the copyright owner. ISBN (10): 1-5275-4031-6 ISBN (13): 978-1-5275-4031-6

    &ŽƌĞǁŽƌĚďLJ:LOOLDP:0XLU͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘džŝŝ  WƌĞĨĂĐĞ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘džŝǀ  ŚĂƉƚĞƌϭ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭ ^ƚĂƌůŝŶŐ͛Ɛ&ůƵŝĚWŚLJƐŝŽůŽŐLJ͗,ŝƐƚŽƌŝĐWĞƌƐƉĞĐƚŝǀĞ ŚĂƉƚĞƌƐƵŵŵĂƌLJ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭ dŚĞŽƌŝŐŝŶĂů^ƚĂƌůŝŶŐƉƌŝŶĐŝƉůĞ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘Ϯ dŚĞƐĐŝĞŶĐĞŽĨ^ƚĂƌůŝŶŐĨŽƌĐĞƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϯ ZĞǀŝƐŝŶŐƚŚĞ^ƚĂƌůŝŶŐƉƌŝŶĐŝƉůĞ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϲ >ĂƌŐĞƉŽƌĞƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϵ dŚĞƌĞǀŝƐĞĚŽƌƐƚĞĂĚLJͲƐƚĂƚĞ^ƚĂƌůŝŶŐƉƌŝŶĐŝƉůĞ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϵ dŚĞŝŶǀĞƌƐĞĚĞƉĞŶĚĞŶĐĞŽĨƉĞƌŝĐĂƉŝůůĂƌLJʋŝŽŶ:ʆƌĞƐƚƌŝĐƚƐĂďƐŽƌƉƚŝŽŶƚŽ ĂƚƌĂŶƐŝĞŶƚƉŚĞŶŽŵĞŶŽŶŝŶŵŽƐƚƚŝƐƐƵĞƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϮ džĐĞƉƚŝŽŶƐƚŽƚŚĞ͚ŶŽƐƚĞĂĚLJͲƐƚĂƚĞĂďƐŽƌƉƚŝŽŶ͛ƌƵůĞ͗ĞĨĨĞĐƚŽĨůŽĐĂů ĞƉŝƚŚĞůŝĂůƚƌĂŶƐƉŽƌƚ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϴ džĐĞƉƚŝŽŶƐƚŽƚŚĞ͚ŶŽƐƚĞĂĚLJͲƐƚĂƚĞĂďƐŽƌƉƚŝŽŶ͛ƌƵůĞ͗ĚŝƐĐŽŶƚŝŶƵŽƵƐ ĐĂƉŝůůĂƌŝĞƐŽĨƚŚĞƐŝŶƵƐŽŝĚĂůƚŝƐƐƵĞƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϴ džĐĞƉƚŝŽŶƐƚŽƚŚĞ͚ŶŽƐƚĞĂĚLJͲƐƚĂƚĞĂďƐŽƌƉƚŝŽŶ͛ƌƵůĞ͗ƌĞŶĂůŵĞĚƵůůĂ͘͘͘ϭϵ džĐĞƉƚŝŽŶƐƚŽƚŚĞ͚ŶŽƐƚĞĂĚLJͲƐƚĂƚĞĂďƐŽƌƉƚŝŽŶ͛ƌƵůĞ͗ŚLJƉŽĚĞƌŵŽĐůLJƐŝƐϭϵ ŶŝŵĂůƌĞƐĞĂƌĐŚ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϮϬ ZĞĨĞƌĞŶĐĞƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘Ϯϯ  ŚĂƉƚĞƌϮ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘Ϯϱ ^ƚĂƌůŝŶŐ͛Ɛ&ůƵŝĚWŚLJƐŝŽůŽŐLJŝŶůŝŶŝĐĂůWƌĂĐƚŝĐĞ͗,ŝƐƚŽƌŝĐWĞƌƐƉĞĐƚŝǀĞ Chapter Summary͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘Ϯϱ ^ƚĂƌůŝŶŐ͛ƐŝŶĨůƵĞŶĐĞŽŶĐůŝŶŝĐĂůƉƌĂĐƚŝĐĞ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘Ϯϲ dŚĞ'ƌĞĂƚtĂƌϭϵϭϰͲϭϵϭϴ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘Ϯϲ dŚĞ^ĞĐŽŶĚtŽƌůĚtĂƌϭϵϯϵͲϭϵϰϱ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘Ϯϳ DŝůŝƚĂƌLJƚƌĂƵŵĂĂƐĂƚĞƐƚŝŶŐŐƌŽƵŶĚĨŽƌƌĞƐƵƐĐŝƚĂƚŝǀĞĨůƵŝĚƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘Ϯϳ dŚĞĞƐƐĞŶƚŝĂůŝƚLJŽĨƚŚĞůLJŵƉŚĂƚŝĐƐLJƐƚĞŵƚŽƌĞĐŽǀĞƌLJĨƌŽŵƐŚŽĐŬ͘͘͘͘͘͘͘Ϯϴ ĞdžƚƌĂŶƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘Ϯϴ ^ƚĂƌĐŚĞƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘Ϯϵ

˜‹

‘–‡–•

dŚĞůŝďĞƌĂůǀƐ͘ƌĞƐƚƌŝĐƚŝǀĞĨůƵŝĚĚĞďĂƚĞ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϯϬ /ŶƚĞŶƐŝǀĞĂƌĞhŶŝƚƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϯϬ ,LJƉĞƌǀŽůĂĞŵŝĐŚLJƉĞƌĚLJŶĂŵŝĐƚŚĞƌĂƉLJ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϯϭ WŚLJƐŝŽůŽŐLJůĂďŽƌĂƚŽƌLJĂĚǀĂŶĐĞƐƉĂƐƐƵŶĂƉƉƌĞĐŝĂƚĞĚ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϯϮ ,ĂǀĞĂ/'Z͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϯϯ dŚĞ:ĐƵƌǀĞĂŶĚƚŚĞ:ƉŽŝŶƚ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϯϰ &ůƵŝĚďŽůƵƐƚŚĞƌĂƉLJ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϯϱ >ŝďĞƌĂůǀƐƌĞƐƚƌŝĐƚŝǀĞĨůƵŝĚƚŚĞƌĂƉLJƌĞǀŝƐŝƚĞĚ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϯϲ ŝŽƉŚLJƐŝĐĂůŽůůŽŝĚKƐŵŽƚŝĐWƌĞƐƐƵƌĞƚŚĞƌĂƉLJƐĐĞƉƚŝĐŝƐŵŐĂŝŶƐ ƚƌĂĐƚŝŽŶ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϯϲ ZĞĨĞƌĞŶĐĞƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϯϳ  ŚĂƉƚĞƌϯ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϰϯ ŝŽůŽŐŝĐĂů^ŽůƵƚŝŽŶƐ ŚĂƉƚĞƌ^ƵŵŵĂƌLJ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϰϯ tĂƚĞƌĂŶĚƚŚĞŚLJĚƌŽŐĞŶŝŽŶĐŽŶĐĞŶƚƌĂƚŝŽŶ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϰϰ Ɖ,Žƌ΀,н΁͍͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϰϱ DŝƚŽĐŚŽŶĚƌŝĂůŚLJĚƌŽŐĞŶŝŽŶŐƌĂĚŝĞŶƚ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϰϲ ĐŝĚƐĂŶĚďĂƐĞƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϰϲ ^ŽůƵƚŝŽŶƐĂŶĚ^ŽůƵƚĞƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϰϲ ^ƚƌŽŶŐĞůĞĐƚƌŽůLJƚĞƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϰϳ tĞĂŬĞůĞĐƚƌŽůLJƚĞƐĂŶĚďƵĨĨĞƌƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϰϵ ĂƌďŽŶĚŝŽdžŝĚĞŝŶǁĂƚĞƌ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϱϬ ůŽŽĚŐĂƐĞƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϱϭ &ĞŶĐůͲ^ƚĞǁĂƌƚŵĞƚŚŽĚ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϱϮ hƌĞĂƐLJŶƚŚĞƐŝƐŝŶĂĐŝĚͲďĂƐĞďĂůĂŶĐĞ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϱϯ KƐŵŽƚŝĐĐŽŶĐĞŶƚƌĂƚŝŽŶŽĨƐŽůƵƚĞƐǁŝƚŚŝŶƚŚĞďŽĚLJǁĂƚĞƌ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϱϰ hƌĞĂ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϱϱ DĞĂƐƵƌŝŶŐŽƐŵŽůĂůŝƚLJ͕ĐĂůĐƵůĂƚŝŶŐŽƐŵŽůĂƌŝƚLJ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϱϲ ZĞŐƵůĂƚŝŽŶŽĨŽƐŵŽůĂƌŝƚLJďLJǁĂƚĞƌƌĞƐŽƌƉƚŝŽŶ;ĂŶƚŝĚŝƵƌĞƐŝƐͿ͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϱϲ ZĞŐƵůĂƚŝŽŶŽĨŽƐŵŽůĂƌŝƚLJ͖ƌĞůĞĂƐĞĨƌŽŵĂŶƚŝĚŝƵƌĞƐŝƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϱϳ ůĚŽƐƚĞƌŽŶĞ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϱϴ ƌŐŝŶŝŶĞͲǀĂƐŽƉƌĞƐƐŝŶĂĞŵŝĂŝŶĐƌŝƚŝĐĂůŝůůŶĞƐƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϱϴ dŽƚĂůďŽĚLJǁĂƚĞƌ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϲϬ ůŝŶŝĐĂůƵƐĞŽĨƚŽƚĂůďŽĚLJǁĂƚĞƌĞƐƚŝŵĂƚŝŽŶĂŶĚŵŽĚŝĨŝĞĚ ďŽĚLJǁĞŝŐŚƚƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϲϬ ĞůůĨůƵŝĚĂŶĚĞdžƚƌĂĐĞůůƵůĂƌĨůƵŝĚ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϲϮ ^ƚĂƌůŝŶŐĨŽƌĐĞƐďĞƚǁĞĞŶĞdžƚƌĂĐĞůůƵůĂƌĂŶĚŝŶƚƌĂĐĞůůƵůĂƌĨůƵŝĚƐ͘͘͘͘͘͘͘͘͘͘͘͘ϲϮ

Ž—‹†Š›•‹‘Ž‘‰›

˜‹‹

DĂŝŶƚĞŶĂŶĐĞŽĨƚŚĞĞdžƚƌĂĐĞůůƵůĂƌͲŝŶƚƌĂĐĞůůƵůĂƌƐŽůƵƚĞďĂůĂŶĐĞŝƐ ĞŶĞƌŐLJͲĚĞƉĞŶĚĞŶƚ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϲϯ ŽƵďůĞŽŶŶĂŶĞĨĨĞĐƚ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϲϰ ĞůůǀŽůƵŵĞƌĞŐƵůĂƚŝŽŶĂŶĚŝŶƚƌĂĐƌĂŶŝĂůƉƌĞƐƐƵƌĞ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϲϰ ĞůůǀŽůƵŵĞƌĞŐƵůĂƚŝŽŶďĞLJŽŶĚƚŚĞďƌĂŝŶ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϲϱ ,ĞƐƐĞůƐ͛ůƚĞƌŶĂƚŝǀĞDŽĚĞůŽĨǁĂƚĞƌ͕ƐŽĚŝƵŵĂŶĚƉŽƚĂƐƐŝƵŵ ĚŝƐƚƌŝďƵƚŝŽŶ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϲϲ ĞŚLJĚƌĂƚŝŽŶ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϲϳ ,LJƉŽŶĂƚƌĂĞŵŝĂ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϲϳ dŚĞĞdžƚƌĂĐĞůůƵůĂƌĚŝƐƉŽƐŝƚŝŽŶŽĨƐŽůǀĞŶƚĂŶĚƐŽůƵƚĞƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϲϴ KĞĚĞŵĂ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϲϵ ƋƵĂŶƚŝƚĂƚŝǀĞĂƉƉƌŽĂĐŚƚŽĨůƵŝĚĂŶĚĞůĞĐƚƌŽůLJƚĞďĂůĂŶĐĞ ĂƚƚŚĞďĞĚƐŝĚĞ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϳϬ ZĞĨĞƌĞŶĐĞƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϳϮ  ŚĂƉƚĞƌϰ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϳϲ ŶĚŽƚŚĞůŝƵŵ ŚĂƉƚĞƌƐƵŵŵĂƌLJ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϳϲ ŶĚŽƚŚĞůŝƵŵ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϳϳ ŝƐĐŽǀĞƌLJŽĨƚŚĞŐůLJĐŽĐĂůLJdž͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϳϵ ^ƚƌƵĐƚƵƌĞŽĨƚŚĞĞŶĚŽƚŚĞůŝĂůŐůLJĐŽĐĂůLJdž͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϳϵ dŚĞŐůLJĐŽĐĂůLJdžĂŶĚƚŚĞĞŶĚŽƚŚĞůŝĂůƐƵƌĨĂĐĞůĂLJĞƌ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϴϲ sŽůƵŵĞĂŶĚƚŚŝĐŬŶĞƐƐŽĨƚŚĞĞŶĚŽƚŚĞůŝĂůƐƵƌĨĂĐĞůĂLJĞƌ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϴϳ dŚĞĞŶĚŽƚŚĞůŝĂůƐƵƌĨĂĐĞůĂLJĞƌ͖ŝŶĨůĂƚŝŽŶĂŶĚĚŝƐŝŶƚĞŐƌĂƚŝŽŶ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϴϵ &ƵŶĐƚŝŽŶƐŽĨƚŚĞĞŶĚŽƚŚĞůŝĂůŐůLJĐŽĐĂůLJdž͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϵϭ sĂƐĐƵůĂƌĞŶĚŽƚŚĞůŝĂůĐĞůůƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϵϱ /ŶƚĞƌĞŶĚŽƚŚĞůŝĂůũƵŶĐƚŝŽŶƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϵϱ ƋƵĂƉŽƌŝŶĐŚĂŶŶĞůƐĂƌĞƉƌĞƐĞŶƚŝŶĞŶĚŽƚŚĞůŝĂůĐĞůůƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϵϳ ƋƵĂƉŽƌŝŶĐŚĂŶŶĞůƐŽĨĂƐƚƌŽĐLJƚĞĨŽŽƚƉƌŽĐĞƐƐĞƐĂƌĞŝŵƉŽƌƚĂŶƚƚŽ ďůŽŽĚͲďƌĂŝŶƉĞƌŵĞĂďŝůŝƚLJďĂƌƌŝĞƌ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϵϴ ƋƵĂƉŽƌŝŶϮŝƐĞdžƉƌĞƐƐĞĚŝŶƚŚĞƌĞŶĂůĐŽůůĞĐƚŝŶŐĚƵĐƚ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϵϴ ƋƵĂƉŽƌŝŶƐŝŶƚŚĞůƵŶŐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϵϴ ĂƉŝůůĂƌLJĐůĂƐƐĞƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϵϴ ŝƌĐƵůĂƚŝŶŐĞŶĚŽƚŚĞůŝĂůĐĞůůƐĂŶĚĞŶĚŽƚŚĞůŝĂůͲĚĞƌŝǀĞĚŵŝĐƌŽƉĂƌƚŝĐůĞƐϭϬϱ dŚĞƐƵďĞŶĚŽƚŚĞůŝĂůďĂƐĞŵĞŶƚŵĞŵďƌĂŶĞĂŶĚĞdžƚƌĂĐĞůůƵůĂƌŵĂƚƌŝdž ϭϬϱ ZĞĨĞƌĞŶĐĞƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϬϲ  

˜‹‹‹

‘–‡–•

ŚĂƉƚĞƌϱ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϭϮ /ŶƚĞƌƐƚŝƚŝƵŵĂŶĚ>LJŵƉŚĂƚŝĐ^LJƐƚĞŵƐ ŚĂƉƚĞƌƐƵŵŵĂƌLJ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϭϮ /ŶƚĞƌƐƚŝƚŝƵŵ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϭϮ DŽůĞĐƵůĂƌƐƚƌƵĐƚƵƌĞŽĨƚŚĞƚƌŝƉŚĂƐŝĐŝŶƚĞƌƐƚŝƚŝƵŵ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϭϯ ,LJĂůƵƌŽŶĂŶ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϭϰ dŚĞƉĞƌŝǀĂƐĐƵůĂƌŵĂƚƌŝdž͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϭϱ /ŶƚĞƌƐƚŝƚŝĂůǁĂƚĞƌ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϭϲ 'ĞůƐǁĞůůŝŶŐƉƌĞƐƐƵƌĞ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϭϳ ŽůůŽŝĚŽƐŵŽƚŝĐƉƌĞƐƐƵƌĞŽĨƚŚĞŝŶƚĞƌƐƚŝƚŝƵŵ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϭϳ ůďƵŵŝŶĞdžĐůƵƐŝŽŶĂŶĚĂƋƵĞŽƵƐƉŚĂƐĞǀŝƐĐŽƐŝƚLJ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϭϳ /ŶƚĞƌƐƚŝƚŝĂůĨůƵŝĚƉƌĞƐƐƵƌĞ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϭϴ ŶĂƚŽŵŝĐĨĞĂƚƵƌĞƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϭϴ >LJŵƉŚĂƚŝĐǀĂƐĐƵůĂƌƐLJƐƚĞŵ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϭϵ /ŶƚĞƌƐƚŝƚŝĂůĨůƵŝĚĚLJŶĂŵŝĐƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϮϬ >LJŵƉŚĂƚŝĐƐĂŶĚƚŚĞŝŶƚĞƌƐƚŝƚŝĂůƐƚŽƌĂŐĞŽĨƐŽĚŝƵŵ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϮϮ /ŶƚĞƌƐƚŝƚŝĂůĨůƵŝĚŝŶĐƌŝƚŝĐĂůŝůůŶĞƐƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϮϰ WĞƌŝƚŽŶĞĂůĚŝĂůLJƐŝƐ͕ƉĞƌŝƚŽŶĞĂůƌĞƐƵƐĐŝƚĂƚŝŽŶ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϮϲ ZĞĨĞƌĞŶĐĞƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϮϲ  ŚĂƉƚĞƌϲ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϯϬ ^LJƐƚĞŵŝĐĐŝƌĐƵůĂƚŝŽŶ͗dŚĞƉĞƌŝƉŚĞƌĂůǀĂƐĐƵůĂƌůŽŽƉƐ ŚĂƉƚĞƌƐƵŵŵĂƌLJ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϯϬ ^LJƐƚĞŵŝĐĐŝƌĐƵůĂƚŝŽŶ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϯϬ ĞƌĞďƌĂůĐŝƌĐƵůĂƚŝŽŶ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϯϬ ŽƌŽŶĂƌLJĐŝƌĐƵůĂƚŝŽŶ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϯϱ ,ĞƉĂƚŝĐĐŝƌĐƵůĂƚŝŽŶƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϯϲ /ŶƚĞƐƚŝŶĂůŵƵĐŽƐĂůĐŝƌĐƵůĂƚŝŽŶ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϯϳ /ŶƚĞƌƐƚŝƚŝĂůĨůƵŝĚĂŶĚ>LJŵƉŚŝƌĐƵůĂƚŝŽŶŝŶƚŚĞ>ŝǀĞƌĂŶĚŝŶƚĞƐƚŝŶĂů ŵƵĐŽƐĂ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϯϴ ^ƉůĞŶŝĐĐŝƌĐƵůĂƚŝŽŶ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϯϵ ŽŶĞŵĂƌƌŽǁĐŝƌĐƵůĂƚŝŽŶ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϰϬ ZĞŶĂůĐŝƌĐƵůĂƚŝŽŶ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϰϬ ^ĞĐƌĞƚŽƌLJŐůĂŶĚƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϰϯ >LJŵƉŚŶŽĚĞƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϰϯ ZĞĨĞƌĞŶĐĞƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϰϰ  

Ž—‹†Š›•‹‘Ž‘‰›

‹š

ŚĂƉƚĞƌϳ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϰϳ WƵůŵŽŶĂƌLJĐŝƌĐƵůĂƚŝŽŶ ŚĂƉƚĞƌƐƵŵŵĂƌLJ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϰϳ dŚĞƉƵůŵŽŶĂƌLJǀĂƐĐƵůĂƌůŽŽƉ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϰϳ WƌĞĐĂƉŝůůĂƌLJĂŶĚƉŽƐƚĐĂƉŝůůĂƌLJĂŶĂƐƚŽŵŽƐĞƐďĞƚǁĞĞŶďƌŽŶĐŚŝĂůĂŶĚ ƉƵůŵŽŶĂƌLJĂƌƚĞƌŝĂůǀĂƐĐƵůĂƌůŽŽƉƐŵĂLJĐŽŶƚƌŝďƵƚĞƚŽƐŽŵĞ ƉŚLJƐŝŽůŽŐŝĐĂůŽďƐĞƌǀĂƚŝŽŶƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϰϴ WƵůŵŽŶĂƌLJďůŽŽĚĨůŽǁ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϰϴ WƵůŵŽŶĂƌLJďůŽŽĚǀŽůƵŵĞ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϰϵ WƵůŵŽŶĂƌLJŵŝĐƌŽǀĂƐĐƵůĂƌĞŶĚŽƚŚĞůŝƵŵĂŶĚŝŶƚĞƌƐƚŝƚŝƵŵ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϰϵ EĞŐĂƚŝǀĞƉƵůŵŽŶĂƌLJŝŶƚĞƌƐƚŝƚŝĂůƉƌĞƐƐƵƌĞ͗ŬĞĞƉŝŶŐƚŚĞĂůǀĞŽůŝĚƌLJ͘͘ϭϱϯ dŚĞĚŝƐƚƌŝďƵƚŝŽŶŽĨƉƵůŵŽŶĂƌLJŽĞĚĞŵĂ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϱϰ WůĞƵƌĂůĨůƵŝĚ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϱϱ ZĞĨĞƌĞŶĐĞƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϱϳ  ŚĂƉƚĞƌϴ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϱϵ WůĂƐŵĂsŽůƵŵĞ ŚĂƉƚĞƌƐƵŵŵĂƌLJ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϱϵ /ŶƚƌŽĚƵĐƚŝŽŶ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϱϵ ĞĨŝŶŝŶŐĂŶĚŵĞĂƐƵƌŝŶŐƉůĂƐŵĂǀŽůƵŵĞ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϲϬ sŽůƵŵĞ^ĞŶƐŽƌƐĂŶĚŶĞƵƌŽŚƵŵŽƌĂůƌĞƐƉŽŶƐĞ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϲϯ ŝƐŽƌĚĞƌƐŽĨƉůĂƐŵĂǀŽůƵŵĞ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϲϱ ůďƵŵŝŶĂŶĚƉůĂƐŵĂǀŽůƵŵĞ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϲϴ ƌĞǀŝƐĞĚdǁŝŐůĞLJͲ,ŝůůŵĂŶĚŝĂŐƌĂŵ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϳϬ dŚĞƌĂƉĞƵƚŝĐŽƉƚŝŽŶƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϳϮ ZĞĨĞƌĞŶĐĞƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϳϰ  ŚĂƉƚĞƌϵ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϳϳ ůďƵŵŝŶĂŶĚKƚŚĞƌŝƌĐƵůĂƚŝŶŐWƌŽƚĞŝŶƐ ŚĂƉƚĞƌƐƵŵŵĂƌLJ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϳϳ ^ŽůƵďůĞĞdžƚƌĂĐĞůůƵůĂƌƉƌŽƚĞŝŶƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϳϳ ůďƵŵŝŶ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϳϴ ŝƌĐƵůĂƚŝŽŶŽĨĂůďƵŵŝŶ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϳϴ ůŝŶŝĐĂůŵĞĂƐƵƌĞŵĞŶƚŽĨĂůďƵŵŝŶ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϴϬ ůŝŶŝĐĂůŵĞĂƐƵƌĞŵĞŶƚƵƐŝŶŐƌĂĚŝŽůĂďĞůůĞĚĂůďƵŵŝŶ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϴϮ ůďƵŵŝŶĂƐĂďŝŶĚŝŶŐƉƌŽƚĞŝŶ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϴϯ ůďƵŵŝŶĂƐĂŶĂŶƚŝͲŽdžŝĚĂŶƚ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϴϯ ůďƵŵŝŶĂŶĚĐĂƉŝůůĂƌLJƉĞƌŵĞĂďŝůŝƚLJ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϴϯ ůďƵŵŝŶŝŶŵĂůŶƵƚƌŝƚŝŽŶ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϴϱ

š

‘–‡–•

ůďƵŵŝŶŝŶŚĞƉĂƚŝĐĚŝƐĞĂƐĞ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϴϲ ůďƵŵŝŶŝŶƉƌĞŐŶĂŶĐLJ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϴϲ ůďƵŵŝŶŝŶĞdžƚƌĂͲŚĞƉĂƚŝĐĚŝƐĞĂƐĞ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϴϲ dŚĞĐĂƐĞĨŽƌĂůďƵŵŝŶƚŚĞƌĂƉLJ͍͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϴϳ ĐƵƚĞWŚĂƐĞWƌŽƚĞŝŶƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϴϴ ZĞĨĞƌĞŶĐĞƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϴϵ  ŚĂƉƚĞƌϭϬ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϵϯ ŚĂĞŵŽĚLJŶĂŵŝĐƉĂƌĂĚŝŐŵ ŚĂƉƚĞƌƐƵŵŵĂƌLJ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϵϯ ,ŝƐƚŽƌŝĐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϵϯ dŚĞĐƵƌƌĞŶƚůLJͲƉƌĞǀĂŝůŝŶŐƉĂƌĂĚŝŐŵ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϵϲ ĂƌĚŝŽŵLJƚŚŽůŽŐLJ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϵϳ DĞĂŶĐŝƌĐƵůĂƚŽƌLJĨŝůůŝŶŐƉƌĞƐƐƵƌĞ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϵϴ sĞŶŽƵƐĞdžĐĞƐƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϭϵϵ DLJŽĐĂƌĚŝĂůĞŶĞƌŐLJĞdžĐĞƐƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϮϬϮ ĨĨĞĐƚŝǀĞǀĞƌƐƵƐŝŶĞĨĨĞĐƚŝǀĞďůŽŽĚǀŽůƵŵĞƐ͍͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϮϬϱ ĞƐĐƌŝďŝŶŐƉƌĞƐƐƵƌĞͲǀŽůƵŵĞĐŚĂƌĂĐƚĞƌŝƐƚŝĐƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϮϬϳ ƌƚĞƌŝĂůƉƌĞƐƐƵƌĞͲǀŽůƵŵĞĐŚĂƌĂĐƚĞƌŝƐƚŝĐƐŝŶƉƌĂĐƚŝĐĞ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϮϬϴ ^LJƐƚĞŵŝĐǀĂƐĐƵůĂƌǁĂƚĞƌĨĂůů͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϮϭϬ WƌĞƐƐƵƌĞͲǀŽůƵŵĞĂŶĚĨůŽǁĐŚĂƌĂĐƚĞƌŝƐƚŝĐƐŽĨƚŚĞǀĂƐĐƵůĂƌůŽŽƉ͘͘͘͘͘͘͘Ϯϭϭ ,ĂĞŵŽĚLJŶĂŵŝĐĐŽŵƉƵƚĂƚŝŽŶŵŽĚĞůƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘Ϯϭϲ ^ŚŽĐŬ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘Ϯϭϳ ZĞĨƌĂĐƚŽƌLJƐŚŽĐŬ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘Ϯϭϳ ^ŚŽĐŬŝŶŚŝůĚƌĞŶ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘Ϯϭϴ ƐƐĞƐƐŝŶŐƐŚŽĐŬĂŶĚƚŚĞƌĞƐƉŽŶƐĞƚŽƚƌĞĂƚŵĞŶƚ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘Ϯϭϴ ,ĞĂƌƚĨĂŝůƵƌĞ͖ĂŶĂƉƉƌŽƉƌŝĂƚĞŶĂŵĞĨŽƌǀĞŶŽƵƐĐŽŶŐĞƐƚŝŽŶ͍͘͘͘͘͘͘͘͘͘͘͘͘Ϯϭϵ ZĞͲĞǀĂůƵĂƚŝŶŐĐĞŶƚƌĂůǀĞŶŽƵƐƉƌĞƐƐƵƌĞŵŽŶŝƚŽƌŝŶŐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϮϮϭ ďĞƚƚĞƌŚĂĞŵŽĚLJŶĂŵŝĐƉĂƌĂĚŝŐŵ͍͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϮϮϭ ZĞĨĞƌĞŶĐĞƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϮϮϯ  ŚĂƉƚĞƌϭϭ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϮϮϵ ŶŝŵƉƌŽǀĞĚƉĂƌĂĚŝŐŵŽĨ&ůƵŝĚWŚLJƐŝŽůŽŐLJĂŶĚdŚĞƌĂƉLJ ŚĂƉƚĞƌƐƵŵŵĂƌLJ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϮϮϵ ǀĂůŝĚĐůŝŶŝĐĂůƉĂƌĂĚŝŐŵŽĨďŽĚLJǁĂƚĞƌĚŝƐƉŽƐŝƚŝŽŶĂŶĚƌĞƐƉŽŶƐĞƚŽ ŝŶƚƌĂǀĞŶŽƵƐŝŶĨƵƐŝŽŶƐŵƵƐƚĞdžƉůĂŝŶŽďƐĞƌǀĞĚĨĂĐƚƐ͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘͘ϮϮϵ DŽŽƌĞ͛ƐsŽůƵŵĞ450 kDa). A currently-marketed formulation is 6% hydroxyethyl starch 130/0.4 in 0.9% sodium chloride. Common

30

Chapter 2

side effects are stated to include itching, increased or decreased heart rate, flu-like symptoms, swollen lymph nodes, vomiting, chills, or fever. Hydroxyethyl starch deposits have been found in tissues as long as ten years after administration [15].

The liberal vs. restrictive fluid debate In the 1960s Tom Shires and Frank Moore were the proponents of liberal and restrictive perioperative fluid therapies. Tom Shires postulated the existence of a third space in surgical patients which drained fluid from the extracellular space and had to be compensated by liberal infusion. Moore held the view that intravenous infusions were undesirable in the perioperative period and would aggravate oedema. In 1967 they co-authored a famous Editorial entitled Moderation, in which they concurred that fluid should only be used to compensate for losses or for oedema and that “inundation” was unwise. They recommended; “… careful assessment of the patient’s situation, the losses incurred, and the physiologic needs in replacement. The objective of care is restoration to normal physiology and normal function of organs, with a normal blood volume, functional body water, and electrolytes.”[16]

This truce did not bring an end to the debate, and merged with the colloid vs crystalloid debate that even today has experts claiming that the modestly reduced volume requirement when colloids are used to prevent or reverse signs of blood hypovolaemia ought to confer a tangible clinical outcome advantage. There is at least now widespread acceptance that Tom Shire’s third space phenomenon was greatly overstated.

Intensive Care Units With the establishment and roll-out of intensive care units through the 1960s and 1970s it became possible to stabilise and to study patients with shock [17]. Diagnostic right heart catheterisation [18] and cardiac output determination by thermodilution [19] were pioneered at St Thomas’ Hospital, London and made ubiquitous by the development of the Swan Ganz catheter in California [20–22]. At the 1978 Hyland Symposium on pulmonary oedema Civetta explained how physiological

Starling’s Fluid Physiology in Clinical Practice

31

considerations lead to the surprising conclusion that oncotically active substances can only serve to enhance the formation of interstitial oedema [23]. The following year Virgilio presented data from surgical patients which “seriously question the necessity to maintain colloid osmotic pressure by using protein solutions during acute haemodynamic resuscitation” [24], and Lucas and Ledgerwood even claimed that albumin infusion appeared to have negative inotropic effects [25]. The thermal - green dye technique for the estimation of extravascular lung water at the bedside [26] was a particularly powerful tool for the investigation of the effects of fluid therapy. Using the new technology to measure extravascular lung water volume in traumatised and resuscitated patients, Tranbaugh and Lewis concluded in 1982 that elevations in capillary hydrostatic pressure and capillary permeability alterations resulting from lung contusion or sepsis are the primary determinants of interstitial fluid accumulation [27]. By 1983 Tranbaugh and Lewis were able to state firmly that analysis of the Starling microvascular forces operative in the lung did not provide a reason to prefer colloid resuscitation, and they had data confirming that crystalloids were both safe and effective. Of colloids, they said “one wonders how their further use can be justified.”[28]

Hypervolaemic hyperdynamic therapy In an observational study of septic shock patients, Abraham and colleagues claimed to find sequential cardiorespiratory patterns that preceded the hypotensive nadir of the illness. In particular they noted increases in cardiac index but decreases in oxygen delivery (DO2) and oxygen consumption (VO2) before as well as at the time of the hypotensive crisis [29]. In a similar exercise for a larger cohort of highrisk surgical patients the same Los Angeles team convinced themselves that non-survivors could be characterised by; • • • •

lower cardiac index (CI) in the presence of higher right and left ventricular filling pressures, increased alveolar-arterial oxygen content difference and pulmonary shunt fraction increased pulmonary artery pressure and decreased oxygen delivery despite normal arterial blood gases and haemoglobin values. [30]

32

Chapter 2

They tested a hypothesis that increased CI and DO2 are circulatory compensations for increased postoperative metabolism in a prospective but unblinded trial [31]. They concluded that patients treated according to pulmonary artery catheter derived data had reduced complications, duration of hospitalisation, duration in intensive care, and mechanical ventilation, and reduced costs. There was broad acceptance of their claim that survival from high-risk noncardiac surgery was optimised with elevated plasma volume, CI, DO2 greater than 600 ml/min.m2, and VO2 greater than 170 ml/min.m2. In contrast, these values were relatively normal in patients who subsequently died. The optimal goals were more easily attained with colloids, red cells, dobutamine, and vasodilators, according to their capacity to improve tissue perfusion, as reflected by increased flow and oxygen transport [32]. There arose discussion about the adverse consequences of so-called oxygen debt [33] The hyperdynamic approach to septic shock followed [34]. Enthusiasm was soon tempered by the results of a controlled trial performed at St Bartholomew’s Hospital showing that the use of dobutamine to boost the cardiac index and systemic oxygen delivery failed to improve the outcome in a heterogeneous group of critically ill patients and may have been detrimental [35].

Physiology laboratory advances pass unappreciated We have noted in the previous chapter J Rodney Levick’s challenge to the standard teaching of transvascular filtration in a review article of 1991 [36]. Absorption cannot be maintained across most low-pressure exchange segments because there is a rise in pericapillary interstitial oncotic pressure as filtration slows and then ceases, keeping the transvascular oncotic pressure difference smaller than the reduced hydrostatic pressure difference. The critical care community did not notice this remarkable revision of fluid physiology, which makes an oncotic pressure therapy for resuscitation unlikely to be effective for more than a few minutes. In neonatal practice Emery, Greenough and Gamsu asked the pertinent question whether it is the dose of colloid (albumin) or the volume of the solution that achieves resuscitation [37]. Their subjects, sixty hypotensive preterm infants, received their resuscitation fluid at 5 ml kg-1 h-1. Twenty received the full dose of colloid as 20% Human Albumin Solution delivered in one hour, while twenty received plasma and twenty received 4.5% Human Albumin

Starling’s Fluid Physiology in Clinical Practice

33

Solution 15 ml kg-1 h-1 over three hours. The total dose of protein was therefore similar between groups, while the hyperoncotic albumin group received a smaller volume. The degree of resuscitation was similar in all three groups after one hour, but after three hours restoration of blood pressure was better in the higher-volume (15 ml kg-1) groups. They thereby showed that it is the volume infused rather than colloid load that is important in producing a sustained increase in blood pressure. The next question was whether albumin was necessary at all, and it was answered in 1997 by So, Fok, Ng and colleagues. Sixtythree hypotensive preterm neonates were treated with an infusion of either 5% albumin or isotonic sodium chloride at 10 ml kg-1 h-1 until they were adequately resuscitated. Albumin did not reduce the volume of fluid required to achieve resuscitation, and it was associated with post-resuscitation fluid retention [38].

Have a CIGAR Edwin Cohn’s belief in the resuscitative powers of the natural colloid albumin was shown to be misplaced by the Cochrane Injuries Group Albumin Researchers (CIGAR 1998). Analysing 30 randomised controlled trials involving 1419 patients, they concluded; “There is no evidence that albumin administration reduces mortality in critically ill patients with hypovolaemia, burns, or hypoalbuminaemia and a strong suggestion that it may increase mortality.” [39]

Such conclusions were largely met by disbelief. Peter Horsey most persuasively criticised the new evidence-based Cochrane process [40] and the colloid - crystalloid debate continued. Another blow was dealt in 2004 by two landmark papers. In a large randomised clinical trial comparing albumin and saline for fluid resuscitation in a general intensive care patient population it was observed that the volume of human serum albumin solution required to achieve resuscitation on the first day (mean 1.2 litres) was only a little less than the effective volume of 0.9% Sodium Chloride (mean 1.6 litres), and that albumintreated patients received more red cell transfusions in the first two days [41]. In Adamson’s California laboratory it was shown that “colloid osmotic forces opposing filtration across non-fenestrated continuous capillaries are developed across the endothelial glycocalyx and that the oncotic pressure of interstitial fluid does not directly determine fluid

34

Chapter 2

balance across microvascular endothelium.”[42] Since 2004 Starling’s hypothesis that “absorption from the tissues ... ensues on any general lowering of capillary pressures” is untenable, and biophysical colloid osmotic pressure therapy in resuscitation is no longer supported by physiological reason. A prospective cohort study involving more than 1,000 patients found an association between hyperoncotic resuscitation fluids and “renal events” [43]. In 2009 researchers in Amsterdam confirmed by clinical experiments in both septic and non-septic patients that reducing colloid osmotic pressure of plasma does not predispose to pulmonary oedema, and that colloid resuscitation does not reduce the risk [44]. By 2011 the Cochrane Injuries Group could expand the number of randomised controlled trials of albumin to 38 and the number of enrolled patients to 10,842 including 1,958 deaths. They advised that albumin should henceforth only be used within the context of well concealed and adequately powered randomised controlled trials [45].

The J curve and the J point Here I present the original J curve cartoon for clinicians, which illustrates how in steady-state conditions volume filtration is preserved even at the low transendothelial pressure differences that were once believed to allow absorption of tissue fluid to the plasma under a dominant colloid osmotic pressure difference. Raising the plasma colloid osmotic pressure while the transendothelial pressure difference is below the inflection (J point), such as occurs in hypovolaemia, has little effect on the filtration rate [46]. (Figure 2.1) The J point concept helps to explain plasma volume and capillary pressure homeostasis and underpins context-sensitive volume expansion by fluid solutions. The modulation of capillary hydrostatic pressure through the appropriate use of arteriolar vasoconstrictors can enhance the effectiveness of fluid infusion and thereby reduce detrimental effects that accompany excessive fluid administration [47, 48].

Starling’s Fluid Physiology in Clinical Practice

35

Figure 2.1

Fluid bolus therapy A study of hypotensive febrile children in resource-poor African treatment centres found no benefit for 5% albumin resuscitation over 0.9% Sodium Chloride, but more remarkably found harm for fluid boluses compared to a fixed rate intravenous fluid infusion [49]. Several interesting explanations were offered, but the most obvious is that bolus therapy will cause transient elevation of capillary hydrostatic pressures, both systemic and pulmonary, leading to hyperfiltration and consequent interstitial fluid accumulation which cannot be reabsorbed when the capillary pressure drops back. Ultrasound examination of the lungs of children receiving fluid bolus therapy in an Australian hospital confirmed that boluses were increasing pulmonary oedema [50]. Laboratory experiments on septic merino ewes showed that a fluid bolus of isotonic salt solution increased the arterial pressure by about 6% and cardiac output by about 11%, but at the expense of a staggering 240% increase in central venous pressure [51]. What appeared to be improvements in renal

36

Chapter 2

function were transient and diminished with subsequent boluses. Unsurprisingly, study animals retained 70% of the volume and 80% of the sodium infused.

Liberal vs restrictive fluid therapy revisited In 2010 a meta-analysis of post-operative intravenous fluid volume randomised trials could find only nine, involving 801 patients undergoing elective open abdominal surgery. They could find no signal of harm or benefit until they re-analysed data as fluid balance (inputoutput) rather than just input. Patients managed in a state of fluid balance experienced significantly fewer complications (risk ratio 0·59 (95% CI 0·44, 0·8 1)) and a shorter length of stay than those managed in a state of fluid imbalance [52]. The Restrictive versus Liberal Fluid Therapy in Major Abdominal Surgery (RELIEF) trial involving 3,000 patients found no significant difference in the rate of disability-free survival among patients who received a restrictive fluid regimen (intraoperative and 1st 24 hours γ 2.5 l input, interquartile range 2.9-4.9) and those who received a more liberal fluid regimen (γ 5.5 l, 5-7.4). Clinicians were allowed to use goaldirected hemodynamic devices according to local practices [53]. They opined that clinical acumen, combined with an appropriate response to physiological variables, is crucial to guide perioperative fluid balance. As Varadhan and Lobo had suggested, zero or marginally positive fluid balance strategy appeared to avoid both the haemodilution of liberal fluid regimens and the renal injury attributable to very restrictive protocols.

Biophysical Colloid Osmotic Pressure therapy scepticism gains traction In the hope that hydroxyethyl starch might succeed where albumin had failed, a large randomised controlled trial was published in 2012 and confirmed that the plasma substitute hydroxyethyl starch causes more harm than isotonic salt solution and has very little volume advantage in clinical resuscitation [54]. In 2013 it was shown that peri-operative stroke volume optimisation goal-directed fluid therapy is possible with crystalloid, and there is no evidence of a benefit in using hydroxyethyl

Starling’s Fluid Physiology in Clinical Practice

37

starch. The authors declared that “the concept of the 1:3 replacement ratio in hypovolaemic patients is obsolete.” [55] ALBIOS, published in 2014, was an open-label multicentre trial that randomised intensive care patients with sepsis to receive an immediate 60 g (300 ml) 20% human albumin solution while resuscitation crystalloids were given to achieve a venous oxygen saturation goal. Albumin-treated patients showed no reduction in the resuscitation volume required in the first day, and subsequent daily administration of human serum albumin to raise the plasma albumin concentration to 30 g l-1 had no oedemasparing effect as assessed by net fluid balance. Like drowning men grasping at the straw of a post-hoc analysis, the authors nonetheless went on to declare albumin the ideal colloid because it maintains plasma volume and to advocate a role for albumin in the treatment of shock [56]. An analysis of data from more than 11500 surgical patients has confirmed an association between exposure to hyperoncotic albumin in the first 48 hours of postoperative shock and acute kidney injury [57]. Regulatory bodies around the world, including the European Medicines Agency, have imposed severe restrictions on the licensed use of hydroxyethyl starch solutions (EMA/35795/2018 corr. 1), leaving albumin, gelatins and dextrans as biophysical osmotic therapy options [58]. At McMaster University Medical Centre in Ontario, the sudden switch from hydroxyethyl starch to lactated Ringer solution in cardiac perioperative practice was associated with shorter length of stay for patients and no evident detriment [59]. Dextran 70 is currently the only plasma substitute on the World Health Organization's List of Essential Medicines, the most effective and safe medicines needed in a health system. The Cochrane collaboration continues to advise against the use of all plasma substitutes and albumin for resuscitation. In this volume I will explore a new Tentative Theory (in the style of Karl Popper) of fluid physiology more fully and put it in the context of a modern clinical practice.

References 1. Staff. Secrets of the Norfolk prison. 2013. https://www.boston.com/uncategorized/noprimarytagmatch/20 13/01/13/secrets-of-the-norfolk-prison. accessed May 2019.

38

Chapter 2

2. Rosen FS. Edwin J. Cohn and the Development of Protein Chemistry. N Engl J Med. 2003;349:511-512. 3. Karcs S. Harvard Magazine. https://hms.harvard.edu/magazine/ethics/backstory. Accessed May 2019 4. Siegel DC, Cochin A, Geocaris T, Moss GS. Effects of saline and colloid resuscitation on renal function. Ann Surg. 1973;177:51-57. 5. Moss GS, Das Gupta TK, Newson B, Nyhus LM. Effect of hemorrhagic shock on pulmonary interstitial sodium distribution in the primate lung. Ann Surg. 1973;177:211-221. 6. Drinker CK, Yoffey JM. Lymphatics, lymph and lymphoid tissue. Cambridge: Harvard University Press; 1941 7. Cope O, Litwin SB. Contribution of the lymphatic system to the replenishment of the plasma volume following a hemorrhage. Ann Surg. 1962;156:655-667. 8. Bull JP, Ricketts C. Dextran as a plasma substitute. Lancet. 1949;1:134-143. 9. Haynes BW, De Bakey ME. Evaluation of plasma substitutes in clinical shock; dextran. Surg Forum. 1951;631-642. 10. Preventing shock with dextran. Br Med J. 1951;2:591-592. 11. Hedin H, Richter W, Messmer K, Renck H, Ljungström KG, Laubenthal H. Incidence, pathomechanism and prevention of dextran-induced anaphylactoid / anaphylactic reactions in man. Dev Biol Stand. 1980;48:179-189. 12. Murray GF, Solanke T, Thompson WL, Ballinger WF. Hydroxyethyl starch as a plasma expander in hemorrhagic shock. Surg Forum. 1965;16:34-35. 13. Brickman RD, Murray GF, Thompson WL, Ballinger WF. The antigenicity of hydroxyethyl starch in humans. Studies in seven normal volunteers. JAMA. 1966;198:1277-1279. 14. Ballinger WF, Murray GF, Morse EE. Preliminary report on the use of hydroxyethyl starch solution in man. J Surg Res. 1966;6:180-183. 15. Wiedermann CJ, Joannidis M. Accumulation of hydroxyethyl starch in human and animal tissues: a systematic review. Intensive Care Med. 2014;40:160-170. 16. Moore FD, Shires G. Moderation. Ann Surg. 1967;166:300-301. 17. Pearce DJ. Experiences in a small respiratory unit of a general hospital with special reference to the treatment of tetanus. Anaesthesia. 1961;16:308-316. 18. Bradley RD. Diagnostic right-heart catheterisation with miniature catheters in severely ill patients. Lancet. 1964;2:941-942.

Starling’s Fluid Physiology in Clinical Practice

39

19. Branthwaite MA, Bradley RD. Measurement of cardiac output by thermal dilution in man. J Appl Physiol. 1968;24:434-438. 20. Swan HJ, Ganz W, Forrester J, Marcus H, Diamond G, Chonette D. Catheterization of the heart in man with use of a flow-directed balloon-tipped catheter. N Engl J Med. 1970;283:447-451. 21. Forrester JS, Ganz W, Diamond G, McHugh T, Chonette DW, Swan HJ. Thermodilution cardiac output determination with a single flowdirected catheter. Am Heart J. 1972;83:306-311. 22. Santora T, Ganz W, Gold J et al. New method for monitoring pulmonary artery catheter location. Crit Care Med. 1991;19:422426. 23. Civetta JM. A new look at the Starling equation. Crit Care Med. 1979;7:84-91. 24. Virgilio RW, Rice CL, Smith DE et al. Crystalloid vs. colloid resuscitation: is one better? A randomized clinical study. Surgery. 1979;85:129-139. 25. Dahn MS, Lucas CE, Ledgerwood AM, Higgins RF. Negative inotropic effect of albumin resuscitation for shock. Surgery. 1979;86:235241. 26. Sivak ED, Starr NJ, Graves JW, Cosgrove DM, Borsh J, Estafanous FG. Extravascular lung water values in patients undergoing coronary artery bypass surgery. Crit Care Med. 1982;10:593-596. 27. Tranbaugh RF, Elings VB, Christensen J, Lewis FR. Determinants of pulmonary interstitial fluid accumulation after trauma. J Trauma. 1982;22:820-826. 28. Tranbaugh RF, Lewis FR. Crystalloid versus colloid for fluid resuscitation of hypovolemic patients. Adv Shock Res. 1983;9:203216. 29. Abraham E, Shoemaker WC, Bland RD, Cobo JC. Sequential cardiorespiratory patterns in septic shock. Crit Care Med. 1983;11:799-803. 30. Bland RD, Shoemaker WC, Abraham E, Cobo JC. Hemodynamic and oxygen transport patterns in surviving and nonsurviving postoperative patients. Crit Care Med. 1985;13:85-90. 31. Shoemaker WC, Appel PL, Kram HB, Waxman K, Lee TS. Prospective trial of supranormal values of survivors as therapeutic goals in highrisk surgical patients. Chest. 1988;94:1176-1186. 32. Shoemaker WC, Appel PL, Kram HB. Measurement of tissue perfusion by oxygen transport patterns in experimental shock and in high-risk surgical patients. Intensive Care Med. 1990;16 Suppl 2:S135-44.

40

Chapter 2

33. Shoemaker WC, Appel PL, Kram HB. Role of oxygen debt in the development of organ failure sepsis, and death in high-risk surgical patients. Chest. 1992;102:208-215. 34. Shoemaker WC, Appel PL, Kram HB, Bishop M, Abraham E. Hemodynamic and oxygen transport monitoring to titrate therapy in septic shock. New Horiz. 1993;1:145-159. 35. Hayes MA, Timmins AC, Yau EH, Palazzo M, Hinds CJ, Watson D. Elevation of systemic oxygen delivery in the treatment of critically ill patients. N Engl J Med. 1994;330:1717-1722. 36. Levick JR. Capillary filtration-absorption balance reconsidered in light of dynamic extravascular factors. Exp Physiol. 1991;76:825857. 37. Emery EF, Greenough A, Gamsu HR. Randomised controlled trial of colloid infusions in hypotensive preterm infants. Arch Dis Child. 1992;67:1185-1188. 38. So KW, Fok TF, Ng PC, Wong WW, Cheung KL. Randomised controlled trial of colloid or crystalloid in hypotensive preterm infants. Arch Dis Child Fetal Neonatal Ed. 1997;76:F43-6. 39. Cochrane IGAR. Human albumin administration in critically ill patients: systematic review of randomised controlled trials. BMJ. 1998;317:235-240. 40. Horsey P. Albumin and hypovolaemia: is the Cochrane evidence to be trusted. Lancet. 2002;359:70-2; discussion 72. 41. Finfer S, Bellomo R, Boyce N et al. A comparison of albumin and saline for fluid resuscitation in the intensive care unit. N Engl J Med. 2004;350:2247-2256. 42. Adamson RH, Lenz JF, Zhang X, Adamson GN, Weinbaum S, Curry FE. Oncotic pressures opposing filtration across non-fenestrated rat microvessels. J Physiol. 2004;557:889-907. 43. Schortgen F, Girou E, Deye N, Brochard L, CRYCO SG. The risk associated with hyperoncotic colloids in patients with shock. Intensive Care Med. 2008;34:2157-2168. 44. van der Heijden M, Verheij J, van Nieuw Amerongen GP, Groeneveld AB. Crystalloid or colloid fluid loading and pulmonary permeability, edema, and injury in septic and nonseptic critically ill patients with hypovolemia. Crit Care Med. 2009;37:1275-1281. 45. Albumin Reviewers (Alderson P BF, Li Wan Po A, Li L, Blackhall K, Roberts I, Schierhout G. Human albumin solution for resuscitation and volume expansion in critically ill patients. Cochrane Database Syst Rev. 2011CD001208.

Starling’s Fluid Physiology in Clinical Practice

41

46. Woodcock TE, Woodcock TM. Revised Starling equation and the glycocalyx model of transvascular fluid exchange: an improved paradigm for prescribing intravenous fluid therapy. Br J Anaesth. 2012;108:384-394. 47. Tatara T. Context-sensitive fluid therapy in critical illness. J Intensive Care. 2016;4:20. 48. Woodcock TE. Plasma volume, tissue oedema, and the steady-state Starling principle. BJA Education. 2017;17:74-78. 49. Maitland K, Kiguli S, Opoka RO et al. Mortality after fluid bolus in African children with severe infection. N Engl J Med. 2011; 364:2483-2495. 50. Long E, O’Brien A, Duke T, Oakley E, Babl FE, Pediatric RIEDIC. Effect of Fluid Bolus Therapy on Extravascular Lung Water Measured by Lung Ultrasound in Children With a Presumptive Clinical Diagnosis of Sepsis. J Ultrasound Med. 2018 51. Lankadeva YR, Kosaka J, Iguchi N et al. Effects of Fluid Bolus Therapy on Renal Perfusion, Oxygenation, and Function in Early Experimental Septic Kidney Injury. Crit Care Med. 2019;47:e36-e43. 52. Varadhan KK, Lobo DN. A meta-analysis of randomised controlled trials of intravenous fluid therapy in major elective open abdominal surgery: getting the balance right. Proc Nutr Soc. 2010;69:488-498. 53. Myles PS, Bellomo R, Corcoran T et al. Restrictive versus Liberal Fluid Therapy for Major Abdominal Surgery. N Engl J Med. 2018;378:2263-2274. 54. Myburgh JA, Finfer S, Bellomo R et al. Hydroxyethyl starch or saline for fluid resuscitation in intensive care. N Engl J Med. 2012; 367:1901-1911. 55. Yates DR, Davies SJ, Milner HE, Wilson RJ. Crystalloid or colloid for goal-directed fluid therapy in colorectal surgery. Br J Anaesth. 2014;112:281-289. 56. Caironi P, Langer T, Gattinoni L. Albumin in critically ill patients: the ideal colloid. Curr Opin Crit Care. 2015;21:302-308. 57. Udeh CI, You J, Wanek MR et al. Acute kidney injury in postoperative shock: is hyperoncotic albumin administration an unrecognized resuscitation risk factor. Perioper Med (Lond). 2018;7:29. 58. CMDh. Hydroxyethyl-starch solutions for infusion to be suspended – CMDh endorses PRAC recommendation. Suspension due to serious risks of kidney injury and death in certain patient populations. 2018.

42

Chapter 2

https://www.ema.europa.eu/news/hydroxyethyl-starchsolutions-infusion-be-suspended-cmdh-endorses-pracrecommendation. Accessed May 2019 59. Hong M, Jones PM, Martin J et al. Clinical impact of disinvestment in hydroxyethyl starch for patients undergoing coronary artery bypass surgery: a retrospective observational study. Can J Anaesth. 2019;66:25-35.

CHAPTER 3 BIOLOGICAL SOLUTIONS

Chapter Summary An appreciation of the distribution of water (the solvent) and its solutes is fundamental to understanding the physiology of body fluid spaces. Peter Stewart’s quantitative approach to acid - base balance is adopted here. The presence of carbon dioxide and ionic solutes in body water compartments determines the extent to which hydrogen and hydroxide ions are available for chemical reactions, the acid-base status. The compartmentalization of body water and fluid flux across cell membrane barriers is determined by hydrostatic pressure differences and solute concentration gradients. Total body water volume is largely conserved by the anti-diuretic hormone effect of arginine vasopressin acting on vasopressin type 2 receptors in renal collecting ducts. In critical illness, arginine vasopressinaemia predisposes patients to dilutional hyponatraemia if the infused volume is larger than necessary or excessively prescribed as “electrolyte-free water”. Body sodium is conserved by aldosterone regulating the degree of sodium reabsorption in renal distal tubules. Body sodium largely determines the proportion of body water that comprises extracellular fluid volume, but there is significant non-osmotic sodium storage capacity in the interstitium, particularly in the interstitium of the skin, which may have clinical relevance. In addition, volume regulated anion channels enable cells to discharge osmotic molecules to the interstitium to protect intracellular fluid volume when body water tonicity is low. Balancing the body potassium (mostly intracellular) with sodium (mostly extracellular) depends on an adequate availability of magnesium. Rapid extracellular fluid osmolality changes can dangerously disturb the intracellular - extracellular fluid equilibrium, so awareness of the major contributors to plasma osmolality is essential. However, evidence from surgical practice suggests that adaptive mechanisms exist to stabilise

44

Chapter 3

the intracellular volume in the face of excessive intravenous fluid infusions, and an alternative model of body water response to intravenous infusions has been proposed.

Water and the hydrogen ion concentration Water is the important solvent for biological fluids which can therefore be considered aqueous solutions of several solutes. Salts exist only as ions, never as charge-free molecules, in biological solution. Biological solutions are always electrically-neutral, that is to say that anions always balance cations. While more advanced scientists may distinguish ionic activity and ionic concentration, for our purposes we will consider them to be one and the same thing. Canadian physiologist Peter Stewart proposed a quantitative approach to whole body acid base analysis for biologists and clinicians in 1978 [1] [2, 3]. The important claims of the Stewart approach are as follows; 1. Acid-base balance for aqueous physiological solutions should be defined by the ratio of hydroxyl ion to hydrogen ion; a solution is acidic when the hydrogen ion concentration is higher than the hydroxyl ion concentration, and neutral when the concentrations are equal. 2. The negative logarithm of the hydrogen ion concentration (pH) is an unhelpful representation of the hydrogen ion concentration [H+], which is better considered directly. 3. The contribution of strong electrolyte cations and anions to [H+] and other dependent acid-base variables can be adequately estimated by the net strong ion difference (SID). 4. [H+] is determined by three independent variables; ် the strong ion difference, ် the total weak acid present, and ် the partial pressure of carbon dioxide (pCO2). 5. Bicarbonate ion concentration [HCO3-] is a dependent variable. None of the dependent variables determines any other dependent variable, although their quantitative behaviours are necessarily correlated. An infusion of sodium bicarbonate into blood alkalinises because it increases the strong ion difference between sodium and chloride. 6. Solutions separated by membranes can interact in acid-base terms only by their independent variables. Interaction of

Biological Solutions

45

intracellular and extracellular acid-base balance can only occur by the cell membrane altering these independent variables in the extracellular and intracellular fluid. It should be noted that pH was originally conceived to be a measure of the hydrogen ion concentration of aqueous solutions, but such free protons quickly react with water to form hydronium (H3O+); acidity of an aqueous solution is therefore more accurately characterized by its hydronium concentration. For biologic considerations the hydronium ion can be used interchangeably with the hydrogen ion; choosing one over the other has no significant effect on the mechanism of reaction and so for the sake of tradition and consistency I shall stick with [H+].

pH or [H+]? It is important to distinguish the clinical terminology which takes a blood pH of 7.4 ([H+] circa 40 nmol l-1) to be normal, and so describes acidaemia to be present when [H+] > 40 nmol l-1 and alkalaemia when [[H+] < 40 nmol l-1. Notice that the hydrogen ion concentration at the neutral pH 7.0 is 100 nmol l-1. In chemistry terms, blood and extracellular fluid is therefore normally an alkaline fluid with low hydrogen ion concentration. The intracellular fluid hydrogen ion concentration is greater than extracellular, and [HCO3-] lower, because of the strong ion differences. Mammalian skeletal muscle intracellular fluid has a higher [H+] than most at around 100 nmol l-1 (pH 6.8-7.1). A 50% increase of extracellular [H+] to 60 nmol is a serious acidaemia though it only causes a small change in the arterial pH from 7.40 to 7.22. Systemic vascular resistance falls, pulmonary artery pressure rises, cardiac contractility falls, and the haemoglobin - oxygen dissociation changes impair tissue oxygenation. It is a curious and useful fact that the activities of clotting factors are inversely proportional to hydrogen ion concentration, so a priority in treating the bleeding patient is to reverse acidaemia. Though hydrogen ions can easily be viewed as A Bad Thing because they threaten to denature proteins by attaching to histidine, they have a vital role in delivering the constant supply of cell energy that is life.

46

Chapter 3

Mitochondrial hydrogen ion gradient Every cell contains mitochondria, which are continually storing vital energy into adenosine triphosphate (ATP) molecules, the process known as oxidative phosphorylation of adenosine diphosphate (ADP) that occurs within the mitochondrial matrix. The electron transport chain of proteins is bound into the inner mitochondrial membrane and as each protein in turn passes electrons along, the protein pumps a hydrogen ion (proton) from the mitochondrial matrix out into the space between the inner and outer mitochondrial membranes - the mitochondrial intermembrane space. The “chemiosmotic coupling hypothesis” was proposed by Peter D. Mitchell, winning him a Nobel Prize in 1978. It explains how the electron transport chain and oxidative phosphorylation are coupled by the proton gradient across the mitochondrial inner membrane. The outward flow of protons from the mitochondrial matrix to the intermembrane space creates an electrochemical gradient (proton gradient) enabling proton inward flow through the ATP synthase complex, oxidatively phosphorylating ADP to ATP within the mitochondrial matrix.

Acids and bases A substance is called an acid if it raises [H+] when added to a solution and a base if it lowers [H+]. An acid is alternatively defined as a compound that dissociates into a hydrogen ion and an anion that is called the conjugate base; it is therefore a hydrogen ion (or proton) donor. We can turn that on its head and also state that a base is an anion capable of binding a hydrogen ion.

Solutions and Solutes We can classify the solutes as electrolytes or non-electrolytes. The electrolytes are subdivided into strong electrolytes which are completely dissociated in solution or weak electrolytes that are partly dissociated and obey a dissociation equilibrium. Non-electrolytes do not dissociate into ions in solution, and molecules of salts do not exist in solution. Solutions must at all times be electrically neutral. The movement of water between compartments is

Biological Solutions

47

passive and so the distribution of water depends on solute concentrations. Osmolar gradients determine the intracellular to extracellular distribution of water. With little or no osmolar gradient between plasma and the interstitium, the colloid osmotic pressure gradient of larger molecules determines the intravascular to interstitial distribution of water.

Strong electrolytes Sodium ion. The normal range of extracellular fluid sodium concentration is about 135-145 mmol l-1. It is estimated that one third of the total body sodium is bound into bone, but two thirds are present in exchangeable forms in the extracellular fluids. Renal sodium reabsorption under the influence of aldosterone occurs in the distal convoluted tubules and maintains both extracellular fluid volume and osmolarity. The Paradox of Sodium's Volume of Distribution, which looks at that exchangeable two thirds, is that the major extracellular cation sodium has a rather larger volume of distribution than inulin or mannitol, which are polysaccharides presumed to be confined to the extracellular fluid compartment. There is emerging evidence that this anomaly is largely due to sodium’s ionic affinity for the glycosaminoglycan called hyaluronate. Sodium hyaluronate (hyaluronan) is distributed widely in the extracellular matrix of mammalian connective, epithelial, and neural tissues, as well as intravascular endothelial glycocalyx. It has often been taught that excess sodium intake inevitably leads to oedema, but the non-ionic storage of excess sodium in surgical patients was first observed in 1986 [4]. The clinical importance of this second sodium store is emerging. Chloride ion. Chloride is normally present in extracellular fluid at a concentration of 95-105 mmol l-1. The volume of distribution of the major extracellular anion chloride is also greater than inulin or mannitol, though not as great as sodium. In recent years chlorine has found itself at the centre of a Hero or Villain debate created by champions of lower-chloride isotonic salt solutions for volume resuscitation. As the cumulative dose of any isotonic salt solution rises the incidence of hyperchloraemia amongst patients rises. Two large pragmatic trials conducted in Nashville, TN were reported in 2018 and showed that hyperchloraemia is a common complication of intravenous fluid therapy with Normal Saline or Ringers Lactate or

48

Chapter 3

Plasma-Lyte A®. Hyperchloraemia occurred in 15-20% of non-critically ill patients [5] and 25-35% of critically ill patients [6]. Even though clinical studies have not proven that dyschloraemia (hypochloraemia 110 mmol l-1) is of itself harmful, an association between hyperchloraemia and indices of acute kidney injury has created widespread concern and a general preference for isotonic salt solutions with less chloride than 0.9% NaCl [7]. One of several recent suggestions is that neutrophil dependent, myeloperoxidase mediated inflammatory conditions such as sepsis, hypovolaemic shock and reperfusion injury could be aggravated by greater availability of hypochlorous acid (HOCl) that targets membrane lipids to create toxic chlorinated lipids [8]. The acidity of the stomach depends on net flux of chloride ion from plasma in the gastric capillaries to the stomach lumen. Note that proton pumping may be a part of that process, but the principle of electric charge neutrality means that gastric luminal [H+] cannot increase without an accommodating increase in chloride anion. Potassium and magnesium ions. Potassium is the major intracellular cation and 98% of the total body potassium is intracellular. The plasma levels of potassium and magnesium are generally poor indicators of the whole-body content of these electrolytes which are the major intracellular cations, but deficiency does eventually manifest as reduced plasma concentrations. The vital membrane-bound enzyme Na/K ATPase maintains the transcellular gradient of sodium and potassium concentrations. The non-availability of oxygen to generate sufficient ATP to fuel this enzyme is rapidly fatal (for example, cyanide poisoning) as excitatory tissues lose their transmembrane electric potential. Magnesium is an essential co-factor for the Na/K ATPase. It is common clinical experience that hypomagnesaemia limits the ability to normalise plasma potassium by giving potassium supplements. Sometimes hypokalaemia only improves after magnesium has been given. The predominant factor seems to be magnesium’s part in the working of several weak inward-rectifier potassium channels found in various isoforms along renal tubular epithelium and only discovered as recently as 1993 [9]. The renal outer medullary potassium channel (ROMK) is the prototypic member of this family, and they play a central role in the regulation of salt and potassium homeostasis. Intracellular magnesium and polyamines enter the inward-rectifier potassium channel cytoplasmic pore and plug the potassium permeation pathway,

Biological Solutions

49

giving rise to the phenomenon of ‘inward rectification’. In simple terms, intracellular magnesium blocks what would otherwise be the inward flow of potassium and so the recovery of potassium from the lumen of the distal tubule. When intracellular magnesium is depleted, the block is lifted allowing potassium to be conserved. Calcium. A bivalent ion present at around 8-10 mEq l-1 in plasma. Of little significance to acid-base balance. Lactate. While normally present at levels of no significance to acid-base balance (less than 2 mmol l-1), lactataemia is an important anionic disorder in critical care medicine.

Weak electrolytes and buffers Weak electrolytes in body fluids are mainly proteins and their normal anionic contribution to the electrolyte balance of plasma is about 6-9 mEq l-1. For the purposes of acid-base analysis weak electrolytes introduce additional equilibria in solution and therefore require more complicated mathematical analysis. Stewart commented that their concentrations normally vary by small amounts, and only slowly, so could in most cases be disregarded. He clearly had no experience of critical care where protein (and particularly albumin) does have to be taken into account! Non-protein weak electrolytes are generally present at very small concentration (of the order of 1 mmol), and so their acid-base effects are negligible. Albumin. In most of us, albumin is the major weak acid of plasma. The hypoalbuminaemia of critical illnesses is an alkalotic state and is one of the more frequent recorded ’abnormalities’ seen on the arterial blood gas analyser in a clinical facility. Albumin and other plasma proteins normally make up about 7% of the plasma volume. Consequently, a reported plasma sodium concentration of 140 mmol l-1 of plasma corrects to (140/0.93=) 150 mmol l-1 of plasma water. In patients who are hyperproteinaemic (or hyperlipidaemic) up to 20% of plasma volume could be non-aqueous, throwing up misleading plasma ion concentration results. For example, in such a patient when sodium is 120 mmol l-1 plasma, the plasma water concentration of sodium is (120/0.8=) 150 mmol l-1.

50

Chapter 3

Phosphate. The other weak acids of note are phosphates, normally present in plasma at around 2.5-4.5 mEq per l. While phosphate abnormalities are generally rare, hypophosphataemia may affect around one in three patients in Intensive Care Units and one in two patients with severe sepsis. Treatment is offered to avoid muscle weakness (diaphragmatic, cardiac and skeletal) and altered mental status (confusion, hallucinations). Severe acute hypophosphataemia (level < 0.3 mmol l-1) may present with seizures, focal neurological deficits, congestive heart failure, and muscle pain.

Carbon dioxide in water The amount of carbon dioxide dissolved in body fluids is in an equilibrium between the rate of production (around 200 ml minute-1) and the rate of excretion by perfused alveolar ventilation. In addition, carbon dioxide reacts with water’s hydroxyl ions to form bicarbonate. Bicarbonate + hydrogen ion carbonic acid Carbon dioxide + water. In the example above bicarbonate is the base to carbonic acid. These entities of the bicarbonate buffering system are in an equilibrium in biological solutions in the presence of the enzyme carbonic anhydrase, making bicarbonate the flexible anion base that occupies the normal strong ion gap or strong ion difference. Notice that at pH 7.40, bicarbonate concentration is of the order of 25 mmol l-1 and hydrogen ion 0.00004 mmol l-1; that is more than half a million bicarbonate anions to every hydrogen ion. A patient in Edinburgh survived acidaemia of pH 6.527 due to acute haemorrhagic shock [10]. The hydrogen ion concentration was 312 nmol l-1 and bicarbonate 4 mmol l-1, that is 13,333 bicarbonate ions to every hydrogen ion. It is not discussed whether bicarbonate solution was infused in this case, but it is important to understand that the main alkalinising effect of infused sodium bicarbonate solution (or sodium hydroxide) is raising the sodium strong ion concentration without affecting the chloride strong ion. The equilibrium presented also reminds us that if we reduce the concentration (tension) of carbon dioxide in the blood by hyperventilation, we will reduce the hydrogen ion concentration; respiratory alkalosis. In the case of severe hypovolaemic shock presented here, pulmonary blood flow will be low and excreting carbon

Biological Solutions

51

dioxide by perfused alveolar ventilation will not be easy! Bicarbonate (or "total CO2" or "carbon dioxide”) is included in calculation of the anion gap, which is not to be confused with the strong ion difference or strong ion gap. Anion Gap = Sodium - (Chloride + Bicarbonate), and the normal range for anion gap is 8-16 mEq l-1. Normal anion gap acidosis is associated with elevated [Cl-] and so often referred to as hyperchloraemic acidosis, but this unfortunate term tends to imply (sometimes falsely) that chloride overdose by infusion of salt solutions is the cause. The mnemonic FUSEDCARS has been suggested for the differential diagnosis: Fistula (pancreatic), Ureteroenterostomy, Saline administration, Endocrine (hyperparathyroidism), Diarrhoea, Carbonic anhydrase inhibitors (acetazolamide), Ammonium chloride, Renal tubular acidosis, Spironolactone. High anion gap metabolic acidoses (HAGMA) are associated with one or more unmeasured anions which may include lactate, ketones, various acids of renal failure, and exogenous poisons such as methanol or ethylene glycol. Patients recovering from critical illness may go through a phase of low anion gap alkalosis due to hypoalbuminaemia.

Blood gases A blood gas analyser uses just two data measurements to calculate other acid base parameters. From the hydrogen ion concentration / pH and the carbon dioxide tension a microprocessor calculates and presents the bicarbonate concentration. As explained above, bicarbonate is now revealed to be a dependent parameter of acid - base status and does not help the clinician to ascertain the cause of a nonrespiratory derangement. Ole Siggaard-Andersen’s base excess is defined as the titratable base of the blood or plasma or model of the extracellular fluid when titrating to pH = 7.40 at pCO2 = 5.33 kPa at 37 °C at constant concentration of total oxygen [11]. Direct titration must therefore be considered to be the reference method. For practical reasons it is more convenient to calculate the base excess either with a nomogram or with an arithmetic equation. Nomograms are now replaced by computer calculation and generally the calculation is based on some version of the Van Slyke equation.

52

Chapter 3

Fencl - Stewart method Stewart proposed instead that effective strong ion difference (in Equivalents l-1) be ascertained from plotting measured data on what he called the pH – pCO2 diagram. By summing the inorganic strong ion difference from measurements of [sodium] + [potassium] - [chloride] and looking at the disparity of the effective SID and the SID expected from the pH-pCO2 nomogram, more than a few mEq per litre must mean that other strong ions are present. Lactate is of course one of the organic strong electrolytes which could contribute to the summed strong ion difference. Notice that Stewart’s nomogram presumes near normality of sodium and protein concentrations, but in modern critical care major abnormalities in these are common. For Stewart, whole-body acid-base balance was the question "What is happening to circulating plasma [H+], and why? “Plasma [H+] is determined by [SID], pCO2, and weak acids. Its normal value is 40 nmol l-1. The “regulators” are therefore the kidneys, lungs and circulation, and liver. The “upsetters” include gastrointestinal tract losses, diet, tissue injury, and capillary leak. Plasma [SID] changes imposed by the upsetters are modified by plasma interaction with interstitial fluid. Interstitial fluid in turn may interact with intracellular fluid. Alveolar ventilation and the circulatory rate of blood regulate alveolar and circulating plasma pCO2. The kidneys regulate circulating plasma [SID] by differential reabsorption of Na+ and Cl-; when circulating plasma [H+] changes due to pCO2 changes, the kidneys produce compensating [SID] changes. Conversely, when circulating plasma [H+] changes due to [SID] changes, respiration in the lungs changes so as to produce compensating plasma pCO2 changes. Later writers abandoned the nomogram but addressed the clinical application of Stewart’s quantitative approach in different ways. Vladimir Fencl was a Harvard physician who extended the utility of Stewart’s quantitative method to critical care by investigating the role of proteins in human acid-base balance, and by developing a mathematical model to characterize the acid-base status of human plasma. In 2000 Fencl presented a formula that “allows one to detect and quantify even the most complex acid-base disturbances seen in critically-ill patients. All the calculations can be done at the bedside with a simple hand-held calculator.” [12]

Biological Solutions

53

Gilfix, Bisque and Magder proposed an approach for bedside interpretation that exploited the popular Base Excess concept [13]. • • • • •

[Na+] is used to calculate the Base Excess due to free water (BEfw) ် 0.3 x (Na-140) [Cl-] is used to calculate the Base Excess due to chloride (BECl) ် 102-(Cl x 140/Na) [Alb] is used to calculate the Base Excess due to albumin (BEalb) ် (0.148 x pH-0.818)(42-[alb]) The difference between the reported Base Excess and the sum of the measured contributors is called the Base Excess due to unmeasured anions (BEua] In this method, lactate would be a common “unmeasured anion”, and of course if actually measured could be used to identify other unmeasured anions.

In the absence of a calculator and for those who prefer mental arithmetic, BEalb can be approximated as (45-[alb]) / 3. [14] The FenclStewart method in one guise or another has been used to offer insights into more complex critical care scenarios like the effects of pump prime solutions [15], lactate buffered high volume haemofiltration [16] and hypoalbuminaemia with hyperchloraemia in pre-eclamptic women [17]. If you don’t fancy writing your own, a good gateway to finding spreadsheets and apps to calculate acid base parameters can be found at http://acidbase.org

Urea synthesis in acid-base balance Nitrogen metabolic waste in the form of ammonium (NH4+) is converted to urea for excretion by the kidneys. The urea cycle was the first biochemical “cycle” to be described, by Hans Krebs and Kurt Henseleit in 1932. The condensation of ammonium with bicarbonate occurs within hepatocyte mitochondria, consumes ATP, and produces carbamoyl phosphate. Carbamoyl phosphate is joined to ornithine, creating citrulline which passes into the hepatocyte’s cytoplasm to be combined with aspartate to form argino-succinate. This large molecule is then lysed to fumarate and arginine, which is in turn split into ornithine (recycled to the mitochondria) and urea which distributes through the total body water and so becomes available for renal excretion. In 1982 Atkinson and Camien pointed out that as both

54

Chapter 3

ammonium and bicarbonate are substrates for urea synthesis, the liver contributes to acid - base balance. When hydrogen ion concentration is elevated (acidaemia, low pH) urea synthesis is decreased and bicarbonate consumption slowed. Conversely when hydrogen ion concentration is low (alkalaemia, high pH) urea synthesis is accelerated and bicarbonate is consumed. Metabolism of 100 g protein yields about l mol of bicarbonate; some of the sulphur of cysteine and methionine is converted to sulphuric acid, but metabolism of proteins produces a large net amount of bicarbonate. Depending on intake, muscle mass and the catabolism/ anabolism state, somewhere between 500 and 800 mmol urea (30 - 48 g) are formed every day.

Osmotic concentration of solutes within the body water On our planet earth water is heated by hydrothermal vents and by underwater volcanic activity, and elutes minerals from the rock. The predominant sea water minerals are of course sodium and chlorine, at a concentration of around 3 to 5% NaCl. Life on earth has evolved from replicating molecules within a salt water environment to membranecontained cells of replicating molecules and finally to multicellular organisms capable of regulating the salt solution environment in which a cell and their replicating molecules can optimally function. Cell membranes are permeable to water molecules, gas molecules and smaller lipophilic molecules including drugs such as anaesthetic agents. Water passes across some cell membranes substantially faster than can be explained by simple diffusion, due to the presence of membrane channel proteins called aquaporins (AQP), discovered by Nobel Laureate Peter Agre in about 1990 [18]. Osmotic concentration (osmolarity) is the measure of solute concentration within the solvent, and is defined as the number of osmoles of solute per litre of solution (osmol l-1, or more usually in biological solutions mosmol l-1). Osmolality is an expression of solute osmotic concentration per mass of solvent. Osmolarity is osmolality multiplied by the difference between the density of the solution and the solute concentration, but as the density of biofluids is very close to 1 kg l-1 they can be considered numerically similar. Tonicity is related to osmolarity, but only takes into account the total concentration of non-membrane penetrating solutes. Tonicity therefore determines the distribution of the solvent

Biological Solutions

55

(water) across a membrane while osmolarity can account for the movement of membrane penetrating solutes. A solution for infusion is isotonic if a cell placed within it does not change volume. Hypotonic solutions cause cells to swell (net inward water movement) while hypertonic solutions shrink cells. The osmolarity of intracellular and extracellular fluid solutions are closely linked.

Urea Haemofiltration and dialysis are used to control uraemia in renal failure. Haemofiltrate contains urea at the same concentration as plasma, so every litre of filtrate from a uraemic 40 mmol l-1 patient removes 40 mmol urea. Presuming that you will need to match a high urea production of 800 mmol per day to stop urea rising in an anuric patient, you will have to draw off and replace 20 litres of filtrate per day. More than 20 litres per day will cause the plasma urea to fall. When the plasma urea is down to 30 mmol l-1 the filtration rate just to stand still is nearer to 30 litres of filtrate per day, and at 15 mmol l-1 it is 53 litres of filtrate per day. This is close to 35 ml/kg/hour for a 70 kg man, a rate once described as high volume haemofiltration [19]. Of course, as the hypermetabolic state subsides and the rate of urea production falls, the necessary daily filtrate replacement declines proportionately, so that at 500 mmol urea production per day the filtration rate to achieve 15 mmol l-1 plasma urea will be just 33 litres per day (about 20 ml/ kg/ hour if there are no breaks). By such simple arithmetic the prescriber of fluids and continuous haemofiltration can make rational decisions about the optimal daily filtration volume and predict its effect on [urea]. At steady-state urea is evenly distributed throughout the total body water, but when serum urea concentration changes acutely it can behave as an effective osmole with a Staverman reflection coefficient (ɐ) approaching 1 across cell membranes and low diffusive permeability (Pd). The properties that allow it to achieve steady-state equilibration are urea transporters in some tissues, a very large area for diffusion (the total cell membrane surface area has been estimated to be around 12 km2) and low daily rate of production (roughly 600 mmol d-1). Urea concentration also equilibrates quickly across most capillary beds, but cerebral capillaries (the blood-brain barrier) exhibit low urea permeability with a Staverman’s reflection coefficient of

Chapter 3

56

around 0.5. If the plasma urea concentration drops too acutely during haemodialysis, a significant cerebral transcapillary urea gradient can develop, raising transendothelial filtration and causing interstitial cerebral oedema (dialysis disequilibrium). Acute increase in plasma urea by infusion of exogenous urea was the first successful hyperosmolar therapy for reduction of intracranial pressure, but mannitol became the preferred agent in the 1960s. Oral urea therapy is currently recommended as a second line treatment for chronic hyponatraemia and syndrome of inappropriate anti-diuresis (SIAD).

Measuring osmolality, calculating osmolarity Clinical laboratories measure the osmolality of a fluid such as plasma by physical methods, typically by freezing point depression or by vapour pressure depression. Plasma osmolarity can be calculated, in mosmol l-1, as the sum of its major contributors in mmol l-1. Sodium and potassium are doubled to account for their balancing cations, plus glucose and urea. The difference between measured osmolality and calculated osmolarity is the osmolar gap. In the intoxicated patient with an osmolar gap >10, a useful differential is; • • •

Osmol gap + anion gap = Ethylene Glycol or Methanol Osmol gap + ketosis = Isopropyl alcohol Osmol gap + lactate = Propylene glycol

Regulation of osmolarity by water resorption (antidiuresis) Hypothalamic osmoreceptors maintain body water osmolarity in the range of 280-295 mosmol kg-1. As plasma osmolarity rises, anti-diuretic hormone is secreted into the blood stream via the posterior pituitary gland. Anti-diuretic hormone is increasingly referred to as argininevasopressin (AVP). Among other effects, AVP activates vasopressin type 2 receptors to increase water permeability of the nephron’s collecting ducts, allowing water reabsorption under the osmotic pressure gradient that exists there. Aquaporin-2 (AQP2) is exclusively expressed in the principal cells of the collecting duct and is the

Biological Solutions

57

predominant AVP-regulated water channel. AQP2 and the vasopressin type 2 receptor are therefore of major importance for the renal regulation of body water balance. Mutations in the AQP2 gene cause profound nephrogenic diabetes insipidus. Arterial baroreceptors are stimulated when the circulating volume is reduced, increasing sympathetic nervous system and reninangiotensin-aldosterone system activity, augmented by non-osmotic AVP secretion.

Regulation of osmolarity; release from antidiuresis There are other neurohumoral mechanisms that regulate body water losses by “escape from antidiuresis”. Escape from antidiuresis is caused by downregulation of aquaporin-2 expression in the face of raised AVP plasma levels. The vasopressin type 2 receptor on the collecting duct epithelial cell membrane seems to become internalised, and the underlying mechanisms may include decreased intrarenal angiotensin II signalling, in combination with increased intrarenal nitric oxide and prostaglandin E2 signalling. The clinical consequence is diuretic resistance. Solute losses occur through the secondary natriuresis induced by water retention. The natriuresis that results in volume regulation of the extracellular fluid is the result of intrarenal hemodynamic changes produced by volume expansion, and these effects are modulated by the major regulator of sodium resorption, aldosterone. The high hydraulic conductance of the proximal tubule and the descending thin limb of Henle’s Loop is attributable to an abundance of aquaporin-1 (AQP1), which accounts for almost 1% of total membrane protein in the renal cortex. AQP1 is not involved in the AVP regulation of renal water transport. AQP1 is the greatest contributor to water movement across the tubule wall, and humans lacking AQP1 are incapable of producing a maximally-concentrated urine.

Chapter 3

58

Aldosterone Aldosterone has circumstances; • •

apparently

conflicting

effects

in

different

It can increase salt reabsorption while the circulatory volume is low without significantly altering potassium balance. It can increase potassium excretion while potassium is high without altering sodium balance or the circulatory volume.

A newly described pathway, the “potassium switch” in the distal nephron, explains the potassium-dependent effects on sodium delivery [20, 21]. The aldosterone-sensitive distal nephron secretes potassium when an increase in aldosterone is accompanied by adequate sodium delivery. Potassium secretion is a two-step process. • •

Active transport of potassium from blood by sodium/ potassium ATPase, and then Passive efflux into the tubular fluid through renal outer medullary potassium channels and big potassium ion channels.

The electrical force driving potassium out of the tubular cell is generated by the epithelial sodium channel (ENaC}. Potassium excretion increases when sodium delivery to the distal nephron (and so the activity of ENaC) are increased. Aldosterone activates ENaC either in response to potassium excess or in response to intravascular volume depletion [22]. Sodium delivery to the distal nephron is different in the two situations. While the circulating volume is low, distal sodium delivery is suppressed as up-stream tubule segments reabsorb more sodium. In potassium excess with normal circulating volume, sodium delivery to the distal nephron is normal.

Arginine-vasopressinaemia in critical illness AVP secretion in response to non-osmotic influences is a normal and mostly appropriate stress response. Some patients who become hypotensive have been shown to have lower AVP in their blood than comparably stressed patients who maintain normal blood pressure [23]. Most importantly, supplementing the inadequate AVP levels by

Biological Solutions

59

administering exogenous AVP (or other vasopressin receptor agonist [24]) corrects the hypotension [25]. Correcting the relative deficiency of the anti-diuretic hormone AVP can even increase urine output while vasopressin type 2 receptors are downregulated. A really important consideration, too often overlooked, is that most of our patients sick enough to be hospitalised will have higher than normal plasma concentrations of AVP; they are appropriately arginine-vasopressinaemic. Medicines that contribute to arginine-vasopressinaemia (or cause the syndrome of anti-diuresis) include; morphine and opioid analgesics. anti-convulsants such as carbamazepine, phenytoin, sodium valproate. • dopamine agonists such as metoclopramide, prochlorperazine and various antipsychotics. • antidepressants (tricyclics and selective serotonin reuptake inhibitors) • 3,4-methylenedioxymethamphetamine (MDMA, ecstasy) One of the commoner consequences of AVP-induced water retention during physiological stress response is a slight reduction in the ECF osmolar concentrations of sodium, potassium, magnesium and albumin. It is not necessary to attempt to correct these slight abnormalities by administering more of the diluted (but not deficient) osmol. • •

An increasingly common cause of hyponatremic encephalopathy in the outpatient setting is use of the recreational drug ecstasy. It is clinically important to know that the majority of deaths or permanent neurological injuries from hyponatremic encephalopathy occur in females, possibly because of oestrogen. It has recently been shown that arginine vasopressin kickstarts erythrocyte production as part of its defence of blood volume [26]. After acute haemorrhage the erythropoietin response only commences when plasma volume compensation leads to anaemia, while the vasopressin response is triggered by hypovolaemia. Erythropoietin response is slow because it stimulates early erythrocyte progenitors. Vasopressin response is rapid because it promotes maturation of the existing intermediate erythrocyte precursors.

Chapter 3

60

Total body water The hypothetical 70 kg man has around 42 kg of total body water, around 60% of his body weight. Most tissues contain 70-80% water, the exceptions being bone and adipose tissue at 10-20%. The major contributors to variation in an individual’s total body water to weight are muscle mass (high in water content) and adiposity (fat is low in water content). This explains why females and older individuals typically have a lower percentage total body water to body weight and may be more prone to disorders of tonicity. As a rule of thumb, the total body water of women can be estimated as 50% of body weight. Total body water is nearer to 45% in elderly women and 55% in elderly men. In very muscular adults, total body water may be greater than the average. For greater precision there are a number of anthropomorphic equations for the calculation of total body water. The Watson formula for total body water was derived from and validated on several hundreds of patients. It uses height, weight, age and gender. [27] Watson formula for males; •

Total Body Water (TBW) litres = 2.447 - 0.09156 x age in years + 0.1074 x height in cm + 0.3362 x weight in kg.

Watson formula for females; •

Total Body Water (TBW) litres = -2.097 + 0.1069 x height + 0.2466 x weight.

Clinical use of total body water estimation and modified body weights The fluid-prescribing clinician has to have an estimate of her patient’s total body water in order to make rational decisions about the magnitude of solvent/ solute imbalances and the doses of fluid and/or electrolytes needed to correct them. Total body water estimates are also needed to decide the appropriate doses of hydrophilic drugs. For non-obese patients, total body water is proportionate to body weight, but with increasing obesity the utility of body weight as a scalar of total body water and cardiac output diminishes; the excess weight is predominantly fat rather than water and takes little of the cardiac

Biological Solutions

61

output. Anaesthetic muscle relaxant drugs and antibiotics are hydrophilic and the preferred scalar of dose should be total body water. In anaesthetic practice the recommended dose of hydrophilic drugs such as non-depolarising muscle relaxants is usually scaled to body weight for the non-obese, and to the ideal body weight for obese patients. Pierre Paul Broca (a French surgeon) created his formula (Broca’s index) in 1871. The Devine formula was proposed by in 1974 and remains the most widely-used today. ƒ ƒ

men: Ideal Body Weight (in kilograms) = 50 + 2.3 kg per inch over 5 feet. women: Ideal Body Weight (in kilograms) = 45.5 + 2.3 kg per inch over 5 feet.

Ideal body weight is the best adjusted weight for following total body water. It is reasonable to presume that the volume of distribution of urea is the total body water and thereby be able to titrate haemofiltration or haemodialysis prescription by ideal body weight to the desired rate of change of urea concentration or to clear other toxins. An estimated lean body mass formula (eLBM) for the normalisation of body fluid volumes was proposed in 1984, the Boer formula [28]. men: eLBM = 0.407W + 0.267H - 19.2 women: eLBM = 0.252W + 0.473H - 48.3 The Peters formula has been proposed for use in anaesthesia and critical care of boys and girls 14 years or younger. The formula first calculates the estimated extracellular fluid volume and then derives the eLBM [29]. • •

Estimated extracellular fluid volume (eECV) = 0.0215·W x 0.6469·H x 0.7236 eLBM = 3.8·eECV

Estimated lean body mass is the best scalar of cardiac output in obese patients and is therefore the preferred scalar of the induction dose of hypnotic agents such as di-isopropyl phenol or thiopentone sodium and for initial dosing of intravenous opiates such as fentanyl and remifentanil. Notice however that these agents are lipophilic and so the measured body weight is the appropriate scalar for setting the steady-

62

Chapter 3

state rate of maintenance infusion, even in obese patients [30]. Tidal volume recommendations for pulmonary ventilation are given in ml/kg predicted body weight (PBW), a parameter calculated from height and gender as height most closely predicts normal lung volumes in men and women [31]. In emergency situations when body weight has not been measured or recorded predicted body weight can rapidly be estimated from height and gender to guide fluid therapy and drug dosing.

Cell fluid and extracellular fluid The widely-taught compartmentalisation of a 70 kg man’s total body water (60% of his body weight = 42 litres) is extracellular 17 litres (about 40% of his TBW) and intracellular 25 litres (about 60% of his total body water). The water in adipose tissue is predominantly extracellular, so the proportion of total body water that is extracellular tends to be higher in women and in the obese. In morbid obesity 5060% of the total body water is extracellular. Skeletal muscle water is about 40-50% of total body water and accounts for almost 75% of intracellular fluid and round 33% of interstitial fluid. Patients with loss of muscle cell mass have a higher proportion of extracellular fluid and higher ratio of plasma to interstitial fluid volume.

Starling forces between extracellular and intracellular fluids Body water distribution between the extracellular and intracellular compartments reflects a steady-state of hydrostatic pressure and osmosis. At equilibrium the difference between the intracellular pressure and extracellular pressure is equal and opposite to the osmotic pressure difference across the cell membrane. The magnitude of diffusive water flux due to the osmotic pressure difference of an impermeable solute across an ideal membrane is proportional to the solute’s concentration difference and to the membrane’s hydraulic conductance (Lp.). In reality the cell membranes are less than ideal barriers, and most solutes are not fully impermeable. A fraction of the partially impermeable solute molecules will therefore be washed through the permeability barrier with solvent flux; this is convective

Biological Solutions

63

transport of solutes. In the 1950s Staverman proposed the reflection coefficient sigma (ɐ) to account for the observed osmotic pressure gradient relative to the ideal osmotic pressure gradient for an impermeable solute. A solute whose sigma approaches zero exerts almost no osmotic pressure (an ineffective osmol), and a solute whose sigma approaches 1 is almost fully effective. Albumin and urea are examples of important solutes whose ɐ for cell membranes approaches 1 in health, and Staverman’s reflection coefficient ɐ for a solute can be thought of as the fraction of molecules that are reflected by the membrane. When almost all the molecules are reflected ɐ approaches 1.0. When half of the molecules are reflected, ɐ is 0.5, and when only one in ten molecules is reflected ɐ is 0.1.

Maintenance of the extracellular-intracellular solute balance is energy-dependent As we observed above, cells need a near-continuous supply of adenosine tri-phosphate (ATP) to extrude permeable Na+ ions (via membrane channels) which are then balanced by an influx of permeable K+ ions; sodium and potassium therefore behave like impermeable effective osmoles sequestered in the ECF and ICF. Magnesium is an important co-factor. The sodium-potassium pump was discovered in 1957 by the Danish scientist Jens Christian Skou, a Nobel Prize winner in 1997. For every ATP molecule consumed, three sodium ions leave the cell and two potassium ions enter; there is thus a net export of a positive charge per cycle creating a membrane potential. Chloride (ClΫ) concentrates in the ECF, while fixed anions predominate in the ICF. The fixed intracellular anions include; • Metabolites such as ATP, phosphocreatine, and sulphate • Nucleotides • Proteins, which provide most of the intracellular anionic equivalence Along with potassium they create the Donnan effect osmotic gradient which would draw water into the cell were it not for the double Donnan effect of sodium potassium ATP-ase.

64

Chapter 3

Double Donnan effect •

The intracellular protein concentration (non-diffusible anion) is higher than extracellular, bringing about the first Gibbs-Donnan equilibrium. With unequal distribution of diffusible ions and electric charge, water tends to move into cells.

•

Active extrusion of sodium by Na-K pump makes sodium the major extracellular cation, and it has low membrane permeability. This brings about a second Gibbs-Donnan equilibrium that tends to move water out of cells.

•

At steady-state the two effects balance out and cell volume remains stable, but if sodium potassium ATP-ase is inhibited cells will swell and rupture due to the first Gibbs-Donnan equilibrium. Water is therefore passively distributed between intracellular and extracellular compartments in proportion to the effective Na+ and K+ content to reach effective osmotic equilibrium (tonicity) and establish cell volume.

Cell volume regulation and intracranial pressure The principle that cell volume is closely linked to plasma tonicity is particularly important in the nervous system; as plasma tonicity falls, cells swell. An acute onset (usually in 48 hours’ duration) hyponatremia also makes the brain vulnerable to injury (osmotic demyelination) if the electrolyte disturbance is corrected too rapidly. The reuptake of organic osmolytes after correction of hyponatremia is slower than the loss of organic osmolytes during the adaptation to hyponatremia. Areas of the brain that remain most depleted of organic osmolytes are the most severely injured by rapid correction. The brain’s reuptake of myoinositol, one of the most abundant osmolytes, occurs much more rapidly in a uremic environment, and patients with uraemia are less susceptible to osmotic demyelination. Cerebral demyelination is a rare complication of overly rapid correction of hyponatremia. The principal risk factors for cerebral demyelination are correction of the serum sodium more than 25 mEq l-1 in the first 48 hours of therapy, correction past the point of 140 mEq l-1, chronic liver disease, and prior hypoxic/anoxic episode. [32]

Cell volume regulation beyond the brain Recently-discovered volume regulated anion channels (VRAC) are not unique to the central nervous system and may prove to have a pivotal role in cell volume regulation in all cell types. Research into the therapeutic potential of hypertonic saline led to the observation that variations in cell volume have quite profound effects on cellular metabolism and gene expression and could, for example, protect against lung injury in a haemorrhagic shock model [35]. A recent meta-

66

Chapter 3

analysis of human studies confirms the expectation of lower-volume resuscitation from sepsis with hypertonic saline, but there was no signal of outcome advantage [36]. Hypertonic sodium is also effective as sodium lactate [37]. VRAC activity could explain the finding from clinical experience that neither electrolyte-free water nor potassium solution infusions increase intracellular fluid volume [38]. Total body water expansion by intravenous fluid infusions of any tonicity appears to be limited to the extracellular fluid volume, which runs contrary to the conventional model exemplified by Twigley & Hillman [39]. Hessels and her colleagues at Groningen have therefore proposed an ‘alternative model’ of water, sodium and potassium distribution.

Hessels’ Alternative Model of water, sodium and potassium distribution Conventional Textbooks and Review Articles presume that an infusion of electrolyte free water will proportionately increase the volume of all compartments of the total body water and reduce osmolarity. In a cohort study of post-surgical patients treated in an Intensive Care Unit with conventional intravenous fluids for 4 days, Hessels and colleagues found that there was a strongly positive accumulation of sodium and total fluid, but a negative balance of electrolyte-free water and potassium. In a sub-study comparing the effects of prescribing potassium to a target of [4.0 mmol l-1] or [4.5 mmol l-1] they found that all the excess potassium of the second group was renally excreted. They reasonably interpreted these observations as showing that excess fluid in clinical practice results in interstitial expansion (extracellular oedema) while the intracellular volume, where potassium is the dominant osmolar cation, is regulated close to its healthy normal. They speculate that the cytosol is able to clear alternative osmolytes when there is volume increase by electrolyte free water infusion, and generate alternative osmolytes when hypertonic saline infusion reduces cell volume. Intracellular volume is thereby conserved in the face of changing body water tonicity [38]. With the usual caveat that “more research needs to be done”, this is an interesting hypothesis. In what appears to be a reanalysis of the same patient cohort, Hessels considers the possibility that excess sodium is stored non-osmotically in the skin, as demonstrated in terrestrial space station residents! [40] The data confirmed “Missing extracellular sodium” ions, and

Biological Solutions

67

intriguingly also revealed “missing extracellular chloride” too. Studies to identify whether the missing ions are held non-osmotically or shifted to the intracellular fluid are warranted.

Dehydration “This patient is dry.” As a clinical diagnosis, dehydration implies deficiency in total body water. There are currently no consensus statements on its definition or assessment, but it is both prevalent and costly within healthcare, particularly amongst elderly and chronically debilitated patients. It’s presence in any disease cohort is often associated with increased morbidity and mortality. It is regrettable that dehydration and volume depletion (or even hypovolaemia) are often conflated in common parlance. Greater discipline would help to promote more rational therapy. The diagnosis of hypertonic dehydration should be made when there is evidence of intracellular water deficit with hypertonicity and a disturbance in water metabolism. The diagnosis is established by laboratory analysis of plasma [Na +] or calculation of serum osmolarity (tonicity). Ideally, plasma osmolality would be measured. Treatment is by encouraging increased enteral intake of water or, if that is not possible, the parenteral infusion of a hypotonic (including electrolyte-free) solution. Volume depletion describes the net loss of total body sodium and a reduction in intravascular volume (extracellular fluid volume depletion) so may be distinguished as isotonic dehydration. The term hypovolaemia should be used to specify circulating volume depletion, which may be due to severe hypertonic or (more frequently) isotonic dehydration. Volume depletion and hypovolaemia from any cause are initially treated by infusion of isotonic salt solutions.

Hyponatraemia In a recent pragmatic trial of isotonic salt solutions in critically-ill patients, hyponatraemia (< 135 mmol l-1) developed during treatment in 35-40% of patients [6]. Presuming that critically-ill patients will usually have raised circulating arginine vasopressin levels (as argued above) this effect is probably due to water reabsorption from the renal collecting ducts via aquaporin-2 channels. Water restriction would be the preferred therapy, but in critically-ill patients may be impractical.

Chapter 3

68

There is a report of use of urea in such cases [41]. I have been critical of UK recommendations for hypotonic intravenous fluid therapy in hospitalised patients [42]. I prefer the advice of Texan nephrologists Moritz and Ayus who advocate isotonic solutions for post-surgical adult and paediatric patients [43–45]. Syndrome of inappropriate antidiuresis. The diagnosis of syndrome of inappropriate antidiuresis (SIAD) is reserved for chronically-hyponatraemic patients with no recognised cause of arginine vasopressinaemia [46]. Some tumours secrete arginine vasopressin (ectopic production), but in most cases the neurohypophyseal response to changes in plasma osmolarity is defective. Patients can be tested by their response to infusion of hypertonic saline. • • • •

In type A SIAD arginine vasopressin levels are high, fluctuating and unrelated to increases in plasma sodium; In type B SIAD there is a slow constant increase in arginine vasopressin unaffected by increases in plasma sodium; Type C SIAD is also referred to as ‘reset osmostat’; there is a rapid progressive rise in plasma arginine vasopressin as plasma sodium rises toward the normal range. That leaves 5% to 10% of SIAD patients for whom there is no demonstrable abnormality in the osmoregulation of arginine vasopressin (type D). We have to look elsewhere for a cause of inappropriate antidiuresis. In some children it appears to be due to an activating mutation of the Vasopressin-2 receptor. In other patients, it may be due to abnormal control of aquaporin-2 water channels in renal collecting tubules or production of an antidiuretic principle other than arginine vasopressin.

The extracellular disposition of solvent and solutes The traditional subdivision of extracellular fluid is • • • •

plasma, interstitial fluid and lymph, dense connective tissue and bone, transcellular fluids (including pleural, peritoneal, cerebrospinal, exocrine secretions),

Biological Solutions

69

• adipose tissue. Fluid therapy predominantly impacts the extracellular fluid volume for reasons discussed above, so the details of extracellular fluid disposition are important and explored in greater detail in the following chapters. Nonetheless, some preliminary comments on pathological body water accumulation are apposite at this point.

Oedema Generalised oedema is often seen as a disorder of whole-body fluid volume regulation. The story goes that with reduced effective arterial blood volume due to a decrease in cardiac output or to vasodilation, the normal central inhibition of the sympathetic nervous system activity and baroreceptor-mediated, non-osmotic arginine vasopressin release is attenuated. The resultant increase in renal adrenergic activity stimulates the renin-angiotensin-aldosterone pathway which reverses vasodilation and reduces sodium delivery to the sites of aldosterone, arginine vasopressin, and natriuretic peptide action. This diminished distal sodium and water delivery is an important factor in the failure to escape from the sodium-retaining effects of aldosterone, the resistance to the natriuretic and diuretic effects of natriuretic peptides, and the diminished maximal solute-free water excretion in patients with oedema. Such an account is necessary if it is presumed that interstitial fluid and plasma are in a filtration - absorption equilibrium determined by Starling forces across the microvascular permeability barrier. Some physiologists and clinicians still take it as proven that capillaries provide a symmetric barrier such that plasma solvent is filtered to the interstitial fluid at the higher-pressure part of the capillary, then absorbed under a dominant osmotic pressure gradient at the lower pressure part. In the twenty-first century hypotheses originally proposed by Charles Michel (London) and Sheldon Weinbaum (New York) were experimentally confirmed, and we now know that capillaries are in fact asymmetric filters. The Starling forces drive filtration of solvent from plasma to interstitium, but when the balance of forces moves towards reabsorption, the transendothelial osmotic pressure gradient rapidly declines. Absorption of filtered fluid by capillaries at steady-state does not occur. The transcapillary colloid

70

Chapter 3

osmotic pressure difference opposes transendothelial filtration but does not reverse it. Only when the steady-state is disturbed by a sudden drop in transcapillary hydrostatic pressure difference is there a volume-limited and transient osmotic “autotransfusion” of interstitial fluid to plasma. The twentieth century view of lymph flow was that it is a minor pathway for return of interstitial fluid to the veins, being the balance of capillary filtration and capillary reabsorption. The current view is that capillary filtration rate is much lower than previously believed and that lymph returns interstitial fluid to veins. The central nervous system was, until recently, believed to have no lymphatic drainage, but a brain glial cell and lymph “glymphatic” system has been described. The lymphatic vessels and channels are now to be seen as an integral part of an extracellular fluid circulation. Water normally enters the body by crossing gastro-intestinal epithelium. From the sub-mucosal interstitium water is either taken into collecting lymphatics or absorbed by diaphragm-fenestrated capillaries to venules. Water can also be absorbed from lymph at lymph nodes, concentrating proteins and lipids in the efferent lymph which is finally discharged into a great vein. With this appreciation generalised oedema can be seen to result from a mismatch between the transendothelial solvent filtration rate Jv and the afferent lymph flow Qlymph, with some water being absorbed to the blood stream at lymph nodes while the remainder returns to central veins as efferent lymph.

A quantitative approach to fluid and electrolyte balance at the bedside On a modest diet around 150 mosmol of electrolytes (mostly sodium, potassium and chloride) must be excreted by the kidneys in addition to the 500 mosmol of urea every day, total 650 mosmol. In humans, the maximum urine concentration is 1200 mosmol kg-1, so at least 500 ml of urine per day (21 ml h-1) are needed to eliminate the solute load. Glucose is normally fully metabolised with little or no urinary excretion, except of course in diabetes mellitus where the additional osmolar load leads to polyuria. An active person takes in about 2.5 litres water per day and produces about 1.8 litres of urine with about 750 mosmol (400 mosmol kg-1), but under restrictive fluid input can easily excrete the 750 mosmol in about 900 ml concentrated urine in 24

Biological Solutions

71

hours. Stress and systemic inflammation increase proteolysis and so increase the nitrogen load to be excreted (around 800 mosmol of urea has been suggested), but the stress hormonal environment inhibits increase in urine output. Maximal concentrating ability is reduced from about 1200 to perhaps only 600 mosmol kg-1. Expect therefore to see a transient rise in plasma urea and urine output reduced to about 1000 ml per day after surgery or in acute illness. Giving liberal fluids during stress has only a small effect on urine output but may reduce plasma urea by dilution. Unless the patient was depleted, most of the extra water and electrolytes are retained as oedema (“fluid overload”). The patient gains weight (1 litre = 1 kg). Weight gain of more than 4 kg (fluid balance > 4 litres) increases the rate of complications including ileus, infection, and arrhythmias. Transient dilutional hyponatraemia, hypoalbuminaemia and hypomagnesaemia (and low urea) are commonly noticed after major surgery and acute illness for which liberal intravenous fluid therapy has been used. It is preferable to prevent iatrogenic water intoxication than to correct by giving sodium or albumin or magnesium. Evaporative loss as sweat may be as little as 10-20 ml h-1 in a temperate climate, but much greater in summer heat or in pyrexial illness or after exercise. Water (and carbon dioxide) produced by metabolism matches water lost by exhalation. (The partial pressures of CO2 and water in exhaled air are similar). In chronic kidney disease with advanced renal damage the urine concentration ability is lost and an isosthenuric urine is produced (i.e. close to plasma osmolality and in the range 250-300 mosmol kg-1). If the obligatory urine output is obtained by dividing the daily osmolar excretion by the maximum urine osmolality, at least 2 l of diuresis would be required to eliminate a normal solute load. This is achieved with a liquid intake of between 2.5 and 3.5 l per day, depending on extrarenal fluid losses. Such patients are at very high risk of renal failure when subjected to stress and systemic inflammation. Some mammals are capable of much higher urinary osmolar concentration than humans. Rats can attain approximately 3,000 mosmol kg-1, hamsters and mice 4,000 mosmol kg-1, and the waterphobic chinchillas 7,600 mosmol kg-1.

72

Chapter 3

References 1. Stewart PA. Independent and dependent variables of acid-base control. Respir Physiol. 1978;33:9-26. 2. Stewart PA. HOW TO UNDERSTAND ACID-BASE: A Quantitative Acid - Base Primer for Biology and Medicine. Elsevier; 1981 3. Stewart PA. Modern quantitative acid-base chemistry. Can J Physiol Pharmacol. 1983;61:1444-1461. 4. Nielsen OM, Engell HC. Changes in extracellular sodium content after elective abdominal vascular surgery. Acta Chir Scand. 1986;152:587-591. 5. Self WH, Semler MW, Wanderer JP et al. Balanced Crystalloids versus Saline in Noncritically Ill Adults. N Engl J Med. 2018;378:819828. 6. Semler MW, Self WH, Wanderer JP et al. Balanced Crystalloids versus Saline in Critically Ill Adults. N Engl J Med. 2018;378:829839. 7. Pfortmueller CA, Uehlinger D, von Haehling S, Schefold JC. Serum chloride levels in critical illness-the hidden story. Intensive Care Med Exp. 2018;6:10. 8. Yu H, Wang M, Wang D et al. Chlorinated Lipids Elicit Inflammatory Responses In Vitro and In Vivo. Shock. 2018 9. Welling PA, Ho K. A comprehensive guide to the ROMK potassium channel: form and function in health and disease. Am J Physiol Renal Physiol. 2009;297:F849-63. 10. Di Rollo N, Caesar D, Ferenbach DA, Dunn MJ. Survival from profound metabolic acidosis due to hypovolaemic shock. A world record. BMJ Case Rep. 2013;2013 11. Siggaard-Andersen O, Gøthgen IH. Oxygen and acid-base parameters of arterial and mixed venous blood, relevant versus redundant. Acta Anaesthesiol Scand Suppl. 1995;107:21-27. 12. Fencl V, Jabor A, Kazda A, Figge J. Diagnosis of metabolic acid-base disturbances in critically ill patients. Am J Respir Crit Care Med. 2000;162:2246-2251. 13. Gilfix BM, Bique M, Magder S. A physical chemical approach to the analysis of acid-base balance in the clinical setting. J Crit Care. 1993;8:187-197. 14. Balasubramanyan N, Havens PL, Hoffman GM. Unmeasured anions identified by the Fencl-Stewart method predict mortality better than base excess, anion gap, and lactate in patients in the pediatric intensive care unit. Crit Care Med. 1999;27:1577-1581.

Biological Solutions

73

15. Liskaser FJ, Bellomo R, Hayhoe M et al. Role of pump prime in the etiology and pathogenesis of cardiopulmonary bypass-associated acidosis. Anesthesiology. 2000;93:1170-1173. 16. Cole L, Bellomo R, Baldwin I, Hayhoe M, Ronco C. The impact of lactate-buffered high-volume hemofiltration on acid-base balance. Intensive Care Med. 2003;29:1113-1120. 17. Ortner CM, Combrinck B, Allie S et al. Strong ion and weak acid analysis in severe preeclampsia: potential clinical significance. Br J Anaesth. 2015;115:275-284. 18. Nielsen S, Frøkiaer J, Marples D, Kwon TH, Agre P, Knepper MA. Aquaporins in the kidney: from molecules to medicine. Physiol Rev. 2002;82:205-244. 19. Clark E, Molnar AO, Joannes-Boyau O, Honoré PM, Sikora L, Bagshaw SM. High-volume hemofiltration for septic acute kidney injury: a systematic review and meta-analysis. Crit Care. 2014;18:R7. 20. Welling PA. Roles and Regulation of Renal K Channels. Annu Rev Physiol. 2016;78:415-435. 21. Hadchouel J, Ellison DH, Gamba G. Regulation of Renal Electrolyte Transport by WNK and SPAK-OSR1 Kinases. Annu Rev Physiol. 2016;78:367-389. 22. Frindt G, Yang L, Uchida S, Weinstein AM, Palmer LG. Responses of distal nephron Na+ transporters to acute volume depletion and hyperkalemia. Am J Physiol Renal Physiol. 2017;313:F62-F73. 23. Landry DW, Levin HR, Gallant EM et al. Vasopressin deficiency contributes to the vasodilation of septic shock. Circulation. 1997;95:1122-1125. 24. Maybauer MO, Maybauer DM, Enkhbaatar P et al. The selective vasopressin type 1a receptor agonist selepressin (FE 202158) blocks vascular leak in ovine severe sepsis*. Crit Care Med. 2014;42:e525-e533. 25. Morales D, Madigan J, Cullinane S et al. Reversal by vasopressin of intractable hypotension in the late phase of hemorrhagic shock. Circulation. 1999;100:226-229. 26. Mayer B, Németh K, Krepuska M et al. Vasopressin stimulates the proliferation and differentiation of red blood cell precursors and improves recovery from anemia. Sci Transl Med. 2017;9 27. Watson PE, Watson ID, Batt RD. Total body water volumes for adult males and females estimated from simple anthropometric measurements. Am J Clin Nutr. 1980;33:27-39. 28. Boer P. Estimated lean body mass as an index for normalization of body fluid volumes in humans. Am J Physiol. 1984;247:F632-6.

74

Chapter 3

29. Peters AM, Snelling HL, Glass DM, Bird NJ. Estimation of lean body mass in children. Br J Anaesth. 2011;106:719-723. 30. Ingrande J, Lemmens HJ. Dose adjustment of anaesthetics in the morbidly obese. Br J Anaesth. 2010;105 Suppl 1:i16-23. 31. Brower RG, Lanken PN, MacIntyre N et al. Higher versus lower positive end-expiratory pressures in patients with the acute respiratory distress syndrome. N Engl J Med. 2004;351:327-336. 32. Achinger SG, Ayus JC. Treatment of Hyponatremic Encephalopathy in the Critically Ill. Crit Care Med. 2017;45:1762-1771. 33. Liotta EM, Romanova AL, Lizza BD et al. Osmotic Shifts, Cerebral Edema, and Neurologic Deterioration in Severe Hepatic Encephalopathy. Crit Care Med. 2018;46:280-289. 34. Mongin AA. Volume-regulated anion channel--a frenemy within the brain. Pflugers Arch. 2016;468:421-441. 35. Rizoli SB, Kapus A, Fan J, Li YH, Marshall JC, Rotstein OD. Immunomodulatory effects of hypertonic resuscitation on the development of lung inflammation following hemorrhagic shock. J Immunol. 1998;161:6288-6296. 36. Orbegozo D, Vincent JL, Creteur J, Su F. Hypertonic Saline in Human Sepsis: A Systematic Review of Randomized Controlled Trials. Anesth Analg. 2019 37. Carteron L, Solari D, Patet C et al. Hypertonic Lactate to Improve Cerebral Perfusion and Glucose Availability After Acute Brain Injury. Crit Care Med. 2018;46:1649-1655. 38. Hessels L, Oude Lansink A, Renes MH et al. Postoperative fluid retention after heart surgery is accompanied by a strongly positive sodium balance and a negative potassium balance. Physiol Rep. 2016;4 39. Twigley AJ, Hillman KM. The end of the crystalloid era? A new approach to peri-operative fluid administration. Anaesthesia. 1985;40:860-871. 40. Hessels L, Oude Lansink-Hartgring A, Zeillemaker-Hoekstra M, Nijsten MW. Estimation of sodium and chloride storage in critically ill patients: a balance study. Ann Intensive Care. 2018;8:97. 41. Decaux G, Andres C, Gankam Kengne F, Soupart A. Treatment of euvolemic hyponatremia in the intensive care unit by urea. Crit Care. 2010;14:R184. 42. GIFTAHo; an improvement on GIFTASuP? New NICE guidelines on intravenous fluids. [editorial]. Anaesthesia 2014;69(5):410. 43. Moritz ML, Ayus JC. Prevention of hospital-acquired hyponatremia: a case for using isotonic saline. Pediatrics. 2003;111:227-230.

Biological Solutions

75

44. Water water everywhere: standardizing postoperative fluid therapy with 0.9% normal saline. [editorial]. Anesth Analg 2010;110(2):293. 45. Ayus JC, Caputo D, Bazerque F, Heguilen R, Gonzalez CD, Moritz ML. Treatment of hyponatremic encephalopathy with a 3% sodium chloride protocol: a case series. Am J Kidney Dis. 2015;65:435-442. 46. Spasovski G, Vanholder R, Allolio B et al. Clinical practice guideline on diagnosis and treatment of hyponatraemia. Intensive Care Med. 2014;40:320-331.

CHAPTER 4 ENDOTHELIUM

Chapter summary The structures of the endothelial filtration barrier, from endothelial surface layer and glycocalyx to basal lamina, are described in more detail. They are responsible for the rate of filtration of plasma solvent and solutes to the interstitial space at the beginning of the vital interstitial fluid/ lymph circulation which is at the heart of understanding fluid physiology. The heterogeneity of microvascular endothelial cells separating plasma from interstitial fluid is often not addressed in discussions of fluid physiology. The discontinuous capillaries of the sinusoidal tissues provide a low-resistance two-way communication between the interstitial spaces of the liver, bone marrow and spleen, and the plasma of the blood that perfuses them. This creates a small reservoir of solvent volume and proteins (including albumin) which can support plasma volume during acute hypovolaemia, or provide a safety valve against hypervolaemia during fluid therapy. In the hyperdynamic circulatory state associated with critical illness, as much as 50% of the left ventricular output perfuses sinusoidal tissues which do not retain albumin within the plasma volume, creating an overall impression of leaky systemic capillaries. Endothelium is the first regulator of the balance between intravascular and extravascular fluid volumes. Endothelium preserves a healthy circulating blood coagulation profile, works with leucocytes to deliver an optimal immune state, continuously regenerates and adapts to tissue demands and is crucial to the healthy function of every traditional organ system structure. Interesting hypotheses of endotheliopathy as a unifying characteristic of critical illness pathophysiology are emerging, along with a hope that therapeutic glycocalyx protection or repair may be feasible in future.

Endothelium

77

Endothelium Endothelial cells are specialised epithelial cells which line blood vessels. The endothelial and epithelial cells that make up the tissues in humans differ in location, structure, and function. In simple terms the endothelial cells are those that lie 'inside' the body, while epithelial cells are those lying 'outside' the body lining the skin, the gastrointestinal tract and genito-urinary tract. The single layer of microvascular endothelial cells (MVECs) that forms capillaries and venules is of particular importance to fluid physiology. It has been argued that medical subspecialisation according to traditional body organ classifications has left endotheliology, the study of diseases of endothelial dysfunction, in a conceptual blind spot. There is a case that the endothelium should be considered an organ system of the body [1]. Seen under a light microscope, capillaries average from 9 to 12 μm in diameter, just large enough to permit passage of erythrocytes whose diameter is around 7 μm. Capillaries are formed by adjoining endothelial cells which are typically just large enough to fully encircle the vessel lumen. Two or three endothelial cells are typically seen in a cross section of a capillary, overlapping at their edges and connected to one another by intercellular tight junction strands and adherens junctions. The tight junction strands have occasional tight junction gaps that permit paracellular fluid movement between plasma and interstitial fluid. No vascular smooth muscle surrounds a capillary. Venules are similar to capillaries in that they have a very thin wall and participate in solvent and solute movement; but unlike capillaries they have diameters that range from approximately 20 - 60 μm, and so it is here that white blood cells are best able to migrate from blood stream to interstitium. Some venules also feature a tunica media with one or two layers of muscle fibres and a histologically-evident tunica adventitia. Capillaries and venules are the microvessels through which solvent and solute transfer between the intravascular space and the interstitium. The microvascular endothelial cell layer has a basal lamina around 80 nm thick which contains type 4 collagen, laminin, fibronectin, and some proteoglycans. Finally, there is a subendothelial layer connecting to the interstitial collagen bundles.

Chapter 4

78

Too small to be seen in light Electron microscopy was invented in 1931 by Knoll and Ruska. The first transmission electron microscope was available to researchers in 1939. The wavelength of an electron can be up to 100,000 times shorter than that of a photon, so electron microscopes can reveal the structure of objects invisible by light. In 1959 Bennett, Luft and Hampton proposed a morphological classification of capillaries based on such EM studies as they had been able to achieve. They proposed: Classification based on basement membrane A - Complete continuous investment of the capillary by basement membrane B - Incomplete discontinuous investment, as seen in liver, spleen and bone marrow Classification based on endothelial cell type 1 - no fenestrations or perforations 2 - intracellular (transcellular) fenestrations or perforations 3 - intercellular (paracellular) fenestrations or gaps Classification based on presence or absence of pericapillary cellular investment Alpha - without a pericapillary cellular investment Beta - with a pericapillary cellular investment, as seen in the central nervous system Bennett Luft and Hampton noted that John Pappenheimer’s work on transcapillary solute movements had been performed on capillaries of type A-1-alpha and predicted the presence of pores of at least 0.3 nm diameter and 0.3 μm length (the thickness of the endothelial cell). Taking into account the population density of Pappenheimer’s pores and the thickness of tissue slices, they expected to see two to six Pappenheimer pores in each section of such capillaries. With great understatement they comment “We deem it significant that we have not visualised such passages …, although we have studied many hundreds of pictures … from this laboratory and others. For these reasons we doubt the pore model of capillary structure and function as proposed by Pappenheimer.” [2]

Endothelium

79

Discovery of the glycocalyx Little attention was paid to the thin luminal lining of the endothelial cells until 1966, when J.H. Luft published his work on “Fine structures of capillary and endocapillary layer as revealed by ruthenium red.” Luft’s endocapillary layer is now known as the endothelial glycocalyx. In 1963, Bennett had suggested glycocalyx as the general term for an “extracellular sugary coating, wherever it may be found.” Curry and Michel proposed a theory “that the molecular sieving properties of the capillary wall reside in a matrix of molecular fibres which covers the endothelial cells and fills the channels through or between them” in 1980. Levick and Michel later proposed that the small pore system of the transvascular semi-permeable membrane is the endothelial glycocalyx layer where it covers the endothelial intercellular clefts, separating plasma from a “protected region” of the subglycocalyx space which is almost protein-free during filtration. The fact that low protein concentration within the subglycocalyx intercellular spaces accounts for the low Jv and lymph flow in most tissues is a critical insight and the basis of the glycocalyx model, a key part of the revision of the Starling principle [3]. The idea that the endothelial surface layer associated with the glycocalyx is just a passive luminal gel that comes between circulating blood cells and the endothelial cell membrane is now giving way to the realization that it is a multilayer, multicomponent, biochemical structure that functions exquisitely as a molecular sieve, a lubrication layer for red blood cell motion, an inhibitor of inflammation and a sensor of plasma flow-induced fluid shear strain. Glycocalyxfocused therapies have potential in vascular diseases and cancer [4].

Structure of the endothelial glycocalyx The endothelial glycocalyx is a web of membrane-bound glycoproteins and proteoglycans on the luminal side of the endothelial cells, associated with various glycosaminoglycans (GAGs) (mucopolysaccharides) which contribute to the volume of the endothelial surface layer. After Charles Michel had hypothesised the existence of such a fibre matrix, researchers at Imperial College London established the existence of a quasi-periodic substructure from electron microscopy images [5]. Weinbaum et al. then presented a quasiperiodic hexagonal arrangement for a glycocalyx with a spacing of 100 nm between proteoglycan clusters [6]. Figure 4.1 shows Kenton

80

Chapter 4

Arkill’s electron microscopic images of continuous glycocalyx layers in various mammalian microvessels (A, B and C). All have lateral spacings at around 20 nm (D) and longer spacings above 100 nm (E). The structural organisation accounts satisfactorily for observed molecular filtering characteristics [7}.

Figure 4.1

A porous outer zone of the endothelial surface layer is now described [8]. It is proposed that a non-stirred layer of albumin molecules accumulates at the boundary of the porous outer zone and the inner endothelial glycocalyx layer [9]. This boundary within the endothelial surface layer may also be involved in sodium regulation [10]. Hyaluronan, which can bind sodium ions, is present in both free and CD44 membrane bound forms. Figure 4.2 is a cartoon illustrating the

Endothelium

81

arrangement.

Figure 4.2

The next cartoon, Figure 4.3, gets in close to the lipid bilayer and illustrates the arrangement of glycocalyx proteoglycans, glycoproteins and GAGs on the surface of endothelial cells. The lower panel is centred on cholesterol and sphingolipid-rich regions where caveolation occurs.

82

Chapter 4

Figure 4.3

Caveolae (caveolin-1) Cave-like structures called caveolae are associated with lipid bilayer regions high in cholesterol and sphingolipids within the membrane. Caveolin-1 is an integral protein of the caveolae and is involved in transmembrane trafficking of lipids and proteins (transcytosis). Albondin (glycoprotein 60) is a 60 kDa albumin receptor molecule, selectively expressed on the plasma membrane of continuous endothelium (except for the brain), and interacts with caveolin-1. Albondin specifically binds native albumin rather than altered albumins. Caveolae-mediated albumin transcytosis is enhanced by atrial natriuretic peptide via the endothelial GC-A receptor. Increased albumin transcytosis is a feature of sepsis and may be a contributor to increased capillary permeability to albumin. Transcellular permeability increases may precede and subsequently trigger paracellular permeability via Src-mediated phosphorylation of caveolin-1. Normally, about half of the transendothelial transport of native albumin is transcellular, via caveolin-1 and albondin, with the remainder being paracellular convection via junction breaks in the

Endothelium

83

intercellular junctions. A series of molecules involved in endothelial nitric oxide synthase signalling localize in caveolae. Glypicans Glypicans and their heparan sulphate GAG chains localize in the regions where caveolae are seen. Syndecans Transmembrane syndecans are shown to cluster in the outer edge of caveolae. Besides heparan sulphate, syndecans also contain chondroitin sulphate and other oligosaccharides close to the point at which the core protein emerges from the lipid bilayer. The anionic oligosaccharides contribute to the net negative charge of glycocalyx. Most of the oligosaccharides are covalently attached to membrane proteins, but some are solute molecules adsorbed in the glycocalyx region by noncovalent interactions. The cytoplasmic domains of syndecans have linker molecules connecting them to cytoskeletal structures. Oligomerization of syndecans helps them make direct associations with intracellular signalling effectors. Matrix metalloproteases may shed syndecans. Plasma level of circulating syndecan is widely presumed to be a marker of glycocalyx shedding [11] and used to define various syndromes of endotheliopathy [12]. Elevated syndecan levels have been found in patients with thermal injuries [13], cardiac surgery [14] and acute traumatic brain injury [15]. Sialoglycoproteins The monosaccharide sialic acid is highly concentrated on endothelial cells as sialoglycoprotein. A glycoprotein with short oligosaccharide branched chains and associated sialic acid caps is shown. Sialic acid residues contribute to microvascular impermeability to both water and albumin [16]. CD44 antigen and hyaluronan Hyaluronan (Hyaluronic Acid, HA) is the most abundant non-sulphated glycosaminoglycan and differs from the other glycosaminoglycans in that it is synthesised by the cell membrane, rather than being secreted through the Golgi apparatus. At around 107 Da, it is much larger than

84

Chapter 4

other glycosaminoglycans. Hyaluronan is very long and weaves into the glycocalyx and binds with CD44 antigen, a cell-surface glycoprotein involved in cell–cell interactions. Transmembrane CD44 can also have the sulphated glycosaminoglycans chondroitin and heparan as well as oligosaccharides attached to it. It localizes amongst caveolae. Another characteristic of hyaluronan is its high anion charge, which attracts a large solvation volume; this makes hyaluronic acid an important determinant of tissue hydration. Hyaluronan gives the gel phase of the glycocalyx some remarkable rheological properties. Solutions of hyaluronic acid are viscoelastic, and their viscosity changes with shear stress. At low shear stress, viscosity is around 100 times higher than the viscosity of the solvent, while under high shear stress, viscosity falls substantially. The sulphated Glycosaminoglycans By removing GAGs and measuring volume reduction (compaction or dehydration) of the endothelial surface layer, the sulphated GAGs heparan sulphate and chondroitin sulphate appear to be major volume contributors, along with hyaluronic acid. Compaction of the endothelial surface layer by removal of GAGs largely preserves its glycocalyx resistance to filtration, despite loss of thickness and possible reduction in permeability. Compaction of the endothelial surface layer and the associated increase in heparan sulphate, hyaluronic acid, or chondroitin sulphate in plasma are widely considered to be markers of glycocalyx injury, variously described as shedding, flaking, degradation or fragmentation. Rapid crystalloid infusion in volunteers or laboratory models, for instance, results in elevated plasma levels of hyaluronic acid. Glycosaminoglycans are present in the extracellular matrix of pulmonary and systemic tissues and are concentrated in lymph, so elevated lymph flow is an entirely plausible explanation for their appearance in plasma. Increased plasma concentrations of GAGs have been found in septic shock patients, and they appear to reduce the antibacterial properties of plasma. Fragments of heparan sulphate have been associated with cognitive dysfunction of sepsis [17]. A cohort of severely injured trauma patients had endogenous heparinisation as evidenced by whole blood thromboelastography. Endogenous heparinisation is presumed to be due to the effect of glycosaminoglycans, including heparan sulphate, on haemostasis [18].

Endothelium

85

The presumption that elevated GAG concentrations in plasma are caused only by liberation from the endothelial surface layer is too simplistic [19]. Heparanase, released from activated mast cells, cleaves heparan sulphates from core proteins. Hyaluronidase and the serine proteases thrombin, elastase, proteinase 3 and plasminogen, as well as cathepsin B lead to loss of hyaluronan from the endothelial surface layer. Plasma proteins, along with sodium and cationic amino acids, are known to associate with the various GAGs. Glycosaminoglycan therapy Sulodexide is a mixture of two highly purified glycosaminoglycans, fastmoving heparin 80% and dermatan sulphate 20%. It is indicated for the treatment of various vascular disorders including chronic venous disease, peripheral arterial disease and diabetic nephropathy, though its therapeutic effect is not fully elucidated. It has anticoagulant, profibrinolytic and anti-aggregative properties and delivers substrates for the reconstruction of vascular endothelial glycocalyx. It also inhibits the activity of enzymes responsible for cleaving GAGs. Sulodexide penetrates endothelial cells and has an effect on the expression of multiple growth factors [20, 21]. Sphingosine-1-phosphate Sphingolipids regulate several aspects of cell behaviour, and cells adjust their sphingolipid metabolism in response to metabolic needs. The final product of sphingolipid metabolism is sphingosine-1phosphate (S1P), a potent bioactive lipid. S1P is involved in the regulation of cell proliferation, cell migration, actin cytoskeletal reorganization and cell adhesion. With a plasma half-life of only 15 minutes, a nearconstant supply of S1P is required to maintain the circulating concentration. S1P has been shown to inhibit protease activity that causes shedding of endothelial glycocalyx constituents such as the sulphated glycosaminoglycans chondroitin and heparan, and syndecan [22]. S1P synthesis by erythrocytes and S1P transport by high density lipoprotein and albumin therefore emerge as a vital factors in nourishing and protecting the endothelial glycocalyx and maintaining vascular homeostasis [23, 24]. S1P acts via S1P1 receptors, and S1P1 receptor antagonism abolishes the protection. The mechanism of protection from loss of glycocalyx components by S1P-dependent pathways has been shown to be suppression of metalloproteinase (MMP) activity. General inhibition of MMPs protected against loss of

Chapter 4

86

chondroitin sulphate and syndecan-1. Specific inhibition of MMP-9 and MMP-13 protected against chondroitin sulphate loss. Other contributors to S1P protection against shock-induced hyperpermeability are the expression of adherens junctional proteins, and protection against mitochondrial membrane depolarization, acting through mitochondrial complex III [25]. Plasma levels of S1P have been measured in patients with dengue fever and sepsis and found to be lower than in healthy subjects [26, 27]. In a clinical study glypican plasma concentrations were significantly positively correlated with plasma levels of syndecan 1 and negatively correlated with plasma levels of the glycocalyx-protective factors apolipoprotein M and sphingosine-1-phosphate [28]. S1P has been shown to attenuate lung ischaemia-reperfusion injury, and the specific S1P receptor involved has been identified [29]. In the future, treatment with S1P receptor agonists is therefore feasible. Other glycocalyx constituents A number of other molecules, derived both from the endothelium and from the plasma and involved in coagulation and inflammation, exist within the glycocalyx layer. Albumin’s role in glycocalyx function is often talked about but is limited. The presence of hyaluronan creates an exclusion zone for albumin in the intravascular gel phase, just as it does within the extravascular gel phase of the interstitium. The hydraulic conductivity of the glycocalyx only increases significantly when plasma albumin is very low, and this may be related to S1P transport. There is as yet no evidence that albumin therapy improves glycocalyx function in clinical situations.

The glycocalyx and the endothelial surface layer In many communications the terms endothelial glycocalyx and endothelial surface layer have been used interchangeably. The quasiperiodic endothelial glycocalyx layer (EGL) is typically less than 0.3 Ɋm thick, with key components associated with the endothelial cell membrane. The glycocalyx or EGL forms the primary molecular filter between circulating blood and the body tissues. Direct optical

Endothelium

87

microscopy reveals endothelial surface layers (ESLs) with porous outer layers that extend 1–2 Ɋm beyond the inner EGL. Such two-layer structures can have permeability properties that do not fully account for measured water and plasma exchange in microvessels. Multilayer models are therefore appropriate in future research into transendothelial transport of solute and solvents and in imaging the components of the EGL and the ESL. It is already becoming apparent that the thickness and distribution of thick ESLs in vessels with diameters larger than 50 Ɋm should not be presumed to reflect functional changes in the inner glycocalyx layer of microvascular endothelial cells [30].

Volume and thickness of the endothelial surface layer The endothelial surface layer is thinner where it covers the microcirculation (as little as 0.2 Ɋm) and thicker in larger vessels (up to 8 Ɋm). The healthy endothelial surface layer is impermeable to Dextran molecules of 70 kDa or more, and the endothelial surface layer–plasma boundary can be visualized as that part of the intravascular space that excludes fluorescein-labelled Dextran 70. Red blood cells are also excluded from the endothelial surface layer, and the intravascular red cell exclusion volume is larger than the Dextran 70 exclusion volume. Dextran 40 is small enough not to be excluded by the endothelial surface layer, and studies measuring the distribution volumes of Dextran 40 and erythrocytes in human subjects indicate an endothelial surface layer in health of about 1700 ml, much larger than the indocyanine green dilution method. From indocyanine green dilution studies of patients, the human endothelial surface layer volume has been estimated to be about 700 ml and, presuming that the endothelial surface area approximates 350 m2, an average thickness of about 2 Ɋm is suggested. Microvascular endothelial surface layer thickness can be measured in sublingual tissues of patients using orthogonal polarisation spectral imaging, and correlates well with dilution estimations. Despite advances, visualisation of the endothelial glycocalyx ultrastructure remains technically demanding. The glycocalyx is notoriously difficult to visualise in electron microscopic preparations: different methods of preparation give different results, and indirect estimates of its thickness based on in vivo observations carry a large standard error.

88

Chapter 4

The in vivo endothelial surface layer is constantly changing in volume with blood pressure and flow variations. Conventional optical and electrical microscopy of such a structure in the order of 10 to 100 nm provide imperfect pictures following fixation. Laboratory methods can be used to assess glycocalyx injury by measuring levels of its degradation products (e.g. syndecan-1, heparan sulphate and hyaluronan sulphate), mostly in the plasma. However, their physiological turnover disqualifies them from being a fully reliable measure of glycocalyx damage. Orthogonal polarization spectral imaging This method was developed to exploit haemoglobin absorption to visualise red blood cells within the microcirculation with reflected light. Vessel diameter and the functional capillary density can be estimated across a wide range of haemoglobin dilution [31]. In a modification called sidestream dark field, a light guide imaging the microcirculation is surrounded by light-emitting diodes (wavelength 530 nm). Light of that wavelength is absorbed by haemoglobin so that erythrocytes can be clearly observed as flowing cells. Video microscopy using sidestream dark-field imaging At the bedside, in vivo video microscopy technologies such as Sidestream Dark Field imaging allow indirect assessment of endothelial surface layer thickness in the oral mucosal microcirculation by measuring the penetration extent of flowing red blood cells into the endothelial surface layer. A video microscope is applied to the buccal mucosa or to a surgically-exposed tissue surface and shows real-time movement of red blood cells as they travel through microvessels. In a commercial system called GlycocheckTM, the video microscope feeds data to a microcomputer which records, detects and analyses blood vessels in the range of 5 to 25 μm. The automatic analysis detects the centre of the lumen of every blood vessel and the associated outer boundaries. The distance between the red blood cell column and this outer boundary is identified as the perfused boundary region (PBR). The PBR value is calculated on 3,000 individual positions, which makes the measurement results sensitive, yet very reproducible. A higher PBR value corresponds with a decrease in the thickness of the glycocalyx layer. The observed perfused vessel density is believed to be indicative of adequate microcirculation. In a patient study the PBR increased

Endothelium

89

(glycocalyx was thinned) after both pulsatile and nonpulsatile cardiopulmonary bypass [32]. The effects of colloid solutions of albumin and hydroxyethyl starch on the mechanical properties of bovine lung endothelial glycocalyx have been studied. Glycocalyx thickness increased significantly between 0.1% and 1% hydroxyethyl starch, similar to the effect of increasing concentrations of albumin. Reflectance interference contrast microscopy revealed a concentration-dependent softening of the glycocalyx in the presence of albumin, but a concentration-dependent increase in stiffness with hydoxyethyl starch. After glycocalyx degradation with hyaluronidase, stiffness was increased only at 4% albumin and 1% hydroxyethyl starch [33]. Intravital microscopy The ability to examine living tissues in situ opens up many opportunities. Advances in probe technology, and the ability to label sub-cellular structures, have extended imaging capabilities from biological processes at the level of the tissue and individual cells to the dynamics of intracellular organelles in several organs [34]. Intravital microscopy of pulmonary capillaries in mice found that the pulmonary ESL was significantly thicker (around about 1 μm) than systemic continuous capillary ESL [35]. The effects of murine sepsis have been imaged with a pulmonary microcirculation imaging system. Real-time cellular-level microscopic imaging of the lung was successfully performed, providing a clear identification of individual erythrocytes flowing in pulmonary capillaries. At subcellular level pulmonary ESL was identified by fluorescence angiography using a dextran conjugated fluorophore to label blood plasma [36].

The endothelial surface layer; inflation and disintegration An inflated endothelial surface layer The volume of the endothelial surface layer is inflated by solvent, and resists compression by, for example, erythrocytes. In addition to solvent being filtered, the GAGs take up water molecules and so swell. The glycoproteins and proteoglycans in the bush-like structures

90

Chapter 4

comprising the glycocalyx have a flexural rigidity that is sufficiently stiff to serve as a molecular filter for plasma proteins and as an exquisitely designed transducer of fluid shearing stresses. That rigidity cannot, however, prevent the buckling of protein structures during the motion of red cells or the penetration of white cell microvilli. In these cellular interactions, the viscous draining resistance of the glycocalyx creates a fluid draining pressure that is essential for preventing adhesive molecular interactions between proteins in the endothelial membrane and circulating cellular components [37]. Disintegration The endothelial surface layer is quite vulnerable and can disintegrate after application of various stressors, such as endotoxins, ischaemia/ hypoxia/ reperfusion, and oxidative stress. Other inflammatory mediators which have been implicated so far include C-reactive protein, A3 adenosine receptor stimulation, tumour necrosis factor, bradykinin, and mast cell tryptase. It appears on the evidence from human studies to date that the endothelial glycocalyx is compromised in systemic inflammatory states such as diabetes, hyperglycaemia, surgery, trauma, and sepsis. Indeed, even normal aging is associated with many aspects of endothelial dysfunction seen in the chronic and acute disease states [38]. Serial measurements of hyaluronan and syndecan are significant prognostic markers for morbidity and survival in sepsis [39]. Glycocalyx disintegration predisposes to leukocyte adhesion, emigration, and tissue infiltration by polymorphonuclear cells, monocyte/macrophages, and lymphocytes. It also leads to hyperactivation of plasma membrane receptors by exposure of ligands and further activation of danger signalling by endothelial cells. In a laboratory study time lapse video microscopy showed that exposure of endothelium to lipopolysaccharide rapidly triggers a surge in reactive oxygen species and exocytosis of lysosomes and Weibel Palade Bodies, which are storage organelles for von Willebrand factor. This was associated with a colocalized loss of endothelial surface layer, which could be prevented by the early blockade of exocytosis [40]. Researchers in New York have used a super resolution fluorescence optical microscope, Nikon’s Stochastic Optical Reconstruction Microscope (STORM), a type of single molecule localization microscopy [41]. STORM can overcome the diffraction barrier in conventional fluorescence microscopy.

Endothelium

91

Therapeutic options for the protection or restoration of the glycocalyx emerge from such studies. N-acetyl cysteine, antithrombin or hydrocortisone, and even sevoflurane anaesthesia could be beneficial. Compacted endothelial surface layer volume can be alleviated by infusion of the GAGs chondroitin sulphate and hyaluronic acid.

Functions of the endothelial glycocalyx Protecting glycocalyx functionality has become a credible therapeutic objective in the treatment of the most severely-ill patients [42]. Mechanosensing The mechanical forces imparted by blood flow to endothelial cells, smooth muscle cells, and circulating blood cells are sensed and can then elicit biochemical responses. The process is referred to as mechanotransduction. The cytoskeleton, largely composed of actin strands, connects the cell nucleus to the cell membrane and membranebound proteins. • •

•

Regulation of blood pressure by nitric oxide mediated vasodilation. Control of vascular permeability by the interaction with plasmaborne sphingosine-1 phosphate (S1P), which has been shown to regulate both the composition of the endothelial glycocalyx and the inter-endothelial junctions. Control of leukocyte recruitment during immunosurveillance and inflammation.

Molecules of the glycocalyx that lines all blood vessel walls are mechanotransducers and modulate blood cell interactions with the endothelial cell surface [43]. Coagulation Classic descriptions of platelet activation and the coagulation cascades start with endothelial surface layer and endothelial cellular damage that exposes sub-endothelial tissues. Platelets bind directly to collagen fibres of the extracellular matrix with collagen-specific glycoprotein Ia/IIa surface receptors, and the bond is

Chapter 4

92

further strengthened by von Willebrand factor (vWF), which is released from the endothelial cells and from platelets. The tissue factor pathway is initiated when circulating factor seven (FVII) complexes with tissue factor expressed on stromal fibroblasts and leukocytes, forming an activated complex that in turn activates FIX and FX. FXa complexes with co-factor FVa to form the prothrombinase complex, which generates thrombin from circulating prothrombin, with the release of activation fragments. Thrombin then activates other components of the coagulation cascade, including FV and FVIII (which forms a complex with FIX), and activates and releases FVIII from being bound to vWF. FVIIIa is the co-factor of FIXa, and together they form the tenase complex, which activates FX; and so, the cycle continues. FVII activation is amplified by thrombin, FXIa, FXII and FXa. Researchers in Amsterdam showed in human volunteer studies that acute hyperglycaemia disintegrates the endothelial glycocalyx (reduced endothelial surface layer volume and raised plasma hyaluronan levels) and activates coagulation, indicated by raised prothrombin activation fragment 1 + 2 [44]. There are five important regulators of platelet activation and the coagulation cascade: •

•

•

•

Prostacyclin (mostly in the form of PGI2) is a vasodilator eicosanoid released by endothelium that activates platelet receptors that inhibit platelet aggregation. The major vasoconstrictor eicosanoid and platelet aggregator is thromboxane (mostly TXA2) Protein C is a vitamin K-dependent serine protease enzyme that is activated by thrombin into activated protein C (APC) while the Protein C and thrombin molecules are bound to a cell surface protein called thrombomodulin. APC, along with protein S and a phospholipid as cofactors, degrades FVa and FVIIIa. Antithrombin is a serine protease inhibitor that degrades the serine proteases thrombin, FIXa, FXa, FXIa, and FXIIa. It is constantly active, but its adhesion to these factors is increased by the presence of heparan sulphate. Tissue factor pathway inhibitor limits the action of tissue factor and inhibits excessive tissue factor-mediated activation of FVII

Endothelium

•

93

and FX. Plasmin is generated by proteolytic cleavage of plasminogen. This cleavage is catalysed by tissue plasminogen activator, which is synthesized and secreted by endothelium. Plasmin proteolyses fibrin into fibrin degradation products.

It is worth noting at this point that blood dilution with normal saline produces an in vitro increase in coagulability as measured by thrombelastography [45]. The strongest responder patients experience a higher incidence of deep vein thrombosis after abdominal surgery, and intra-operative intravenous saline therapy predisposes patients to thrombosis [46]. Intravenous fluid bolus therapy is a contributor to glycocalyx perturbation that has direct effects on coagulation and fibrinolytic responses. The ability of von Willebrand factor (vWF) to trap platelets is due to the dynamic change from a globular conformation to an elongated fibre. Fibre formation is favoured by the anchorage of vWF to the endothelial cell surface, and vWF-platelet aggregates on the endothelium contribute to inflammation, infection, and tumour progression. The endothelial glycocalyx controls platelet recruitment through the tethering of vWF. Acute severe trauma patients may present with coagulopathy attributable to auto-heparinisation from endothelial glycocalyx disintegration [18]. The proposed syndrome of shock induced endotheliopathy (SHINE) includes coagulopathy [12]. Sites of low shear stress in the arterial tree are more susceptible to atheroma, presumably because of reduced nitric oxide generation. Chronically impaired endothelial glycocalyx mechanotransduction resulting in reduced nitric oxide production could therefore be an early step in the atherothrombotic process. Hyperglycaemia, hyperlipidaemias and hyperhomocysteinaemia may all be contributory. Exercise, by increasing blood flow and shear stress to generate nitric oxide, is protective against atheroma [47]. Leucocyte-endothelial cell adhesion The leucocyte–endothelial cell adhesion cascade brings about the migration of circulating immune cells from plasma to the interstitium. In the presence of bacterial antigens, circulating white blood cells

Chapter 4

94

release macrophage migration inhibition factor (an inflammatory cytokine), which then binds to CD74 in the endothelial cell membrane and on other immune cells to trigger an acute immune response. Trauma has been associated with pituitary release of macrophage migration inhibitory factor. Circulating leucocytes have long been known to cross the capillary and venular endothelium, a process called trafficking, and the roles of the cellular adhesion molecules that facilitate this migration are emerging. Clearly, white blood cells are capable of penetrating the endothelial glycocalyx in a way that their red blood cell cousins cannot. A sequence of molecular interactions between leucocyte and endothelium brings about leucocyte migration out of the circulation. Three families of cell adhesion molecules (CAMs) are recognised. They are 1. 2. 3.

the selectins, the integrins, and the immunoglobulin superfamily.

Sodium regulation Interest in the non-osmotic storage of sodium by interstitial glycosaminoglycans raises the possibility that GAGs of the endothelial surface layer could have a role as buffers between plasma and interstitial sodium levels [48]. Ion transport through the endothelial glycocalyx layer is probably abnormal in some cardiovascular diseases and researchers at University College London have investigated the response of ion transport to the changing blood flow velocity and the shedding of endothelial glycocalyx constituents [10]. Increasing blood flow brings about a conformational change of endothelial glycocalyx sugar chains that promotes Na+ transport from the inner region of the endothelial glycocalyx layer to the free-flowing plasma. In this text I have presumed that the transendothelial osmotic pressure difference that affects the solvent filtration rate Jv is exclusively due to larger molecules, the colloid osmotic pressure. Based on these findings, the Starling principle is further revised to include flow-dependent noncolloid osmotic pressure differences. The researchers estimate that physiological flow regulates the osmotic part of transendothelial water flux by about 8% compared with the stationary situation.

Endothelium

95

Vascular endothelial cells If we are to think of the endothelium as an organ, weight estimates range from several hundred grams to as much as a kilogram in the average adult. Endothelial cells are flattened polygons in shape, joined to one another edge to edge, and in the capillary beds average only 0.3 microns in thickness. Capillary endothelial surface area has been variously estimated at 400 to 800 m². In ball park terms, we may think of the endothelial surface area as being at least two full size tennis courts. The endoplasmic reticulum of endothelial cells is within a few nanometres of the cell membrane, and contains a small store of calcium ions. Cigar-shaped Weibel-Palade bodies are unique to endothelial cells, and appear to be storage organelles for the glycoprotein von Willebrand factor and for the leucocyte-adhesive protein P-selectin, among others. Their number, and their contents, vary from tissue to tissue, but in general terms we may see them as part of the endothelial response to coagulation, inflammation, vascular tone and angiogenesis [49]. Endothelial cell shape is maintained by an actin cytoskeleton, which is organised into three distinct systems; the cortical web, the junction band and basal stress fibres. The cortical web is the thin layer of actin filaments just beneath the cell membrane, and is linked to glycoproteins that pass through the cell membrane and are part of the endothelial glycocalyx. This linkage enables information about blood flow to be transmitted to the inside of the endothelial cell. The junctional band is a prominent ring of actin running around the cell perimeter. It is attached to the intercellular junction proteins via alpha actin microfilaments and anchors them in place. The junctional band is therefore important for cell to cell adhesion. These junctions are also served by cytoskeletal microtubules to facilitate both maintenance of barrier function and modulation of signal transduction in response to the tethering and contractile forces exerted on the endothelium.

Interendothelial junctions Tight junctions are formed by the fusion of the outer layers of the plasma membranes and are composed of occludins, claudins and junctional adhesion molecules coupled to cytoplasmic proteins and

96

Chapter 4

linked to the endothelial cell actin cytoskeleton by the zonula occludens family. The tight junctions occlude what would otherwise be a paracellular pathway for glycocalyx-filtered solvent and solutes, and so tight junction breaks (also known as tight junction gaps) are the site of paracellular convection. An increase in the length or number of intercellular tight junction breaks lowers the resistance of the endothelial barrier to solvent flow. Adherens junctions feature cadherins such as vascular endothelial cadherin (VE-cadherin), that bind intracellular catenin proteins (including p120-catenin, a VE-cadherin stabilising protein) that in turn bind to other protein partners in the actin cytoskeleton. Calciumdependent association of cadherin proteins occurs at adhesion junctions, regulating paracellular transport of cells and solutes between the blood and the interstitium. Adherens junctions, and specifically the VE-cadherin found there, regulate paracellular permeability and leucocyte transmigration. Platelet and endothelial adhesion molecule-1 (PeCAM-1) also contributes to interendothelial adhesion and permeability [50]. Inflammation predisposes to disassembly of intercellular junctions, a major contributor to ‘capillary leak’ and oedema of sepsis. London proposed that stabilising vascular barrier function in the presence of inflammatory cytokines could be a life-saving therapeutic strategy [51]. Connexins are trans-membrane proteins that allow intercellular communication and the transfer of ions and small signalling molecules between cells. Connexins have been shown to form functional hemichannels but have roles independent of channel functions. Endothelial cells are tethered to the extracellular matrix (ECM) via interaction between cell surface integrins and their ECM ligands, which are organised in focal adhesion plaques. Figure 4.4 is a cartoon that summarises interendothelial adhesion and the mechanisms that affect it. Regulation of nitric oxide (NO) synthesis is the major vasodilation pathway.

Endothelium

97

Figure 4.4

Aquaporin channels are present in endothelial cells We have noted in Chapter 3 the existence of membrane channels that mediate rapid water flux through cell membranes. Peter Agre at Johns Hopkins University shared the 2003 Nobel Prize in Chemistry for his discovery of the specific transmembrane water pore that was later called aquaporin-1. Most aquaporins are specific for cell membrane water permeability, but aqua-glyceroporins can additionally conduct some very small uncharged solutes such as glycerol, CO2, ammonia, and urea across the membrane. All aquaporins are impermeable to charged solutes. Aquaporins are synthesised within endothelial cells and inserted into the endothelial cell membrane. AQP1 is widely expressed in microvascular endothelia outside of the brain, but it is only in renal vasa recta that endothelial AQP1 is of clear physiological significance. With that exception, the aquaporins of microvascular endothelia appear to have no great functional significance. AQP4 has been identified in the capillaries of intestinal mucosa, is up-regulated by plasma hypertonicity induced by hypertonic saline infusion, and may play a part in the treatment of oedema [52].

98

Chapter 4

Aquaporin channels of astrocyte foot processes are important to blood-brain permeability barrier AQP1 is expressed in choroid plexus epithelial cells where it may assist with cerebrospinal fluid secretion. AQP4 is present in astrocyte foot process membranes adjacent to capillary basement membranes, is constitutively active and therefore this cerebral perivascular pool of AQP4 is believed to be involved in normal regulation of brain water flux. AQP4 and other aquaporins have been implicated in ischaemic cerebral oedema formation [53–55].

Aquaporin 2 is expressed in the renal collecting duct In Chapter 3 we met AQP2 which is expressed in the principle cells of the collecting ducts within the kidney and is key to body water balance regulation by arginine vasopressin. Arginine Vasopressin phosphorylates the aquaporin-2 water channel; and increases the number of aquaporin-2 water channels in the membrane. The permeability of the membrane to water (its hydraulic conductivity) is thereby increased and water reabsorption enhanced. Diabetes insipidus can be a consequence of AQP2 deficiency.

Aquaporins in the lung Aquaporins are expressed in the lung and airways: AQP1 in pulmonary microvascular endothelia, AQP3 and AQP4 in airway epithelia, and AQP5 in type I alveolar epithelial cells, submucosal gland acini, and a subset of airway epithelial cells. However, they are not essential to the intravascular - extravascular disposition of water within lung and airways [56].

Capillary classes We looked historically at a system of classification of capillaries that was proposed by Bennett in the early days of electron microscopy. For the purposes of fluid physiology it makes more sense to classify the tissue capillaries in terms of their large pore size so that we can understand which macromolecules and cellular elements they may be

Endothelium

99

permeable to [57]. The physiologic upper limit of pore size in the capillary walls of most non-sinusoidal blood capillaries to the transcapillary passage of lipidinsoluble endogenous and non-endogenous macromolecules ranges between 5 and 12 nm. Macromolecules larger than the physiologic upper limits of pore size generally do not accumulate within the interstitial fluid and lymph of these tissues. At the other extreme, the reticuloendothelial sinusoidal blood capillaries of myeloid bone marrow permit non-endogenous macromolecules as large as 60 nm in diameter, and such molecules can distribute into the bone marrow interstitial space via the phago-endocytic route. The interstitial fluid circulation carries these macromolecules to the lymphatics of periosteal fibrous tissues, and thence to the locoregional lymph nodes. As in my original communication, I adopt Sarin’s classification of the blood capillary microvasculature [57]. Non-sinusoidal (continuous) microvasculature There is a continuous surface layer that includes the anionic glycocalyx matrix on the endothelial cell surfaces, and a continuous anionic basement membrane (featuring lamina densa and lamina reticularis) that is rich in sulphated proteoglycans condensed from the interstitial matrix. Breaks (where present) within the inter-endothelial cell tight junctions constitute the primary pathways for transvascular fluid filtration, and the increased porosity seen in inflammation may be due to an increase in these normally infrequent discontinuities. An alternative interpretation of pore theory, called the ‘glycocalyxjunction-break model’, proposes that pore size (small or large) is a function of the spaces between the matrix fibres of the endothelial glycocalyx, while the area for fluid exchange is a function of the length of the junction breaks between adjacent endothelial cells. Figure 4.5 is a cartoon of the continuous microvasculature pathways for solute and solvent passage between plasma and the interstitial fluid.

Chapter 4

100

Figure 4.5

• •

•

Lipophilic solutes include gases such as oxygen and carbon dioxide; they easily cross cell membranes along diffusion gradients. Water and small lipophobic solutes from inorganic ions to glucose and smaller hormone molecules move by convection through interendothelial clefts (paracellular transport) and through fenestrations (Latin for windows). In the cartoon I have indicated two-way movement through diaphragm fenestrations, as such pathways are present where epithelial secretions bring extra solvent to the interstitial fluid and endothelial absorption can occur; an exception to the no-absorption rule. Large lipophobic solutes. The glycocalyx matrix is less than perfect (as quantified by Staverman’s reflection coefficient which may be only 0.7 to albumin in pulmonary capillaries), and so some such molecules pass through the filter with the paracellular flow of solvent filtrate. Inflammation degrades the glycocalyx filter, creates more tight junction strand gaps, and loosens adherens junctions increasing the paracellular flow of larger lipophobic solutes. Breakdown of the endothelial surface layer and the interendothelial adhesions creates inflammatory gaps, seen particularly in non-invested venules. Transcellular pathways are available for circulating protein transport via

Endothelium

101

transendothelial channels and protein-specific vesicles (caveolae). Non-sinusoidal non-fenestrated microvasculature These endothelial cells have no fenestrations. Non-sinusoidal nonfenestrated capillaries of muscles, connective tissues, and lungs have macula occludens loose junctions to their intercellular clefts and incomplete cellular investment. The effective pore size there is up to 5 nm, making them permeable to molecules as large as myoglobin [58]. The tissues that can accumulate substantial amounts of interstitial fluid after trauma and sepsis (i.e. the more compliant tissues) are loose connective tissues, muscles, lungs, and gastrointestinal mesentery and mucosa. For example, extra-vascular lung water measured by double indicator dilution can increase from around 500 ml to 2.5 litre in pulmonary oedema, while the loose connective tissues and muscles can expand to many litres of peripheral oedema. Non-fenestrated (continuous) capillaries with multiple tight junctions a. Primary anatomic sites of transvascular flow are zona occludens interendothelial junctions (paracellular) with tight opposition of adjacent endothelial cell membranes. b. Determinants of physiologic pore size are the zona occludens interendothelial cell junctions in series constitute an absolute barrier to the transvascular flow of macromolecules; the bloodbrain barrier. c. Physiologic upper limit of pore size is less than 1 nm. d. Representative tissue microvascular beds include retinal, brainspinal cord, nerve endoneurium, enteric nervous system and lymphoid tissue cortex. Non-fenestrated (continuous) capillaries with loose junctions a. Primary anatomic sites of transvascular flow are macula occludens interendothelial junctions (paracellular) with loose opposition of adjacent endothelial cell membranes at junctions. A single tight junction strand has occasional junction strand gaps which funnel filtered fluid to the interendothelial cleft and the interstitium beyond. b. Determinants of physiologic pore size are the condition of the glycocalyx and the macula occludens interendothelial cell

102

Chapter 4

junctions in series, which constitute a relative barrier to the transvascular flow of macromolecules. c. Physiologic upper limit of pore size is about 5 nm. d. Representative tissue microvascular beds include skin, muscle, cortical bone, adipose tissue, lung, and intestinal mesentery. These are the majority of capillaries within the systemic circulation, and almost all of the capillaries within the pulmonary circulation. Endothelial investment by pericytes was once believed to be a feature exclusive to brain capillaries, but pericytes are found in the lungs and perhaps in other tissues. Even if they don’t create a full investment, pericytes can localise to interendothelial cleft exits where they can contain the sub-glycocalyx protected region and accelerate the steadystate mechanism. Endothelial cells may undergo phenotype changes in response to physical and chemical stresses (including inflammation), which contribute to endothelial dysfunction. In particular, continuous capillaries can develop fenestrations. Non-sinusoidal fenestrated microvasculature The endothelial cells are fenestrated with spaces rich in sulphated proteoglycans. The continuous anionic glycocalyx is rich in sialyated glycoproteins, and the anionic basement membrane is rich in sulphated proteoglycans. Fenestrated capillaries with diaphragmed fenestrae a. Primary anatomic sites of transvascular flow are transcellular diaphragmed fenestrae. The diameters of the fenestrae are 60 to 80 nm, including the closed membranous central diaphragms which are 10 to 30 nm wide. Eight to twelve membranous septae (2 to 7 nm wide) radiate out from the central diaphragm giving the appearance of a cartwheel. The arc widths of fenestrated open spaces are 6 to 12 nm. b. Determinants of physiologic pore size are the arc widths which constitute the barriers to the transvascular flow of macromolecules and the condition of the anionic glycocalyx. c. Physiologic upper limit of pore size is therefore 6-12 nm. d. Representative tissue microvascular beds include exocrine glands, renal peritubular, endocrine glands, intestinal mucosa, peripheral ganglia, nerve epineurium, circumventricular organs, choroid plexus and the ciliary process of the eye.

Endothelium

103

These are tissues that acquire interstitial fluid from cellular exocytosis (for example, endocrine glands) or transepithelial absorption (gastrointestinal mucosa). This fluid can be absorbed into plasma. Levick’s classic analysis included capillaries of the intestinal mucosa that predominantly filter intravascular solvent and solutes to the interstitium during fasting, and absorb when water and nutrients are available in the gut lumen. Renal glomerular capillaries (open fenestrations) The open fenestrations have a diameter of about 65 nm, but in health the glycocalyx that overhangs the edge of the fenestration reduces the effective pore size to about 15 nm. Numerically a very small fraction of the total capillaries, but taking around 20% of the systemic circulation they filter over 100 ml of solvent and solute to the renal tubules every minute in healthy humans; the glomerular filtration rate. Slit diaphragms at the level of podocyte foot processes restrict the filtration of plasma proteins larger than 6 nm in diameter (e.g. haemoglobin and albumin). Sinusoidal reticuloendothelial microvasculature The endothelial cells are of phago-endocytic phenotype and there is a non-continuous anionic glycocalyx at non-endocytic sites. The glycocalyx contains sialyated glycoproteins, but the endothelial cells express uptake receptors for hyaluronic acid, and by actively removing this important glycosaminoglycan they prevent development of an effective endothelial surface layer. The basement membrane is commonly absent. At steady-state there is no colloid osmotic pressure difference between plasma and interstitial fluid of sinusoidal reticuloendothelial tissues. This self-evident fact is often ignored in critical care physiology teaching. In norepinephrine-supported patients with sepsis splanchnic blood flow is around 20% of cardiac output in moderate shock, and rising towards 50% in severe shock [59]. An increase in blood circulatory rate or in the proportion of left ventricular outflow going to sinusoidal tissues will increase the escape rate of larger molecules such as albumin and give the clinician the mistaken impression that systemic capillaries have become leakier. •

Hepatic sinusoidal blood capillary. Primary anatomic sites of transvascular flow are open transcellular fenestrae whose diameters are variable across mammalian species. In the human

104

•

Chapter 4

the range is 50 to 180 nm. With no basement membrane underlying the fenestrae and a lack of glycocalyx matrix fibres within fenestrae, the upper limit of pore size for transendothelial flow is around 180 nm. Much larger molecules can be phagocytosed. Myeloid bone marrow sinusoidal blood capillary. Primary anatomic sites of transvascular flow are the interendothelial junctions. These endothelial cells are only transiently fenestrated during the transcellular passage of blood cells. They are of phago-endocytic phenotype and have macula occludens interendothelial junctions. The physiologic upper limit of pore size for transendothelial flow is therefore only 5 nm, but cells can relatively easily pass by phago-endocytosis. The intra-osseous route for fluid infusion is said to be appropriate for any crystalloid, colloid or blood cell infusion.

Sinusoidal non-reticuloendothelial microvasculature These endothelial cells show only low levels of phago-endocytosis. There is a thin anionic glycocalyx and the basement membrane is discontinuous. •

•

Splenic red pulp arterial capillaries are also known as terminal capillaries. Primary anatomic sites of transvascular flow are the terminal capillary ending openings about 5 Ɋm in diameter. There are macrophages in the terminal arterial pericapillary sheath and within the splenic red pulp reticular meshwork. The basement membrane is sparse. Macromolecules as large as 5 Ɋm pass into splenic red pulp reticulum through the terminal capillary ending openings, and exogenous macromolecules can be phagocytosed by the pericapillary sheath macrophages. Splenic red pulp venous capillaries (sinus capillaries) feature cuboidal endothelial cells. The primary anatomic sites of transvascular flow are interendothelial slits between apical and basal adherens junctions. The basement membrane is ringed, with belts of basement membrane rings 2-3 Ɋm apart. The slits are closed except during active blood cell migration and macrophage phagocytosis. There are few connections between splenic arterioles and venous capillaries and they are a minor pathway in splenic filtration. Exogenous macromolecules in the

Endothelium

105

sinus lumen are phagocytosed at level of the interendothelial slits by finger-like pseudopodia of splenic pulp reticulum macrophages.

Circulating endothelial cells and endothelial-derived microparticles Circulating endothelial cells derive both from mature endothelial cells and endothelial progenitor cells. The number of circulating endothelial cells increases in inflammatory diseases including sepsis [60]. Circulating endothelial cells from septic shock patients convert to fibroblasts [61]. Endothelial damage generates endothelial-derived microparticles which are relevant biomarkers of septic shock-induced disseminated intravascular coagulation [62, 63]. They are found during active cerebrovascular disease [64, 65] and after myocardial infarction with renal impairment [66]. They may also have a physiological role in maintaining vascular function. Endothelial microparticles belong to a family of extracellular vesicles that are dynamic, mobile, biological effectors capable of mediating vascular physiology and function [67]. Endothelial-derived microparticles can exert autocrine and paracrine effects on target cells through surface interaction, cellular fusion, and, possibly, the delivery of their intra-vesicular cargo. Moderate exercise reduces the number of circulating endothelial derived microparticles in healthy women [68].

The subendothelial basement membrane and extracellular matrix The glycocalyx is the first and the major fibre matrix resistor in the current of fluid and solutes between plasma and lymph. The basement membrane and extracellular matrix are the second and third resistances in a series. The basement membrane, where it exists, is a specialized part of the extracellular matrix 60 – 100 nm in thickness, composed of type IV collagen and laminin and closely adherent to the cell membrane. The extracellular matrix is essentially a web of collagen fibrils within the interstitial space upon which glycoproteins such as fibronectin and proteoglycans (protein molecules with GAG side

106

Chapter 4

chains) are arranged, and contain free GAGs. Toll-like receptors are found within the extracellular matrix and are believed to have a pivotal role in the early development of systemic inflammatory response and ventilator-induced lung injury. Integrins and their receptors modulate cell locomotion through the extracellular matrix, and it has been discovered that they can modulate the interstitial pressure. Connective tissues are held under mild compression by collagen fibrils that are beta 1-integrin bonded to fibroblasts and oppose the inherent tendency of interstitium to expand (glycosaminoglycan swelling). Fibril detachment serves to reduce interstitial pressure to more subatmospheric values. Supporting evidence includes a fall in interstitial pressure following treatment of the tissues with beta 1integrin antibodies. In the next Chapter we consider the interstitium and lymphatic system in greater detail.

References 1. Aird WC. Endothelium as an organ system. Crit Care Med. 2004;32:S271-9. 2. Bennett HS, Luft JH, Hampton JC. Morphological classifications of vertebrate blood capillaries. Am J Physiol. 1959;196:381-390. 3. Crocket E. Learners Corner: Endothelial Glycocalyx and the Revised Starling Principle. PVRI Chronicle. 2014;1:issue 2. 4. Tarbell JM, Cancel LM. The glycocalyx and its significance in human medicine. J Intern Med. 2016;280:97-113. 5. Squire JM, Chew M, Nneji G, Neal C, Barry J, Michel C. Quasi-periodic substructure in the microvessel endothelial glycocalyx: a possible explanation for molecular filtering. J Struct Biol. 2001;136:239-255. 6. Weinbaum S, Tarbell JM, Damiano ER. The structure and function of the endothelial glycocalyx layer. Annu Rev Biomed Eng. 2007; 9:121-167. 7. Arkill KP, Knupp C, Michel CC et al. Similar endothelial glycocalyx structures in microvessels from a range of mammalian tissues: evidence for a common filtering mechanism. Biophys J. 2011; 101:1046-1056. 8. Curry FE. Layer upon layer: the functional consequences of disrupting the glycocalyx-endothelial barrier in vivo and in vitro. Cardiovasc Res. 2017;113:559-561.

Endothelium

107

9. Curry FE, Michel CC. The endothelial glycocalyx: Barrier functions versus red cell hemodynamics: A model of steady-state ultrafiltration through a bi-layer formed by a porous outer layer and more selective membrane-associated inner layer. Biorheology. 2019 10. Jiang XZ, Ventikos Y, Luo KH. Microvascular Ion Transport through Endothelial Glycocalyx Layer: New Mechanism and Improved Starling Principle. Am J Physiol Heart Circ Physiol. 2019 11. Torres Filho IP, Torres LN, Salgado C, Dubick MA. Plasma syndecan1 and heparan sulfate correlate with microvascular glycocalyx degradation in hemorrhaged rats after different resuscitation fluids. Am J Physiol Heart Circ Physiol. 2016;310:H1468-78. 12. Johansson PI, Stensballe J, Ostrowski SR. Shock induced endotheliopathy (SHINE) in acute critical illness - a unifying pathophysiologic mechanism. Crit Care. 2017;21:25. 13. Osuka A, Kusuki H, Yoneda K et al. Glycocalyx Shedding is Enhanced by Age and Correlates with Increased Fluid Requirement in Patients with Major Burns. Shock. 2018;50:60-65. 14. Pesonen E, Passov A, Andersson S et al. Glycocalyx Degradation and Inflammation in Cardiac Surgery. J Cardiothorac Vasc Anesth. 2019;33:341-345. 15. Gonzalez Rodriguez E, Cardenas JC, Cox CS et al. Traumatic brain injury is associated with increased syndecan-1 shedding in severely injured patients. Scand J Trauma Resusc Emerg Med. 2018;26:102. 16. Betteridge KB, Arkill KP, Neal CR et al. Sialic acids regulate microvessel permeability, revealed by novel in vivo studies of endothelial glycocalyx structure and function. J Physiol. 2017;595:5015-5035. 17. Hippensteel JA, Anderson BJ, Orfila JE et al. Circulating heparan sulfate fragments mediate septic cognitive dysfunction. J Clin Invest. 2019 18. Ostrowski SR, Johansson PI. Endothelial glycocalyx degradation induces endogenous heparinization in patients with severe injury and early traumatic coagulopathy. J Trauma Acute Care Surg. 2012;73:60-66. 19. Plevris JN, Haydon GH, Simpson KJ et al. Serum hyaluronan--a noninvasive test for diagnosing liver cirrhosis. Eur J Gastroenterol Hepatol. 2000;12:1121-1127. 20. Broekhuizen LN, Lemkes BA, Mooij HL et al. Effect of sulodexide on endothelial glycocalyx and vascular permeability in patients with type 2 diabetes mellitus. Diabetologia. 2010;53:2646-2655.

108

Chapter 4

21. Becker BF, Jacob M, Leipert S, Salmon AH, Chappell D. Degradation of the endothelial glycocalyx in clinical settings: searching for the sheddases. Br J Clin Pharmacol. 2015;80:389-402. 22. Zeng Y, Adamson RH, Curry FR, Tarbell JM. Sphingosine-1phosphate protects endothelial glycocalyx by inhibiting syndecan-1 shedding. Am J Physiol Heart Circ Physiol. 2014;306:H363-72. 23. Curry FR, Adamson RH. Tonic regulation of vascular permeability. Acta Physiol (Oxf). 2013;207:628-649. 24. Zhang L, Zeng M, Fan J, Tarbell JM, Curry FR, Fu BM. Sphingosine-1phosphate Maintains Normal Vascular Permeability by Preserving Endothelial Surface Glycocalyx in Intact Microvessels. Microcirculation. 2016;23:301-310. 25. Alves NG, Trujillo AN, Breslin JW, Yuan SY. Sphingosine-1Phosphate Reduces Hemorrhagic Shock and Resuscitation-Induced Microvascular Leakage by Protecting Endothelial Mitochondrial Integrity. Shock. 2018 26. Gomes L, Fernando S, Fernando RH et al. Sphingosine 1-phosphate in acute dengue infection. PLoS One. 2014;9:e113394. 27. Winkler MS, Nierhaus A, Holzmann M et al. Decreased serum concentrations of sphingosine-1-phosphate in sepsis. Crit Care. 2015;19:372. 28. Fisher J, Linder A, Bentzer P. Elevated plasma glypicans are associated with organ failure in patients with infection. Intensive Care Med Exp. 2019;7:2. 29. Stone ML, Sharma AK, Zhao Y et al. Sphingosine-1-phosphate receptor 1 agonism attenuates lung ischemia-reperfusion injury. Am J Physiol Lung Cell Mol Physiol. 2015;308:L1245-52. 30. Curry FE. The Molecular Structure of the Endothelial Glycocalyx Layer (EGL) and Surface Layers (ESL) Modulation of Transvascular Exchange. Adv Exp Med Biol. 2018;1097:29-49. 31. Harris AG, Sinitsina I, Messmer K. Validation of OPS imaging for microvascular measurements during isovolumic hemodilution and low hematocrits. Am J Physiol Heart Circ Physiol. 2002;282:H15029. 32. Koning NJ, Vonk AB, Vink H, Boer C. Side-by-Side Alterations in Glycocalyx Thickness and Perfused Microvascular Density During Acute Microcirculatory Alterations in Cardiac Surgery. Microcirculation. 2016;23:69-74. 33. Job KM, O’Callaghan R, Hlady V, Barabanova A, Dull RO. The Biomechanical Effects of Resuscitation Colloids on the

Endothelium

109

Compromised Lung Endothelial Glycocalyx. Anesth Analg. 2016;123:382-393. 34. Masedunskas A, Milberg O, Porat-Shliom N et al. Intravital microscopy: a practical guide on imaging intracellular structures in live animals. Bioarchitecture. 2012;2:143-157. 35. Yang Y, Schmidt EP. The endothelial glycocalyx: an important regulator of the pulmonary vascular barrier. Tissue Barriers. 2013;1. http:// dx.doi.org/10.4161/tisb.23494 36. Park I, Choe K, Seo H et al. Intravital imaging of a pulmonary endothelial surface layer in a murine sepsis model. Biomed Opt Express. 2018;9:2383-2393. 37. Weinbaum S, Zhang X, Han Y, Vink H, Cowin SC. Mechanotransduction and flow across the endothelial glycocalyx. Proc Natl Acad Sci U S A. 2003;100:7988-7995. 38. Machin DR, Bloom SI, Campbell RA et al. Advanced age results in a diminished endothelial glycocalyx. Am J Physiol Heart Circ Physiol. 2018;315:H531-H539. 39. Anand D, Ray S, Srivastava LM, Bhargava S. Evolution of serum hyaluronan and syndecan levels in prognosis of sepsis patients. Clin Biochem. 2016;49:768-776. 40. Zullo JA, Fan J, Azar TT et al. Exocytosis of Endothelial LysosomeRelated Organelles Hair-Triggers a Patchy Loss of Glycocalyx at the Onset of Sepsis. Am J Pathol. 2016;186:248-258. 41. Xia Y, Fu BM. Investigation of Endothelial Surface Glycocalyx Components and Ultrastructure by Single Molecule Localization Microscopy: Stochastic Optical Reconstruction Microscopy (STORM). Yale J Biol Med. 2018;91:257-266. 42. Milford EM, Reade MC. Resuscitation Fluid Choices to Preserve the Endothelial Glycocalyx. Crit Care. 2019;23:77. 43. Tarbell JM, Simon SI, Curry FR. Mechanosensing at the vascular interface. Annu Rev Biomed Eng. 2014;16:505-532. 44. Nieuwdorp M, van Haeften TW, Gouverneur MC et al. Loss of endothelial glycocalyx during acute hyperglycemia coincides with endothelial dysfunction and coagulation activation in vivo. Diabetes. 2006;55:480-486. 45. Heather BP, Jennings SA, Greenhalgh RM. The saline dilution test--a preoperative predictor of DVT. Br J Surg. 1980;67:63-65. 46. Janvrin SB, Davies G, Greenhalgh RM. Postoperative deep vein thrombosis caused by intravenous fluids during surgery. Br J Surg. 1980;67:690-693.

110

Chapter 4

47. Noble MI, Drake-Holland AJ, Vink H. Hypothesis: arterial glycocalyx dysfunction is the first step in the atherothrombotic process. QJM. 2008;101:513-518. 48. Olde Engberink RH, Rorije NM, Homan van der Heide JJ, van den Born BJ, Vogt L. Role of the vascular wall in sodium homeostasis and salt sensitivity. J Am Soc Nephrol. 2015;26:777-783. 49. Metcalf DJ, Nightingale TD, Zenner HL, Lui-Roberts WW, Cutler DF. Formation and function of Weibel-Palade bodies. J Cell Sci. 2008;121:19-27. 50. Stoyanoff TR, Rodríguez JP, Todaro JS, Colavita JPM, Torres AM, Aguirre MV. Erythropoietin attenuates LPS-induced microvascular damage in a murine model of septic acute kidney injury. Biomed Pharmacother. 2018;107:1046-1055. 51. London NR, Li DY. Robo4-dependent Slit signaling stabilizes the vasculature during pathologic angiogenesis and cytokine storm. Curr Opin Hematol. 2011;18:186-190. 52. Radhakrishnan RS, Shah SK, Lance SH et al. Hypertonic saline alters hydraulic conductivity and up-regulates mucosal/submucosal aquaporin 4 in resuscitation-induced intestinal edema. Crit Care Med. 2009;37:2946-2952. 53. Zador Z, Stiver S, Wang V, Manley GT. Role of aquaporin-4 in cerebral edema and stroke. Handb Exp Pharmacol. 2009159-170. 54. Zeynalov E, Chen CH, Froehner SC et al. The perivascular pool of aquaporin-4 mediates the effect of osmotherapy in postischemic cerebral edema. Crit Care Med. 2008;36:2634-2640. 55. Nakayama S, Migliati E, Amiry-Moghaddam M, Ottersen OP, Bhardwaj A. Osmotherapy With Hypertonic Saline Attenuates Global Cerebral Edema Following Experimental Cardiac Arrest via Perivascular Pool of Aquaporin-4. Crit Care Med. 2016;44:e702-10. 56. Verkman AS. Role of aquaporins in lung liquid physiology. Respir Physiol Neurobiol. 2007;159:324-330. 57. Sarin H. Physiologic upper limits of pore size of different blood capillary types and another perspective on the dual pore theory of microvascular permeability. J Angiogenes Res. 2010;2:14. 58. Inagawa R, Okada H, Takemura G et al. Ultrastructural Alteration of Pulmonary Capillary Endothelial Glycocalyx During Endotoxemia. Chest. 2018;154:317-325. 59. De Backer D, Creteur J, Silva E, Vincent JL. Effects of dopamine, norepinephrine, and epinephrine on the splanchnic circulation in septic shock: which is best. Crit Care Med. 2003;31:1659-1667.

Endothelium

111

60. Yoo JW, Moon JY, Hong SB, Lim CM, Koh Y, Huh JW. Clinical significance of circulating endothelial cells in patients with severe sepsis or septic shock. Infect Dis (Lond). 2015;47:393-398. 61. Tapia P, Gatica S, Cortés-Rivera C et al. Circulating Endothelial Cells From Septic Shock Patients Convert to Fibroblasts Are Associated With the Resuscitation Fluid Dose and Are Biomarkers for Survival Prediction. Crit Care Med. 2019 62. Mastronardi ML, Mostefai HA, Meziani F, Martínez MC, Asfar P, Andriantsitohaina R. Circulating microparticles from septic shock patients exert differential tissue expression of enzymes related to inflammation and oxidative stress. Crit Care Med. 2011;39:17391748. 63. Delabranche X, Boisramé-Helms J, Asfar P et al. Microparticles are new biomarkers of septic shock-induced disseminated intravascular coagulopathy. Intensive Care Med. 2013;39:1695-1703. 64. El-Gamal H, Parray AS, Mir FA, Shuaib A, Agouni A. Circulating microparticles as biomarkers of stroke: A focus on the value of endothelial- and platelet-derived microparticles. J Cell Physiol. 2019 65. Kandiyil N, MacSweeney ST, Heptinstall S, May J, Fox SC, Auer DP. Circulating Microparticles in Patients with Symptomatic Carotid Disease Are Related to Embolic Plaque Activity and Recent Cerebral Ischaemia. Cerebrovasc Dis Extra. 2019;9:9-18. 66. Mörtberg J, Lundwall K, Mobarrez F, Wallén H, Jacobson SH, Spaak J. Increased concentrations of platelet- and endothelial-derived microparticles in patients with myocardial infarction and reduced renal function- a descriptive study. BMC Nephrol. 2019;20:71. 67. Curtis AM, Edelberg J, Jonas R et al. Endothelial microparticles: sophisticated vesicles modulating vascular function. Vasc Med. 2013;18:204-214. 68. Serviente C, Burnside A, Witkowski S. Moderate-intensity exercise reduces activated and apoptotic endothelial microparticles in healthy midlife women. J Appl Physiol (1985). 2019;126:102-110.

CHAPTER 5 INTERSTITIUM AND LYMPHATIC SYSTEMS

Chapter summary The structures of the triphasic interstitium and the lymphatic pumping mechanism are described in more detail. They are responsible for the rate of clearance of interstitial fluid to the plasma at the end of the vital interstitial fluid/ lymph circulation. Interstitium and the lymphatics are, together, the second regulator of the balance between intravascular and extravascular fluid volumes. Interstitium and the lymph system work with leucocytes to deliver an optimal immune state, continuously regenerate and adapt to tissue demands and are crucial to the healthy function of every traditional organ structure. Non-osmotic sodium storage in the interstitium is emerging as an important factor in sodium balance. Interesting hypotheses of interstitial matrix biology and lymphatic function as a unifying characteristic of critical illness pathophysiology are emerging. Finally, we consider the therapeutic potential of peritoneal dialysis, a treatment that enables a clinician to directly modify interstitial fluid volume and to remove unwanted solutes.

Interstitium The interstitium is the extracellular matrix within which reside tissue parenchymal cells. It accounts for about one sixth of the total body volume. As appreciation of the roles of the interstitium has grown there have been headline-grabbing calls to classify the interstitium as an organ. Oxygen and solutes in plasma traverse a transendothelial barrier to diffuse through interstitium to cells, while carbon dioxide and metabolic waste solutes diffuse from cells to the plasma. The fact of an extravascular circulation of extracellular fluid draws our attention to

Interstitium and Lymphatic Systems

113

the way interstitium channels the flow. The regulatory role of interstitium in salt and water homeostasis is an emerging subject of interest to clinicians. The perivascular extracellular matrix forms an organ specific vascular niche that orchestrates mechano-, growth factor, and angiocrine signalling required for tissue homeostasis and organ repair [1].

Molecular structure of the triphasic interstitium The composition of the interstitium is controlled by the regulation of synthesis and turnover of each of its individual components, driven by cytokines and growth factors. The major structural elements are collagen fibre bundles which are visible to light microscopy and can extend for long distances. The collagen triple helix consists of three intertwined polypeptide chains that entangle water molecules - collagen hydration. Collagen fibre bundles can therefore be considered one of the interstitial aqueous phases. Collagen bundles of several interstitial spaces have been reported to be associated with thin, flat cells (spindle shaped in cross section) that have scant cytoplasm and an oblong nucleus, and express the transmembrane phosphoglycoprotein CD34 [2]. These cells lack the ultrastructural features of endothelial differentiation yet appear to channel the flow of interstitial fluid [3]. Endothelial cell membranebound integrins can act upon collagen fibrils in the adjacent (perivascular) extracellular matrix, exposing glycosaminoglycans (GAGs) to take up water and thereby lower the interstitial pressure. The interstitial gel phase is largely composed of coiled and twisted proteoglycan filaments, barely visible on electron microscopy but holding 99% of the interstitial water in association with glycosaminoglycans, mostly hyaluronan. The other 1% of interstitial water is within a gel-free phase through which water can flow alongside collagen fibre bundles with their associated CD34+ interstitial cells. This space appears microscopically as fluid vesicles and rivulets. The proportion of the gel-free water phase is increased in interstitial oedema, and in the most severe cases up to 50%. Interstitial gel-free solvent and solutes are drawn into collecting lymphatics in order to complete the circulation of interstitial fluid.

Chapter 5

114

Interstitial matrices are open and loose or compact and rigid. The primitive interstitium forms a loose open network of molecules that facilitate cell division, cell movement, and cell sorting. In disease states the first stages of repair are the creation of a primitive open and loose matrix which allows cells to enter and repair tissue damage. The open and loosely organized matrix is called provisional matrix [4]. .

The component molecules of the interstitium interact by entanglement, cross-linking, and charge-dependent interactions to form bioactive polymers. These polymers determine the biomechanical properties of each tissue and interact with cells to affect cell phenotype. Matrices that are soft and compliant predominantly feature proteoglycans and associated hyaluronan, while matrices that are stiff and rigid contain more collagens and other fibrous proteins. The abundance of various molecules characterises the different tissue types and results in mechanical and chemical properties of each tissue environment.

Hyaluronan Hyaluronan restricts the movement of water and forms a diffusion barrier that regulates transport of substances through intercellular spaces. Hyaluronan takes part in the partitioning of plasma proteins between vascular and extravascular spaces, and creates the excluded volume phenomenon that affects solubility of macromolecules in the interstitium, changes chemical equilibria, and stabilizes the structure of collagen fibres. Other functions include matrix interactions with hyaluronan binding proteins such as: x x x x x x

hyaluronectin, glial hyaluronan binding protein, brain enriched hyaluronan binding protein, collagen VI, tumor necrosis factor-stimulated gene 6 protein (TSG-6), inter-alpha-trypsin inhibitor.

Cell surface interactions involving hyaluronan include its coupling with CD44 as seen in the endothelial glycocalyx, and with RHAMM (Hyaluronan-mediated motility receptor), which has been implicated in developmental processes, tumour metastasis, and pathological reparative processes. Fibroblasts, mesothelial cells, and certain types

Interstitium and Lymphatic Systems

115

of stem cells surround themselves in a pericellular "coat", part of which is constructed from hyaluronan, in order to shield themselves from bacteria, red blood cells, or other matrix molecules. For example, with regards to stem cells, hyaluronan, along with chondroitin sulphate, helps to form the stem cell niche. Stem cells are protected from the effects of growth factors by a shield of hyaluronan and minimally sulphated chondroitin sulphate. During progenitor division, the daughter cell moves outside of this pericellular shield where it can then be influenced by growth factors to differentiate even further. Interstitial hyaluronan washout when lymph flow is raised during systemic inflammation could well contribute to elevated plasma hyaluronan concentrations [5]. Collagen fibrils provide tensile strength to compact and rigid interstitium, while in skin and large arteries elastin is also present to confer elasticity. Glycosaminoglycans (GAGs) are contained within interfibrillar spaces. Sulphated GAGs anchor to a linear core protein to form brush-shaped proteoglycans. The sulphated GAGs include chondroitin, keratan, dermatan and heparan. Syndecans are heparan sulphate proteoglycans attached to the cell membrane. Proteoglycans are strung along the long non-sulphated GAG hyaluronic acid string, which at around 5 microns in length is very much longer than the sulphated GAGs. This arrangement has been called the string of pearls model. All GAGs have fixed negative charges in the form of carboxyl groups and can bind cations, most notably sodium. Finally, the interstitium also contains glycoproteins, many of which are cell membrane components, such as fibronectin.

The perivascular matrix The subendothelial basement membrane contains collagen type IV isoforms, laminin isoforms, heparan sulphate proteoglycans (perlecan or agrin), and nidogen-1 and/ or nidogen-2. Collagen type IV confers structural stability while laminins are biologically active. Collagen and laminins are connected by perlecans and nidogens. Laminins transduce signals that control cell migration, survival, proliferation, and differentiation. Laminin Ƚ-chain distribution and expression are different between organs and capillary types. Laminin isoforms bind to growth factors (VEGF, platelet-derived growth factor [PDGF], fibroblast

116

Chapter 5

growth factor [FGF], and epidermal growth factor [EGF]) with high affinity, through their heparin-binding domains located in the Ƚ chain laminin-type G domain. In addition, basement membrane-bound proteoglycans and glycosaminoglycans such as hyaluronan, chondroitin sulphate, and heparan sulphate, with terminal sialic acids, contribute to binding and sequestering of growth factors and morphogens with their surface receptors. A fibrillar interstitial matrix underlies the endothelial basement membrane and serves to further interconnect the endothelium with its neighbouring tissue. In general, it is composed largely of the fibrillar collagen types I and III together with chondroitin sulphate and dermatan sulphate proteoglycans such as decorin and biglycan, and again multi-adhesive glycoproteins, although its composition may change from one capillary bed to another. The composition of this latter layer is primarily controlled by resident tissue fibroblasts, that act as a signalling hub for tissue homeostasis and repair. Like the endothelium, they have an organ-specific phenotype where anatomical localization is related to the specific transcriptome profiles, such as embryonic patterning genes, as well as function. While these cells intimately communicate with tissue cells and the immune system to preserve an anti-inflammatory and trophic milieu, they can respond to injury by differentiating into proliferating myofibroblasts, converting to nonproliferating, matrix-producing phenotype. Where the role of the latter in scar formation is undisputed, the way by which tissue fibroblasts use and regulate their extracellular microenvironment for homeostasis is not well understood.

Interstitial water Interstitial water and solutes of the gel phase occupy the spaces within the proteoglycan/ hyaluronan matrix. The effective radius of these spaces, known as their hydraulic radius, is as small as 3 nm in cartilage to 300 nm in the vitreous body of the eye. The hydraulic radius of a matrix determines its resistance to the flow of solvent and solutes through that part of the interstitium. The Wharton’s Jelly of the umbilical cord is an open and loosely organised matrix with a hydraulic radius of about 30 nm and is a good example of the gel nature of interstitium. The interstitial gel restricts water mobility and so stabilises tissue shape. It also prevents interstitial fluid displacement

Interstitium and Lymphatic Systems

117

by gravity and slows the spread of organisms such as bacteria.

Gel swelling pressure GAGs attract water molecules and confer the ability of the interstitial gel phase to swell by taking up water. The gel swelling pressure is defined as the subatmospheric pressure which precisely balances the suction effect of the interstitial GAGs, and is an osmotic pressure largely due to sodium ions attracted by the fixed negative charges - the GibbsDonnan effect. As Arthur Guyton discovered during his classic investigations of the Starling forces, many tissues maintain a subatmospheric interstitial fluid pressure in health and so the GAGs are normally under-saturated with water. The lymphatic system maintains this state of under-saturation by pumping fluid away from the gel phase of the interstitium into the aqueous lymph. Reduction of the pumping capacity of the lymphatic system therefore predisposes to fluid retention and oedema. In collagen-rich tissues such as skin the swelling tendency of the interstitial matrix is further counteracted by tissue fibroblasts which tension the collagen fibrils under the regulation of collagen-binding integrins at the cell membrane contact points. The tension of collagen fibres restricts the swelling of GAGs. Collagenbinding integrins only have a limited role in adult connective tissue homeostasis because of the relative paucity of cell-binding sites in the mature fibrillar collagen matrices. Their importance may be greater in connective tissue remodelling, such as wound healing [6].

Colloid osmotic pressure of the interstitium Interstitial fluid can be harvested from nylon wicks implanted subcutaneously, for instance in arm and leg. In one study of anaesthetised children the mean plasma colloid osmotic pressure was 26 mmHg while the interstitial colloid osmotic pressure was about 14 mmHg [7].

Albumin exclusion and aqueous phase viscosity Albumin is largely excluded from the interstitial gel phase and the collagen phase but is present in the free flowing interstitial aqueous

118

Chapter 5

phase and in the lymph within lymphatic vessels [8, 9]. Water and small solutes (Na+, Cl-, and urea) move easily between aqueous and gel phases, and between intracellular fluid and extracellular fluid according to prevailing osmotic, hydrostatic, and electrochemical forces. The presence of proteins will affect viscosity to the flow and accumulation of water in hypoproteinaemic oedematous conditions [10]. As interstitial fluid accumulates and the aqueous phase expands relative to the gel phase, this mechanism becomes increasingly relevant.

Interstitial fluid pressure Interstitial fluid pressure varies from tissue to tissue and with the rate of fluid exchange. Integrin activation and subsequent conformational changes to collagen allow the GAGs to become hydrated. This brings about an acute reduction in interstitial fluid pressure in inflammatory conditions, increasing the transendothelial pressure difference and thereby increasing Jɋ by as much as 20-fold independently of other causes of capillary leak. Water absorption from the gut lumen is associated with increased mucosal interstitial fluid pressure which promotes water transfer to the plasma by fenestrated mucosal capillaries. Fluid secretion, for instance by endocrine and salivary glands, reduces their interstitial pressure and so increases transendothelial filtration to supply the water needed to continue the secretion. As noted above, the matrix compressive effect of fibroblasts via collagen-binding integrins has only limited effect on the regulation of interstitial fluid pressure. In a rat thermal injury model acutely reduced interstitial fluid pressure was shown to be a major factor in oedema formation [11]. High-dose Vitamin C attenuated the pressure drop and reduced post-burn oedema.

Anatomic features Researchers in the United States have used probe-based confocal laser endomicroscopy (pCLE), an in vivo imaging technology that provides real-time histologic assessment of tissue structures in patients. They

Interstitium and Lymphatic Systems

119

looked at gastrointestinal tract and urinary bladder submucosae, the dermis, peri-bronchial and peri-arterial soft tissues and fascia. They emphasise the existence of an interstitial (i.e. prelymphatic) space defined by a complex lattice of thick collagen bundles that are intermittently lined on one side by fibroblast-like cells that stain with endothelial markers in the tissues they examined [3].

Lymphatic vascular system The lymphatic system pumps fluid from tissues and returns it to blood vessels. Lymphatics also transport lymphocytes and dendritic cells to the lymphoid organs. The lymphatic system vasculature consists of thin-walled capillaries and larger vessels that are lined by endothelial cells. There are unique lymphatic markers that differentiate lymph vessels from blood vessels. These include Prox1, a transcription factor required for programming the phenotype of the lymphatic endothelial cell, and LYVE-1, a CD44 homologue. Vascular endothelial growth factor receptor 3 is a receptor for vascular endothelial growth factors (VEGF) C and D, and is not detected on blood vascular endothelial cells. VEGF-C and VEGF-D regulate lymphangiogenesis by activating VEGFR3, a cell-surface tyrosine kinase receptor, leading to initiation of a downstream signalling cascade. During human gestation, lymph sacs appear at 6–7 weeks. They appear to sprout from embryonic veins. VEGFR-3 deletion in mice leads to defects in blood vessel remodelling and embryonic death [12]. The afferent lymphatic vessels are of two types, initial and collecting. They differ anatomically (i.e. the presence or absence of surrounding smooth muscle cells and semilunar lymphatic valves), in their expression pattern of adhesion molecules and in their permissiveness to fluid and cell entry [13]. A lymphangion is defined as the functional unit of a lymph vessel that lies between two lymphatic valves [14]. Efferent lymphatic vessels conduct lymph away from lymph nodes, to further lymph nodes or to the lymphatic trunks. They also feature semilunar valves to ensure one-way flow and an investment of smooth muscle to pump the contained fluid. The right and left lumbar trunks and the intestinal trunk constitute the cisterna chyli. The left lymphatic duct, more often called the thoracic duct when seen in the chest, originates on the 12th thoracic vertebra from the confluence of right and left lumbar trunks, then traverses the diaphragm at the aortic

120

Chapter 5

aperture and ascends the superior and posterior mediastinum between the descending thoracic aorta and the azygos vein. The left lymphatic duct averages about 5 mm diameter as it passes behind the left carotid artery and left internal jugular vein at the 5th thoracic vertebral level and drains into the venous angle of the left subclavian and internal jugular veins. There are two valves at the junction of the duct with the left subclavian vein that prevent the flow of venous blood into the duct when central venous pressure exceeds thoracic duct lymph pressure. Efferent lymph from the right thorax, right arm, head, and neck is conducted by the smaller right lymphatic duct. Recently a non-muscular lymphatic endothelial vessel network in the dura mater of the mouse brain was discovered by researchers in Helsinki [15]. The dural lymphatic vessels absorb cerebrospinal fluid from the adjacent subarachnoid space and brain interstitial fluid via the glymphatic system described by Iliff [16]. Dural lymphatic vessels conduct fluid into deep cervical lymph nodes via foramina at the base of the skull, where solvent and small solutes can be absorbed to lymph node venules while efferent lymph flows to the right thoracic duct.

Interstitial fluid dynamics Intrinsic lymphatic pumping is regulated by four major factors: preload, afterload, spontaneous contraction frequency and contractility. The similarity to the Frank–Starling relationship for the heart is obvious. •

•

• •

Preload is the end-diastolic pressure (or volume) within the valved muscular lymphangion. Increasing the ‘filling pressure’ over a physiologic range increases the amplitude of contraction and so enhances pump output. Afterload; The lymphatic pump must adapt to elevated outflow pressures resulting from partial outflow obstruction, increased central venous pressure and/or gravitational shifts. Lymphangions in series can propel lymph against higher pressures than individual lymphangions. Contraction frequency of collecting lymphatics is exquisitely sensitive to pressure, and changes as small as 0.5 cm H2O can double the contraction frequency. Contractility is often used in the lymphatic context to describe

Interstitium and Lymphatic Systems

121

the enhancement of amplitude or frequency of contraction in response to a pressure increase or agonist activation. The cardiac parallel is the concept of inotropy and inotropic agents. There are of course extrinsic pump mechanisms operative in vivo. Leg muscles, for example, contribute significantly to the energy expended on pumping lymph to the inguinal, femoral and iliac lymph nodes. Lymph flow in the thoracic duct is supported by the cycle of breathing. The thoracic duct smooth muscle is capable of contracting with sufficient force to propel lymph toward the jugular venous junction at 1-3 ml min-1 which is just about sufficient to move the normal daily efferent lymph volume of around 4 litres [17]. Lymphatic muscle contractions, like cardiac muscle contractions, can occur spontaneously, but in health they are subject to neural modulation. Sympathetic adrenergic nerve fibres appear to be the dominant neural innervation of the lymphatic vasculature. Ƚadrenergic stimulation of contractile lymphatic vessels consistently increases tone, amplitude and frequency, while Ⱦ-adrenergic receptor activation decreases them. Substance P, commonly associated with afferent nerve endings, augments tone and increases frequency. Muscarinic receptors promote an increase in frequency, but the inhibitory effect of endothelial nitric oxide synthase (eNOS) activation seems to be predominant. Mu receptor agonists like endorphins and morphine reduce the spontaneous contractility of smooth muscle everywhere. Serotonin (5-HT) can either inhibit or increase spontaneous lymphatic contractions depending on the species and the state of serotonin receptor expression. Other inhibitory factors include vasoactive intestinal peptide and calcitonin gene related peptide. Contraction synchrony within a lymphangion generates a systolic pressure pulse that can open the outflow valve and eject lymph. Lymphatic contractions are triggered by an action potential achieved in a pacemaker lymphatic microvascular cell, and the action potential propagates rapidly from cell to cell over the length of the lymphangion. Electrical coupling between the cells is presumably through connexins that form intercellular gap junctions. Application of gap-junction blockers in mesenteric lymphatic vessel segments leads to uncoordinated contractions. Valve function is critical. Collecting lymphatics contain bicuspid (semilunar) valves whose leaflets extend from a ring-shaped base and

122

Chapter 5

insert into the vessel wall. The valve opening is a tapered funnel. A dilated sinus downstream from the valve facilitates valve opening and partially balances the high resistance of the narrow orifice created by the valve leaflets. Valves are spaced at semi-regular intervals, and the factors that control their spacing are not known. Barrier function of lymphatic vessels was once disregarded, presuming they were impermeable to fluid and solute. More recent analyses of collecting lymphatic endothelial junction proteins reveal no major differences from those of blood vessels. Collecting lymphatics are not only permeable to solute and fluid, their albumin permeability is comparable to that of postcapillary venules. Like venules, lymphatic permeability is actively regulated because it can be modified by several signalling pathways, including nitric oxide. Lymphatic capillaries are an order of magnitude more permeable than collecting lymphatic microvessels, most likely due to their discontinuous pattern of junctional adhesion proteins, facilitating fluid and solute absorption from the interstitium. Lymphatic contractile dysfunction is often contingent with inflammatory states such as trauma, sepsis, burns, and even major surgery. It is a likely contributor to the accumulation of interstitial fluid or oedema seen in these conditions.

Lymphatics and the interstitial storage of sodium It has long been taught that body sodium content directly determines the extracellular fluid volume and therefore the effective circulating fluid volume. Long-term blood pressure regulation, it was taught, relies on renal mechanisms to retain or excrete sodium in order to keep the effective circulating fluid volume within very narrow margins of equilibrium. Clinicians therefore use isotonic salt solutions to resuscitate patients with reduced effective circulating fluid volume (hypovolaemia) and are cautioned that excessive sodium administration must cause oedema. Recent investigations in humans confirm animal laboratory evidence that some sodium is in fact stored within the body without commensurate water retention [18–20]. This phenomenon was observed with salt solution infusions in surgical patients as long ago as 1986 [21]. The sodium store is now shown to be an interstitial reservoir that buffers the free extracellular sodium and is regulated by

Interstitium and Lymphatic Systems

123

extrarenal, tissue-specific mechanisms for the release and storage of sodium. Immune cells from the mononuclear phagocyte system, including macrophages and dendritic cells, are the local sensors of interstitial electrolyte concentration. The major anatomic site of this sodium regulation is the interstitium of the skin, with its substantial volume of interstitial fluid and lymphatic vasculature forming a vessel network that can be expanded or reduced according to long term sodium intake. Skin macrophages and lymphatics are now known to act in concert as systemic regulators of body fluid volume and long-term blood pressure. Interstitial electrolyte concentrations are higher than in blood, and macrophages regulate local interstitial electrolyte composition via a tonicity-responsive enhancer–binding protein which induces vascular endothelial growth factor (VEGF-C) production as tonicity rises. Acting on VEGF Receptor 3, VEGF-C stimulates lymphangiogenesis to extend the capillary network and enhance the capacity for interstitial fluid clearance. At the same time VEGF-C stimulates VEGF Receptor 2 on blood capillaries promoting endogenous nitric oxide synthesis and so increasing local blood flow. Free sodium ions are thus presented via the blood stream for renal excretion and the extracellular space is protected from major sodiuminduced fluid volume fluctuations. A recent study has shown that fluid leaving the skin as lymph is isosmotic to plasma, even after a high sodium intake, but raises the possibility that the skin can differentially control its own electrolyte microenvironment by creating local gradients that may be functionally important [22]. To investigate the effects of sodium intake on the endothelial surface layer, twelve healthy male volunteers were randomised to low sodium (less than 50 mmol per day) or high sodium (more than 200 mmol per day) diets for eight days. There was no measurable effect on arterial pressure, perfused boundary region (endothelial surface layer thickness) or glycosaminoglycan excretion. Body weight increased by around 2.5 kg with high salt intake, suggesting an extracellular volume expansion. Plasma volume measured by the central volume of distribution of radiolabelled albumin was unaffected. Subjects who had followed a low sodium diet were then given 540 ml of 2.4% (hypertonic) sodium chloride as an acute sodium load. This challenge increased the volume of distribution of albumin by 250 ml and increased the transcapillary escape rate of albumin from 7 to 10 % per

124

Chapter 5

hour. There was no acute effect on arterial pressure or perfused boundary region. The authors’ interpretation of their data was that acute intravenous sodium loading was associated with increased microvascular permeability, suggesting functional damage to the endothelial surface layer [23]. In my opinion there are other plausible interpretations, including a natriuretic peptide effect. In the same experiment plasma sodium concentration at the end of hypertonic saline infusion was as predicted by standard sodium kinetics, but four hours later had decreased by 1.8 mmol l-1 against a predicted fall of less than 1 mmol l-1. The authors therefore concluded that healthy individuals are able to osmotically inactivate significant amounts of sodium after hypertonic saline infusion [24].

Interstitial fluid in critical illness William Sibbald’s laboratory in London, Ontario delivered some remarkable insights into systemic inflammatory effects on fluid physiology during the 1980s. Their experimental model used mature, awake sheep with observations over several days. An early report was on the effects of infusing zymosan-activated plasma, which is known to contain the complement-derived neutrophil activator, C5a. It therefore promotes sterile inflammation. Pulmonary artery pressure and lung lymph flow increased immediately after the infusion was commenced, and it was remarked that lymph to plasma total protein ratios remained essentially unchanged. If we were to presume that lymph is simply a plasma filtrate, we might deduce that the increased transendothelial filtration rate of solvent Jɋ was paralleled by a change in reflection coefficients and permeability to proteins. However, we now appreciate that the composition of lymph is greatly influenced by processes within the interstitium [25]. The polymorphonuclear leukocyte count in arterial blood fell immediately because of sequestration within the pulmonary microvasculature, and parallels with the likely pathophysiology of acute lung injury in systemic inflammatory conditions were drawn [26]. In a subsequent experiment bacterial septicaemia caused a very similar picture of pulmonary hypertension, increased lung lymph flow and neutrophil extravasation into the pulmonary parenchyma [27]. Dobutamine infusion increased cardiac output and slightly reduced the lymph protein concentration, but did not increase the pulmonary lymph flow rate [28]. Eicosanoid-based treatments with ibuprofen [29] or with Prostaglandin E1 [30]

Interstitium and Lymphatic Systems

125

suggested some therapeutic potential for reducing sepsis-induced pulmonary transendothelial filtration and lymph flow. Sibbald’s team also studied the effects of systemic sepsis on peripheral microcirculatory fluid exchange by examining changes in flow (Qlymph) and lymph-to-plasma [L/P] total protein and albumin ratios from lymph draining the efferent duct of a prefemoral lymph node in sheep, before and during surgically-induced peritonitis. After baseline study, peritonitis was produced by caecal ligation, perforation, and devascularisation. By 24 hours blood cultures revealed a polymicrobial bacteraemia. The haemodynamic response to the septic insult during the 72-hour study period was characterized by an increase in heart rate and an initial fall in stroke volume index; yet, the mean blood pressure remained unchanged from baseline levels throughout the study protocol. The intrapulmonary shunt fraction increased by 48 hours, as did both the Qlymph (2.6 +/- 1.9 ml/h to 6.8 +/- 4.6 ml/h) and the calculated lymph albumin clearance (1.6 +/- 1.2 ml/h to 3.1 +/- 1.7 ml/h). Although the calculated serum to interstitial colloid osmotic pressure gradient fell, both the [L/P] total protein and albumin ratios were unchanged from baseline throughout 72 hours of study. Further, [L/P] total protein ratios were unrelated to Qlymph; as Qlymph (experimental/baseline) increased with sepsis, [L/P] total protein ratio (experimental/baseline) did not fall. The conclusion was that systemic sepsis results in increased transendothelial solvent filtration that is primarily a consequence of an increase in microvascular permeability [31]. Robert Demling in Boston made important contributions to the pathophysiology of oedema. We see in his work an appreciation of acute disequilibrium of Starling forces moving to a steady-state. His laboratory demonstrated that a marked increase in fluid flux after sustained protein depletion is unrelated to colloid osmotic pressure. They drew attention to the possible contribution of decreasing viscosity of the interstitial matrix leading to a more rapid interstitial fluid accumulation [10] [32]. The surgical research team in Denver Colorado developed a hypothesis that mesenteric ischaemic/reperfusion primes polymorphonuclear leucocytes which can then be provoked, for example by endotoxin, to cause distant organ injury by migrating across the endothelium cell and releasing reactive oxygen species [33]. A variant of this hypothesis is

126

Chapter 5

that the shock-injured gut releases biologically active factors into mesenteric lymph, and these factors activate circulating neutrophils to injure distant endothelial cells: the gut-lymph hypothesis [34, 35].

Peritoneal dialysis, peritoneal resuscitation Peritoneal dialysis clears both proteins and smaller molecules. Typically, the daily clearance of proteins in chronic ambulatory peritoneal dialysis patients is 5 g, of which around 4 g is albumin and does not cause hypoproteinaemia [36]. While the clearance of smaller solutes is fairly constant, the more variable protein clearance is attributed to peritoneal vascular surface area and size-selective membrane permeability. Chronic peritoneal dialysis patients often have raised plasma and dialysate levels of hyaluronan. The levels are further increased during episodes of peritonitis, and the concentration of hyaluronan in the dialysate effluent has been presumed to be a surrogate marker for peritoneal inflammation [37]. It seems likely that frequent exposure to dialysis fluids brings about peritoneal inflammation, leading to interstitial matrix remodelling and angiogenesis and peritoneal fibrosis that ultimately leads to dysfunction of the peritoneum as a dialysis membrane [38]. Peritoneal resuscitation is an interesting concept. Following haemorrhagic shock mesenteric oedema is seen and mesenteric lymph flow increases. Reducing post-shock mesenteric oedema and mesenteric lymph flow could have therapeutic advantages. Adjunctive direct peritoneal resuscitation with dialysis fluid, or even with normal saline, have been shown in rat experiments to achieve these objectives [39]. Intravenous hypertonic saline resuscitation is another way to minimise intestinal oedema [40].

References 1. Witjas FMR, van den Berg BM, van den Berg CW, Engelse MA, Rabelink TJ. Concise Review: The Endothelial Cell Extracellular Matrix Regulates Tissue Homeostasis and Repair. Stem Cells Transl Med. 2018

Interstitium and Lymphatic Systems

127

2. Sidney LE, Branch MJ, Dunphy SE, Dua HS, Hopkinson A. Concise review: evidence for CD34 as a common marker for diverse progenitors. Stem Cells. 2014;32:1380-1389. 3. Benias PC, Wells RG, Sackey-Aboagye B et al. Structure and Distribution of an Unrecognized Interstitium in Human Tissues. Sci Rep. 2018;8:4947. 4. Wight TN. Provisional matrix: A role for versican and hyaluronan. Matrix Biol. 2017;60-61:38-56. 5. Berg S, Engman A, Hesselvik JF, Laurent TC. Crystalloid infusion increases plasma hyaluronan. Crit Care Med. 1994;22:1563-1567. 6. Zeltz C, Gullberg D. The integrin-collagen connection--a glue for tissue repair. J Cell Sci. 2016;129:653-664. 7. Guthe HJ, Indrebø M, Nedrebø T, Norgård G, Wiig H, Berg A. Interstitial fluid colloid osmotic pressure in healthy children. PLoS One. 2015;10:e0122779. 8. Reed RK, Lepsøe S, Wiig H. Interstitial exclusion of albumin in rat dermis and subcutis in over- and dehydration. Am J Physiol. 1989;257:H1819-27. 9. Wiig H, Reed RK, Tenstad O. Interstitial fluid pressure, composition of interstitium, and interstitial exclusion of albumin in hypothyroid rats. Am J Physiol Heart Circ Physiol. 2000;278:H1627-39. 10. Demling RH, Harms B, Kramer G, Gunther R. Acute versus sustained hypoproteinemia and posttraumatic pulmonary edema. Surgery. 1982;92:79-86. 11. Tanaka H, Lund T, Wiig H et al. High dose vitamin C counteracts the negative interstitial fluid hydrostatic pressure and early edema generation in thermally injured rats. Burns. 1999;25:569-574. 12. El-Chemaly S, Levine SJ, Moss J. Lymphatics in lung disease. Ann N Y Acad Sci. 2008;1131:195-202. 13. Baluk P, Fuxe J, Hashizume H et al. Functionally specialized junctions between endothelial cells of lymphatic vessels. J Exp Med. 2007;204:2349-2362. 14. van Helden DF. The lymphangion: a not so ‘primitive’ heart. J Physiol. 2014;592:5353-5354. 15. Aspelund A, Antila S, Proulx ST et al. A dural lymphatic vascular system that drains brain interstitial fluid and macromolecules. J Exp Med. 2015;212:991-999. 16. Iliff JJ, Wang M, Liao Y et al. A paravascular pathway facilitates CSF flow through the brain parenchyma and the clearance of interstitial solutes, including amyloid Ⱦ. Sci Transl Med. 2012;4:147ra111.

128

Chapter 5

17. Scallan JP, Zawieja SD, Castorena-Gonzalez JA, Davis MJ. Lymphatic pumping: mechanics, mechanisms and malfunction. J Physiol. 2016;594:5749-5768. 18. Bhave G, Neilson EG. Body fluid dynamics: back to the future. J Am Soc Nephrol. 2011;22:2166-2181. 19. Wiig H, Swartz MA. Interstitial fluid and lymph formation and transport: physiological regulation and roles in inflammation and cancer. Physiol Rev. 2012;92:1005-1060. 20. Wiig H, Luft FC, Titze JM. The interstitium conducts extrarenal storage of sodium and represents a third compartment essential for extracellular volume and blood pressure homeostasis. Acta Physiol (Oxf). 2018;222 21. Nielsen OM, Engell HC. Changes in extracellular sodium content after elective abdominal vascular surgery. Acta Chir Scand. 1986;152:587-591. 22. Nikpey E, Karlsen TV, Rakova N, Titze JM, Tenstad O, Wiig H. HighSalt Diet Causes Osmotic Gradients and Hyperosmolality in Skin Without Affecting Interstitial Fluid and Lymph. Hypertension. 2017;69:660-668. 23. Rorije NMG, Olde Engberink RHG, Chahid Y et al. Microvascular Permeability after an Acute and Chronic Salt Load in Healthy Subjects: A Randomized Open-label Crossover Intervention Study. Anesthesiology. 2018;128:352-360. 24. Olde Engberink RH, Rorije NM, van den Born BH, Vogt L. Quantification of nonosmotic sodium storage capacity following acute hypertonic saline infusion in healthy individuals. Kidney Int. 2017;91:738-745. 25. Dzieciatkowska M, D’Alessandro A, Moore EE et al. Lymph is not a plasma ultrafiltrate: a proteomic analysis of injured patients. Shock. 2014;42:485-498. 26. Sharkey P, Judges D, Driedger AA, Cheung H, Finley RJ, Sibbald WJ. The effect of an infusion of zymosan-activated plasma on hemodynamic and pulmonary function in sheep. Circ Shock. 1984;12:79-93. 27. Judges D, Sharkey P, Cheung H et al. Pulmonary microvascular fluid flux in a large animal model of sepsis: evidence for increased pulmonary endothelial permeability accompanying surgically induced peritonitis in sheep. Surgery. 1986;99:222-234. 28. Gnidec AG, Finley RR, Sibbald WJ. Effect of dobutamine on lung microvascular fluid flux in sheep with “sepsis syndrome”. Chest. 1988;93:180-186.

Interstitium and Lymphatic Systems

129

29. Gnidec AG, Sibbald WJ, Cheung H, Metz CA. Ibuprofen reduces the progression of permeability edema in an animal model of hyperdynamic sepsis. J Appl Physiol (1985). 1988;65:1024-1032. 30. Sibbald WJ, Campbell D, Raper RR, Rutledge FS, Cheung H. The effects of prostaglandin E1 on lung injury complicating hyperdynamic sepsis in sheep. Am Rev Respir Dis. 1989;139:674681. 31. Avila A, Warshawski F, Sibbald W, Finley R, Wells G, Holliday R. Peripheral lymph flow in sheep with bacterial peritonitis: evidence for increased peripheral microvascular permeability accompanying systemic sepsis. Surgery. 1985;97:685-695. 32. Demling RH, Kramer GC, Gunther R, Nerlich M. Effect of nonprotein colloid on postburn edema formation in soft tissues and lung. Surgery. 1984;95:593-602. 33. Moore EE, Moore FA, Franciose RJ, Kim FJ, Biffl WL, Banerjee A. The postischemic gut serves as a priming bed for circulating neutrophils that provoke multiple organ failure. J Trauma. 1994;37:881-887. 34. Senthil M, Brown M, Xu DZ, Lu Q, Feketeova E, Deitch EA. Gut-lymph hypothesis of systemic inflammatory response syndrome/multipleorgan dysfunction syndrome: validating studies in a porcine model. J Trauma. 2006;60:958-65; discussion 965. 35. Deitch EA, Xu D, Kaise VL. Role of the gut in the development of injury- and shock induced SIRS and MODS: the gut-lymph hypothesis, a review. Front Biosci. 2006;11:520-528. 36. Balafa O, Halbesma N, Struijk DG, Dekker FW, Krediet RT. Peritoneal albumin and protein losses do not predict outcome in peritoneal dialysis patients. Clin J Am Soc Nephrol. 2011;6:561-566. 37. Yung S, Chan TM. Hyaluronan--regulator and initiator of peritoneal inflammation and remodeling. Int J Artif Organs. 2007;30:477-483. 38. Rosengren BI, Sagstad SJ, Karlsen TV, Wiig H. Isolation of interstitial fluid and demonstration of local proinflammatory cytokine production and increased absorptive gradient in chronic peritoneal dialysis. Am J Physiol Renal Physiol. 2013;304:F198-206. 39. Matheson PJ, Mays CJ, Hurt RT, Zakaria ER, Richardson JD, Garrison RN. Modulation of mesenteric lymph flow and composition by direct peritoneal resuscitation from hemorrhagic shock. Arch Surg. 2009;144:625-634. 40. Radhakrishnan RS, Shah SK, Lance SH et al. Hypertonic saline alters hydraulic conductivity and up-regulates mucosal/submucosal aquaporin 4 in resuscitation-induced intestinal edema. Crit Care Med. 2009;37:2946-2952.

CHAPTER 6 SYSTEMIC CIRCULATION: THE PERIPHERAL VASCULAR LOOPS

Chapter summary In this Chapter, we examine the integrated glycocalyx - endothelial cell - interstitium - lymphatic arrangements of some of the systemic peripheral vascular loops. Pathologic fluid accumulation as oedema and effusions is an important feature of the Organ System Failures and Dysfunctions.

Systemic circulation The systemic circulation features several peripheral vascular loops in parallel, each with its own general anatomy of arteries and veins, and with many individual variations. Each vascular loop includes a dynamic microcirculation of several billion capillary bed units continually fluctuating between their perfused (flow) and collapsed (no flow) state under the influence of systemic and local factors. For a comprehensive understanding of fluid physiology, we must include in our consideration of the microcirculation the associated interstitium and lymphatic system. While organ blood flow can be measured as the average net perfusion of these units, it is important to remember that dysregulation of the perfusion of individual units can still be pathological.

Cerebral circulation Capillaries of the blood-brain barrier. Cerebral capillaries are a very special subset of the non-fenestrated, continuous basement membrane type capillary.

Systemic circulation

• • • •

131

They lack aquaporins. Their endothelial cell membranes are tightly opposed by zona occludens tight junctions with few breaks, resulting in very small effective pore size of barely 1 nm. They feature a pericapillary cellular investment, as noted by Bennett (“Class Beta”). The pericapillary cellular investment is of astrocyte end feet (glial cells) and pericytes. The pericyte-endothelial cell combination that forms the blood brain barrier acts as a tight epithelium, actively transporting small ions between plasma and brain interstitial fluid and CSF and being mainly responsible for the fluid distribution between blood and brain ISF. The absence of water channels through the tight endothelial barrier of the brain is in contrast to the abundance of aquaporin-4 (AQP4) in the astrocytic foot processes where they are in close contact with the capillaries; the perivascular pool of AQP4. There is evidence that astrocytic AQP4 is an important determinant of brain water homeostasis.

The interplay between endothelial cells, glial cells and pericytes in neuroinflammation affects vascular permeability and appears to be of pathophysiological significance in traumatic brain injury, epilepsy, and neurodegenerative disorders [1, 2]. The hydroxyethyl starches have been found to modulate neuroinflammation and there was hope they could have a positive therapeutic role in ‘cerebral resuscitation’, but initial trials have not suggested a benefit [3]. Orosomucoid is a glycoprotein that contributes to permeability of the blood brain barrier [4, 5] and so has some therapeutic potential. Capillaries of the blood-cerebrospinal fluid barrier There is a second barrier within the central nervous system, the bloodcerebrospinal fluid barrier of the choroid plexus. The periventricular organs that lack the blood–brain barrier seem to be critical for changes in cerebrospinal fluid (CSF) and also in the systemic extracellular fluid sodium concentration. Choroid plexus is anatomically located in the inferior horn of the lateral ventricle, and passes into the interventricular foramen to the third ventricle. There is choroid plexus in the fourth ventricle beneath the cerebellum. The blood-cerebrospinal fluid barrier begins with fenestrated capillaries that filter solvent and solutes to the central stroma of the choroid plexus which is contained by the continuous basement membrane of epithelial cells that have

132

Chapter 6

tight intercellular junctions. Within the central stroma are dendritic cells, fibroblasts and macrophages. Aquaporin-1 is expressed within the cell membrane of the choroid plexus’ epithelial cells and transfers the solvent within the central stroma to the subarachnoid space where it is called cerebrospinal fluid. Solutes and some solvent are transferred to the cerebrospinal fluid by other transcellular routes. The epithelial cells of the choroid plexus contain plenty of mitochondria, a Golgi apparatus and a smooth endoplasmic reticulum, which together confer high synthetic and secretory capacity [6]. Extracellular fluid circulation in the central nervous system The central nervous system’s extracellular fluid circulation does not have a classical lymphatic drainage system. Jeffrey Iliff and colleagues reported that a substantial portion of subarachnoid cerebrospinal fluid cycles through the brain interstitial space as recently as 2012 [7]. Cerebrospinal fluid in the subarachnoid space enters the parenchyma along paravascular spaces that surround penetrating arteries, and brain interstitial fluid is cleared along paravenous drainage pathways. Iliff called this the glymphatic system. The constant immune surveillance of the central nervous system occurs within the meningeal compartment. Researchers in Charlottesville, Virginia, were searching for T cell pathways into and out of the meninges when they discovered functional lymphatic channels lining the dural sinuses [8]. Researchers in Helsinki also identified dural lymphatic channels in mice [9]. These structures are able to carry fluid and cells from the cerebrospinal fluid to the deep cervical lymph nodes. They express the molecular hallmarks of lymphatic endothelial cells. The secretion and composition of the cerebrospinal fluid is tightly regulated by the choroid plexuses, which are complex structures comprised of a plexus of fenestrated capillaries surrounded by a layer of cuboidal epithelial cells, with an intervening stromal space. The epithelial cells are polarised, with the apical cerebrospinal fluid facing side possessing microvilli and tight junctions that constitute the bloodcerebrospinal fluid barrier (BCSFB), which is in many respects analogous in function to the BBB of the cerebrovascular endothelium. Under physiological conditions, cerebrospinal fluid is actively secreted, largely independently of choroidal blood flow, through the concerted activity and involvement of numerous membrane proteins. These

Systemic circulation

133

include apical Na+/K+ ATPase, the aquaporin-1 (AQP1) water channel, and numerous secondary ion transporters and channels localized specifically to basolateral and/or apical Arachnoid granulations, also called arachnoid villi, are projections of the epithelial arachnoid membrane into the venous sagittal sinus and its associated venous lacunae. The Starling forces here favour transcellular movement of cerebrospinal fluid to the blood circulation. Arachnoid granulations were long-thought to be responsible for drainage, but studies on mice have demonstrated many perineural sites for the egress of cerebrospinal fluid. It may be that the cerebral lymphatic system also removes cerebral interstitial solvent and solutes, rather than the traditional arachnoid granulation - venous route alone [10]. The current description of extracellular fluid circulation in the brain, by astroglial-mediated interstitial fluid bulk flow (the glymphatic system), is therefore as follows [11, 12]: •

•

•

• • • •

Cerebrospinal fluid is formed by choroid plexus and leaves the subarachnoid space along the periarterial space of various penetrating arterial branches, driven by pulsation caused by the heartbeat. Aquaporin-4 water channels of astrocytic foot processes that invest the arterial side of the microcirculation facilitate the influx of cerebrospinal fluid from the periarterial glymphatic into the brain parenchyma, where it enters the brain interstitial fluid. Aquaporin-4 water channels of astrocytic foot processes that invest the venous side of the microcirculation facilitate brain interstitial fluid collection into the glymphatic perivenous space of cerebral veins. Lymph vessels accompany the veins within meninges and perivenous glymph is absorbed into them. The lymph vessels conduct their lymph to the cervical lymph nodes. At the cervical nodes, fluid can be absorbed from lymph into the bloodstream, or Efferent lymph fluid from the cervical nodes drains to the superior vena cava.

134

Chapter 6

Notice that the fluid between the endothelium and the astrocytic foot processes comes from the glymphatic system in addition to transendothelial filtrate, and so there is uncoupling of the inverse ‘extravascular dilution’ relationship between ci (or ȫi) and Jv, and prevents ȫi from approaching ȫp. An Exception to the no-absorption rule occurs. The colloid osmotic pressure difference supporting absorption of perivascular fluid to the plasma is protected. This path of fluid return to the circulation is however restricted by the fact that healthy cerebral endothelial cells are tightly opposed by zona occludens tight junctions with few tight junction gaps, and by the absence of aquaporin channels through the endothelial cell membrane. Cerebral oedema Cerebral oedema is the accumulation of extracellular water within the interstitium of the brain and is often accompanied by swelling of brain cells. Cerebral oedema raises intracranial pressure, displaces blood from the incompliant cranial cavity and causes herniation of the brain stem through the foramen magnum. It can therefore be rapidly fatal. Broadly, there are three pathophysiologies, but two or three may be at play in any clinical condition. Vasogenic oedema is due to failure of the capillary permeability barrier, with normally-excluded plasma proteins (and water) gaining access to the interstitial space. Cellular oedema can be caused by plasma hypo-osmolarity (free water excess) but more typically is due to inadequate energy availability to sustain the neuronal cell membrane ion pumps which regulate cell volume, causing cells to swell (cytotoxicity). Interstitial cerebral oedema results from cerebrospinal fluid outflow obstruction, or acute hydrocephaly. Cerebral oedema is visualised and assessed by X-radiologic computerised axial tomography. Oedema appears as low-density areas on an unenhanced scan. The white matter is primarily affected by vasogenic oedema, grey matter by cellular oedema. Magnetic resonance imaging seems to be more sensitive at detecting brain abscess in the cerebritis phase of its development as well as at detecting associated cerebral oedema. Ischaemic or haemorrhagic stroke lead to localised cerebral oedema which is initially cytotoxic because of energetic failure in neuronal cell membranes. Traumatic brain injury and hypoxic brain injury are

Systemic circulation

135

primarily conditions of cellular oedema with subsequent vasogenic oedema. Meningitis may cause acute hydrocephaly as well as vasogenic oedema. Primary and metastatic tumours of the brain are associated with cerebral edema because the blood vessels that develop to supply them (vasogenesis) are not of the blood-brain barrier type: they allow higher rates of transendothelial solvent filtration, and are more permeable to albumin and other proteins. I am sometimes challenged by readers to explain how osmotic diuretics (mannitol, hypertonic saline, urea etc) reduce cerebral oedema, at least acutely, if there is a no-absorption rule. I think the most obvious explanation is that oedema occurs when there is breakdown of the normal blood brain barrier which includes the endothelial glycocalyx layer that confers asymmetry, and therefore the glycocalyx model cannot operate. But it is also interesting to consider that the healthy brain capillaries are potentially an Exception to the no-absorption rule, as explained above. If it were not for their exceptionally tight interendothelial junctions, significant absorption across the cerebral endothelial glycocalyx could be possible. Alcohol intoxication Alcohol impairs blood brain barrier function. In a laboratory experiment sphyngosine-1-phosphate enhanced barrier function and even restored junctions in the presence of alcohol [13].

Coronary circulation The coronary microcirculation is compressed during systole, so myocardial perfusion largely occurs during diastole. As a consequence, adequate diastolic aortic pressure and diastolic time are needed for adequate perfusion. The microvascular endothelial cells of the heart and the pericardium are covered by a continuous glycocalyx. One of the earliest experiments to demonstrate tissue oedema after glycocalyx degradation was performed on the heart [14] and there followed experiments to assess glycocalyx protection by hydrocortisone [15] and sevoflurane [16]. The extensive lymphatic system of the heart runs parallel to the coronary arteries and drains to mediastinal lymph nodes. Myocardial infarction is followed by lymphatic transport dysfunction and myocardial oedema, in spite of remodelling of the cardiac

136

Chapter 6

lymphatic network as a response to endogenous lymphangiogenic factors. Laboratory rat experiments suggest that targeted lymphangiogenic therapy improves both precollector remodelling and capillary lymphangiogenesis, with shorter time to resolution of myocardial oedema and inflammation, and consequent prevention of cardiac fibrosis and dysfunction [17]. It is possible that cardiac lymphatics can be targeted therapeutically to restore lymphatic drainage in the heart to limit myocardial oedema and chronic inflammation [18]. Pericardial effusion Haemorrhage, inflammation, metastatic tumours, and of course changes in the localized Starling Forces within pericardial vasculature all cause pericardial effusion Pericardial effusions are typically subclinical; however, the major concern is the development of cardiac tamponade which may occur if the effusion develops rapidly and overwhelms the capacity of pericardial lymphatics to drain excess fluid. Additionally, the pericardial sac can gradually stretch with slowly developing effusions, thus accommodating larger volumes of fluid without placing pressure on the heart. As an effusion expands, heart sounds may become fainter and the pericardial friction rub associated with acute pericarditis may vanish. The consistency of the effused fluid varies depending on the aetiology. In cases resulting from acute pericarditis the effused fluid can be serous, fibrinous, or haemorrhagic depending on the root cause of the inflammation. Cases due to deranged Starling fluid typically result in a serous fluid.

Hepatic circulations Liver sinusoidal endothelial cells (LSEC) are the most permeable endothelial cells of the mammalian body, and have the highest endocytosis capacity of any human cell [19]. Arterial blood enters the liver via the hepatic artery and interlobar arteries to perfuse the peribiliary plexuses which drain into the hepatic sinusoids. The portal vein carries venous blood from the intestinal mucosa and delivers it via interlobar veins to the hepatic sinusoids. The ratio of hepatic artery to

Systemic circulation

137

portal vein contributions to the hepatic sinusoid blood flow is around 1:3. The sinusoids drain via hepatic veins to the inferior vena cava. Open fenestrations in the hepatic sinusoidal capillaries permit albumin and macromolecules as large as chylomicrons and lipoproteins to pass between plasma and the interstitial fluid of the perisinusoidal (Disse) space, which therefore has the same osmolality as plasma. As small molecules and gases also diffuse freely through the fenestrae, the space of Disse can be considered an extravascular compartment of the plasma volume. The fenestrations are 50-150 microns in diameter and grouped together to form sieve plates which are evident on electron microscopy. Being limited by fibrous capsules, the sinusoidal tissues have little or no compliance to accommodate interstitial fluid expansion. Filtration will be dependent on hydrostatic pressure gradients, as there is no colloid osmotic pressure difference to oppose filtration, and return to the circulation is via the lymphatics. The liver is observed to account for around 50% of the body’s total lymph production, with higher than average protein concentration, and is therefore the major site of transcapillary escape of plasma proteins and probably of other macromolecules when capillary function is unimpaired. In resuscitated hyperdynamic septic shock patients, hepatic blood flow is increased to around 50% of the cardiac output. This redistribution of blood from non-sinusoidal tissue capillary beds to sinusoidal tissues will cause an apparent change in systemic capillary permeability.

Intestinal mucosal circulation The diaphragm fenestrated capillaries of the ileal mucosa are regulated to a lower mean capillary pressure than the non-fenestrated capillaries that perfuse the ileal smooth muscle from the same arterial supply. Their ability to absorb water and nutrients that are presented by feeding is thereby optimised. When fluid is ingested and crosses the intestinal epithelium, the colloid osmotic pressure of the ileal interstitial fluid drops, both by dilution of interstitial albumin and by reduced fractional exclusion of albumin. The protein content of ileal lymph is consequently reduced and the lymph flow rate increased. As a result, the Starling forces across the capillary walls change from favouring a low level of solvent filtration when the gut is not absorbing fluid (no absorption) to a state of sustained solvent uptake when the gut is called upon to absorb ingested fluid.

138

Chapter 6

The absorption of digested food brings a little more complexity because of osmotic gradients set up by the absorbed small solutes. For example, during glucose uptake, the high interstitial glucose concentration causes the effective osmotic pressure to oppose and even exceed the Starling pressures causing solvent absorption. It is known that the epithelial absorption of glucose and amino acids in the ileum is confined to the outer third of the villus, and so it seems likely that fluid absorbed through epithelium here might flow through the villus interstitium to be taken up into the extensive capillary bed of the lower two-thirds of the villi and thence into the lacteals [20].

Interstitial fluid and Lymph Circulation in the Liver and intestinal mucosa The liver has portal, sublobular, and capsular lymphatic vessels. Most of the hepatic lymph drains into the portal lymphatic vessels, while the remainder drains through sublobular and capsular lymphatic vessels. The hepatic lymph primarily comes from the hepatic sinusoids. Fluid Ƥltered out of the sinusoids into the space of Disse ƪows along collagen fibres traversing the limiting plate and connecting the space of Disse with the interstitial space in the portal tracts, or around the sublobular veins. Similar channels traverse the hepatocytes intervening between the space of Disse and the hepatic capsule. From these interstitial spaces, prelymphatic vessels feed the portal, sublobular and capsular lymphatic vessels which take lymph to the cisterna chili and thoracic duct [21]. The intestine has two independent lymphatic networks: • •

A network containing the lacteals drains the villi and the submucosal lymphatic network. A network containing the lymphatics that drain the intestine muscular layer.

These systems deliver their lymph into a common network of collecting lymphatics originating near the mesenteric border [22]. The mesenteric lymphatic system is the subject of research because it may be an amplifier of systemic sepsis, delivering lipid-mediated proinflammatory stimuli and mitochondrial danger-associated molecular patterns (DAMPs) to the circulation [23–25]. Pharmacologic inhibition

Systemic circulation

139

of lymph flow [26] or adjunctive resuscitation by peritoneal dialysis [27] could be protective. Mesenteric lymph is also an important component of the pumped efferent lymph flow in response to haemorrhage [28]. Theories of ascites formation •

•

• •

The underfill theory. The initiating event is portal hypertension; venous congestion and interstitial fluid accumulation cause hypovolaemia which both limits the elevated portal pressure and causes retention of sodium and water. Sodium retention increases the plasma volume and portal pressure, with the subsequent formation of ascites. The overflow theory. Sodium and water retention are the primary events in ascites formation with portal hypertension resulting; plasma volume expansion to the point of overflow from the hepatic sinusoids then causes ascites formation. The vasodilation theory. The initiating event is splanchnic vasodilation leading to sodium and water retention. The lymph imbalance theory. As we have discussed in the context of steady-state Starling physiology, interstitial fluid accumulates when Jv exceeds Qlymph. Diuretic therapy for ascites has been shown to increase lymph flow [29].

Whichever theory is correct, excess interstitial fluid has free access to the Glisson’s capsule on the liver surface. Ascites related to postsinusoidal obstruction, such as cardiac failure or Budd-Chiari syndrome, is protein rich because the sinusoidal capillaries are leaky to large molecules. The ascites of cirrhosis is from mesenteric continuous capillary filtrate and so is protein-poor.

Splenic circulation Spleen is the largest lymphatic structure. The spleen parenchyma features white pulp embedded in red pulp. The red pulp is a reticular connective tissue containing all types of blood cells. Human spleens appear to have a totally open circulation system, as connections from capillaries to sinuses have not been found in the red pulp. Red pulp fibroblasts appear to prevent blood clotting in the open circulation, but how they work is not clear. Splenic arterioles and “sheathed” capillaries

140

Chapter 6

in the red pulp are surrounded by lymphocytes, especially by B cells. Capillary sheaths are composed of endothelial cells, pericytes, special stromal sheath cells, macrophages and B lymphocytes. In the white pulp, T and B lymphocytes form accumulations, the periarteriolar lymphatic sheaths and the follicles, located around the central arteries.

Bone marrow circulation The arrangement of microvessels within bone marrow has proved difficult to ascertain. In a recent study of human iliac crest bone marrow, adipose tissue was found to contain irregularly distributed haematopoietic areas. The haematopoietic areas are served by networks of sinusoidal endothelium, which are loosely connected to the non-sinusoidal capillary beds in areas of pure adipose tissue [30].

Renal circulation Capillaries of the renal glomeruli The renal arteries divide to form low resistance afferent arterioles which supply blood at high pressure (capillary pressure circa 55 mmHg) to the glomerular capillaries. These vessels have a full basement membrane and endothelial glycocalyx layer, but they are generously fenestrated. Anatomically, the fenestrations are as wide as 65 nm, but their effective pore size is only about 15 nm, due to glycocalyx flanking over the open fenestrations. The effective pore size for glomerular filtration beyond the capillary basement membrane is limited to about 6 nm by filtration slit diaphragms at the level of podocyte foot processes. Thus, albumin and larger molecules are normally not filtered into tubular fluid. It has been presumed that albuminuria is an index of capillary permeability. Protein filter function is impaired by hyperglycaemia, and probably by other kidney injuries. Blood leaves the glomerular capillaries via high resistance efferent arterioles that carry blood to capillaries in the renal medulla (the vasa recta) or to an anastomotic arrangement of capillaries in the renal cortex (the peritubular plexus).

Systemic circulation

141

Why does the glomerular capillary pressure have to be so high? After around 20% of plasma water has been filtered away by the glomerular capillary, the protein concentration and colloid osmotic pressure of plasma entering the efferent arterioles is raised. From an afferent plasma colloid osmotic pressure of 27 mmHg, blood leaving glomerular capillaries may reach 32 mmHg. Pressure in the proximal tubules is around 10 to 15 mmHg. To enable filtration, glomerular capillary pressure has to be greater than the sum of proximal tubular pressure and glomerular capillary plasma colloid osmotic pressure. From the figures given here, that sum is 42 - 47 mmHg. From a glomerular capillary pressure of 55 mmHg, the effective filtration pressure at the glomerulus is therefore only 8-13 mmHg. Glomerular filtration is very vulnerable to arterial hypotension. Renal peritubular capillaries and the ascending vasa recta Efferent arteriolar resistance is kept high to support glomerular capillary pressure, but the pressure in downstream capillaries is much lower. The capillaries of the kidney cortex and medulla (peritubular capillaries and the ascending vasa recta), are fenestrated and so capable of fluid absorption. Transendothelial osmotic transport driven by gradients of small solutes occurs in the renal vasa recta, and the expression of AQP1 in the endothelial cell membrane here is of clear physiological significance. Renal Cortex The peritubular capillary Starling forces have been investigated in rat and dog kidneys. During fluid deprivation capillary pressure is low, and the plasma colloid osmotic pressure within the capillary is 20–30% higher than in systemic arterial blood, because the peritubular vessels are fed with plasma concentrated by glomerular filtration. The interstitial colloid osmotic pressure is kept low by the macromolecular impermeability of the fenestrated peritubular capillaries (Staverman’s reflection coefficient for albumin close to 1) and the plentiful production of protein-free interstitial fluid by the tubular epithelium, which flushes the relatively small amount of plasma protein entering the cortical interstitial space into the cortical lymphatics. Rehydration by intravenous isotonic salt solution infusion reduces the cortical peritubular capillary plasma colloid osmotic pressure and increases the capillary pressure, but the net pressure favouring solvent absorption is

142

Chapter 6

maintained by a large increase in interstitial fluid pressure and a further fall in interstitial colloid osmotic pressure. Renal Medulla The renal medulla has no lymphatic vessels, so that interstitial fluid has to be absorbed back to the blood stream. Fluid that is absorbed from the collecting ducts into the renal medullary interstitium must therefore be continuously absorbed into the blood stream by the ascending vasa recta capillaries. The medullary interstitial colloid osmotic pressure is kept very low by dilution of any soluble proteins or other macromolecules present. In addition, capillary pressure has been measured and found to be low. The very high rate of solvent absorption to the blood stream is further supported by relatively high interstitial pressure and plasma colloid osmotic pressure. There are, however, problems in explaining the very low medullary interstitial colloid osmotic pressure. The ascending vasa recta are much more permeable to albumin and other proteins than the almost impermeable cortical peritubular capillaries, and the medulla does not enjoy the benefit of protein removal via lymphatics. There has been a suggestion that prelymphatic channels may take interstitial fluid to the renal cortex, but such channels are yet to be identified. It has been demonstrated that labelled albumin is cleared from the medullary interstitium directly into blood, and it has been calculated that convective flow of large solutes like albumin can account for this efficient clearance. Staverman’s reflection coefficient for albumin in the ascending vasa recta is around 0.67, low enough to permit interstitial albumin molecules to enter the plasma yet high enough to preserve a colloid osmotic pressure difference that supports solvent absorption. As the solvent absorbed to the interstitium from the collecting ducts is essentially protein-free, a very low medullary interstitial colloid osmotic pressure is in fact attainable in the presence of leaky capillaries and the absence of lymphatics. Thus, the subglycocalyx accumulation of proteins that would limit absorption according to the MichelWeinbaum model does not happen and absorption is sustained! Another problem is the vulnerability of the ascending vasa recta to be collapsed by the combination of low capillary pressure and high interstitial pressure. In one experiment these vessels did not collapse until capillary pressure became subatmospheric. Their stability is probably due to tethering of the endothelial cells by fine processes to

Systemic circulation

143

the basal lamina of neighbouring descending vasa recta and tubules.

Secretory glands The basement membrane of these capillaries is continuous, and diaphragmed fenestrations are induced by vascular endothelial growth factors. Their upper pore size is in the range of 6–12 nm. The epithelial cell transport of secretions provides the interstitial supply of fluid with low protein concentration that creates exceptions to the principle of no-absorption.

Lymph nodes There are around 600-700 lymph nodes, grouped in the neck, axillae, groin, thoracic mediastinum, and mesenteries of the GI tract. Lymph nodes are exceptions to the no-absorption rule because their interstitial fluid is continuously replenished by the flow of prenodal lymph that has a low protein concentration. They are the site of absorption of around half of the formed lymphatic fluid to the blood stream. They host immunoprotective cell lines, the T lymphocytes and B lymphocytes. The superficial (e.g. inguinal) nodes are generally supplied by a single artery that enters at the hilum of the node and then radiates outwards to the lymph node pulp. By contrast, the deep mesenteric lymph nodes have several separate arteries penetrating the lymph node capsule, entering the trabeculae and then running towards the centre of the node. This arrangement would seem to be better suited to higher volume flows of blood and lymph, with greater immune processing capacity. High endothelial venules, which are much plumper than other post-capillary venules, feature glycoprotein receptors called lymphocyte homing receptors within their glycocalyx. These receptors control the extravasation of lymphocytes as well as other leukocytes, and help regulate both non-specific and specific immune responses [31].

144

Chapter 6

References 1. Giannoni P, Badaut J, Dargazanli C et al. The pericyte-glia interface at the blood-brain barrier. Clin Sci (Lond). 2018;132:361-374. 2. Woodcock T, Morganti-Kossmann MC. The role of markers of inflammation in traumatic brain injury. Front Neurol. 2013;4:18. 3. Murkin, JM. “Primum non nocere”: the role of hydroxyethyl starch 130/0.4 in cerebral resuscitation. [editorial]. Can J Anaesth 2012;59(12):1089. 4. Yuan W, Li G, Zeng M, Fu BM. Modulation of the blood-brain barrier permeability by plasma glycoprotein orosomucoid. Microvasc Res. 2010;80:148-157. 5. Schött U, Solomon C, Fries D, Bentzer P. The endothelial glycocalyx and its disruption, protection and regeneration: a narrative review. Scand J Trauma Resusc Emerg Med. 2016;24:48. 6. Marques F, Sousa JC. The choroid plexus is modulated by various peripheral stimuli: implications to diseases of the central nervous system. Front Cell Neurosci. 2015;9:136. 7. Iliff JJ, Wang M, Liao Y et al. A paravascular pathway facilitates CSF flow through the brain parenchyma and the clearance of interstitial solutes, including amyloid Ⱦ. Sci Transl Med. 2012;4:147ra111. 8. Louveau A, Smirnov I, Keyes TJ et al. Structural and functional features of central nervous system lymphatic vessels. Nature. 2015;523:337-341. 9. Aspelund A, Antila S, Proulx ST et al. A dural lymphatic vascular system that drains brain interstitial fluid and macromolecules. J Exp Med. 2015;212:991-999. 10. Ma Q, Ineichen BV, Detmar M, Proulx ST. Outflow of cerebrospinal fluid is predominantly through lymphatic vessels and is reduced in aged mice. Nat Commun. 2017;8:1434. 11. Iliff JJ, Goldman SA, Nedergaard M. Implications of the discovery of brain lymphatic pathways. Lancet Neurol. 2015;14:977-979. 12. Tarasoff-Conway JM, Carare RO, Osorio RS et al. Clearance systems in the brain-implications for Alzheimer disease. Nat Rev Neurol. 2015;11:457-470. 13. Alves NG, Yuan SY, Breslin JW. Sphingosine-1-phosphate protects against brain microvascular endothelial junctional protein disorganization and barrier dysfunction caused by alcohol. Microcirculation. 2019;26:e12506. 14. van den Berg BM, Vink H, Spaan JA. The endothelial glycocalyx protects against myocardial edema. Circ Res. 2003;92:592-594.

Systemic circulation

145

15. Chappell D, Jacob M, Hofmann-Kiefer K et al. Hydrocortisone preserves the vascular barrier by protecting the endothelial glycocalyx. Anesthesiology. 2007;107:776-784. 16. Chappell D, Heindl B, Jacob M et al. Sevoflurane reduces leukocyte and platelet adhesion after ischemia-reperfusion by protecting the endothelial glycocalyx. Anesthesiology. 2011;115:483-491. 17. Henri O, Pouehe C, Houssari M et al. Selective Stimulation of Cardiac Lymphangiogenesis Reduces Myocardial Edema and Fibrosis Leading to Improved Cardiac Function Following Myocardial Infarction. Circulation. 2016;133:1484-97; discussion 1497. 18. Brakenhielm E, Alitalo K. Cardiac lymphatics in health and disease. Nat Rev Cardiol. 2019;16:56-68. 19. Poisson J, Lemoinne S, Boulanger C et al. Liver sinusoidal endothelial cells: Physiology and role in liver diseases. J Hepatol. 2017;66:212227. 20. Bernier-Latmani J, Petrova TV. Intestinal lymphatic vasculature: structure, mechanisms and functions. Nat Rev Gastroenterol Hepatol. 2017;14:510-526. 21. Ohtani O, Ohtani Y. Lymph circulation in the liver. Anat Rec (Hoboken). 2008;291:643-652. 22. Miller MJ, McDole JR, Newberry RD. Microanatomy of the intestinal lymphatic system. Ann N Y Acad Sci. 2010;1207 Suppl 1:E21-8. 23. Diebel LN, Liberati DM, Ledgerwood AM, Lucas CE. Systemic not just mesenteric lymph causes acute lung injury following hemorrhagic shock. Surgery. 2008;144:686-93; discussion 693. 24. Diebel LN, Liberati DM, Lucas CE, Ledgerwood AM. Systemic not just mesenteric lymph causes neutrophil priming after hemorrhagic shock. J Trauma. 2009;66:1625-1631. 25. Yi J, Slaughter A, Kotter CV et al. A "clean case" of systemic injury: mesenteric lymph after hemorrhagic shock elicits a sterile inflammatory respose. Shock. 2015;44:336-340. 26. Langness S, Costantini TW, Morishita K, Eliceiri BP, Coimbra R. Modulating the Biologic Activity of Mesenteric Lymph after Traumatic Shock Decreases Systemic Inflammation and End Organ Injury. PLoS One. 2016;11:e0168322. 27. Matheson PJ, Mays CJ, Hurt RT, Zakaria ER, Richardson JD, Garrison RN. Modulation of mesenteric lymph flow and composition by direct peritoneal resuscitation from hemorrhagic shock. Arch Surg. 2009;144:625-634.

146

Chapter 6

28. Boulanger BR, Lloyd SJ, Walker M, Johnston MG. Intrinsic pumping of mesenteric lymphatics is increased after hemorrhage in awake sheep. Circ Shock. 1994;43:95-101. 29. Henriksen JH, Schlichting P. Increased extravasation and lymphatic return rate of albumin during diuretic treatment of ascites in patients with liver cirrhosis. Scand J Clin Lab Invest. 1981;41:589599. 30. Steiniger BS, Stachniss V, Wilhelmi V et al. Three-Dimensional Arrangement of Human Bone Marrow Microvessels Revealed by Immunohistology in Undecalcified Sections. PLoS One. 2016; 11:e0168173. 31. Jalkanen S, Reichert RA, Gallatin WM, Bargatze RF, Weissman IL, Butcher EC. Homing receptors and the control of lymphocyte migration. Immunol Rev. 1986;91:39-60.

CHAPTER 7 PULMONARY CIRCULATION

Chapter summary In this Chapter, we examine the integrated glycocalyx - endothelial cell - interstitium - lymphatic arrangements of the pulmonary circulation. Fluid pathophysiology in this vascular loop manifests as pulmonary oedema and pleural effusion.

The pulmonary vascular loop The pulmonary vascular loop runs in series with the numerous parallel systemic vascular loops and has several characteristics: •

• • •

•

Pulmonary artery pressure is low over a wide range of pulmonary artery blood flow rates as pulmonary microcirculation units are passively recruited or derecruited. Consequently, the calculated pulmonary vascular resistance does not indicate a state of arteriolar tone. Capillary density in the pulmonary parenchyma is higher than in any other tissue, delivering blood to around 250 million alveoli. Smaller pulmonary arteries constrict when the alveolar oxygen content is low, diverting blood to better oxygenated alveoli (hypoxic pulmonary vasoconstriction). Blood cell transit through pulmonary capillaries is relatively prolonged, and the interaction of pulmonary endothelium with circulating cellular elements is favoured. As in other capillary beds, endothelial glycocalyx regulates exposure of endothelial surface adhesion molecules, thereby serving as a barrier to neutrophil adhesion and extravasation. Circulating activated neutrophils are “deprimed” within the pulmonary interstitium as a defence against the amplification of

148

•

•

Chapter 7

systemic inflammation. Staverman’s reflection coefficient sigma for albumin is only about 0.7 in pulmonary microvascular endothelium, suggesting an endothelial surface layer that is effectively thinner than in systemic capillaries. Studies suggest that removal of GAGs from the pulmonary glycocalyx contributes little to in vivo pulmonary endothelial barrier hydraulic conductivity and solute permeability. Glycocalyx is, however, the mechanotransducer of pulmonary vascular pressure, and is through that mechanism involved in the regulation of pulmonary endothelial permeability.

Precapillary and postcapillary anastomoses between bronchial and pulmonary arterial vascular loops may contribute to some physiological observations Pulmonary blood flow During severe exercise pulmonary artery blood flow can increase fourfold or more, with little increase in what is a remarkably low arterial pressure. Pulmonary perfusion is heavily influenced by posture and by the pull of gravity. With increasing blood flow, previously unperfused lung units can become perfused (recruited) with little change in pressure. Oxygen to maintain the integrity of pulmonary tissues comes from the alveoli and conducting airways, and from bronchial arteries which operate at systemic pressure. Bronchial artery-supplied blood to the pleura and hilar tissues drains to azygos veins, thence to the superior vena cava. It can therefore be considered independent of the main pulmonary circulation. Bronchial artery blood serving the capillary units of more peripheral bronchi and bronchioles, around two thirds of the bronchial artery flow, is drained by postcapillary anastamoses to the pulmonary veins. For respiratory physiologists this creates an admixture of venous blood with the arterialized blood from the alveolar capillary networks; the physiological shunt. Precapillary anastomoses from the bronchial arteries to the pulmonary arteries are also observed. Called ‘Sperr arterien’, they seem to have little significance in health, but are able to protect against pulmonary

Pulmonary circulation

149

parenchymal ischaemia, for example pulmonary embolism, by supplementing deficient pulmonary arterial flow.

Pulmonary blood volume After the development of cardiac catheterisation technologies, researchers at Harvard University were able to apply the StewartHamilton method to measure pulmonary blood volume in patients. To avoid the problem of sampling from the left atrium, whose blood is not fully mixed, they measured the pulmonary artery to brachial artery volume, and then subtracted the separately-measured left atrial to brachial artery volume [1]. Just four of their patients had normal haemodynamics, and the pulmonary blood volume averaged 450 ml (250 ml m-2). De Frietas et al. described a single injection technique in 1964 [2]. Indocyanine Green dye is rapidly injected into a central vein and blood for sampling is withdrawn through catheters placed in the pulmonary artery and left atrium by two identical syringes driven by the same motor at a constant speed. In a remarkable experiment the same researchers found that neither sudden increase nor sudden decrease in total blood volume significantly altered the pulmonary blood volume, leading the authors to conclude that rapid volume changes are absorbed by variations in the systemic venous blood volume. The pulmonary blood volume is preserved within narrow limits [3].

Pulmonary microvascular endothelium and interstitium Pulmonary endothelial glycocalyx Inagawa and colleagues have imaged the pulmonary microvascular endothelium of the lungs of laboratory mice before and after exposure to endotoxin (lipopolysaccharide). The glycocalyx reminded them of moss, entirely covering the endothelial cell surface, before lipopolysaccharide was given. They noted that the glycocalyx appeared to be thinner in lung capillaries compared to other tissues such as heart and kidney. Endotoxinaemia severely disrupted the glycocalyx, which in surviving animals at 48 hours appeared to be peeling away.

150

Chapter 7

Numerous spherical structures containing glycocalyx were reported on the endothelial surface: my interpretation is that the investigators were seeing endothelial microparticles (EMPs) [4] which have been observed in patients with septic shock [5]. Intravital microscopy has been used to show pulmonary glycocalyx degradation and neutrophil adhesion in endotoxinaemic mice [6]. Severe pulmonary microvascular endothelial glycocalyx dysregulation or fragmentation is therefore one possible aetiology of acute lung injury [7]. Neutrophil depriming role of the pulmonary circulation Another acute lung injury hypothesis is based on neutrophil physiology without primary glycocalyx damage. Researchers at Cambridge University have demonstrated that unprimed neutrophils pass through the healthy human pulmonary vasculature at about the same speed as red blood cells, but “primed” neutrophils [8] are retained and then 'deprimed’ before release into the systemic circulation. This physiological depriming mechanism appears to have failed in many patients with the clinical syndrome of acute respiratory distress, resulting in increased numbers of primed neutrophils within the systemic circulation. This may contribute to the remote organ damage typically observed in patients with acute respiratory distress syndrome [9]. Such new insights lay the groundwork for future research into possible therapeutic strategies and the role of the pulmonary circulation in systemic inflammatory diseases [10]. It has been calculated that the process of neutrophil depriming takes about 35 minutes [11]. Pulmonary vascular pericytes The pulmonary microvascular endothelium is continuous and nonfenestrated, and associated with pericytes derived from mesenchymal stem cells. The pulmonary pericytes facilitate appropriate angiogenesis, and abnormal pericyte coverage of pulmonary microvessels is a pathological feature of pulmonary arterial hypertension [12]. They have also been implicated in the pathophysiology of other chronic lung diseases, including asthma and pulmonary fibrosis [13]. Starling forces While pulmonary artery catheterisation was commonplace in Intensive Care Units, it was well known that pulmonary capillary pressure is

Pulmonary circulation

151

substantially lower than capillary pressures in systemic circulatory tissues. As the lungs reside in a sub-atmospheric pressured body space, the interstitial fluid pressure there is lower than that in subcutaneous tissue and fluid from pulmonary capillaries is primarily drawn into the perivascular interstitial space by virtue of the subpleural pressure there. In the progression of pulmonary oedema, fluid next accumulates in the interstitial space around the airways forming “peribronchial cuffs”. The Staverman reflection co-efficient sigma for albumin in pulmonary capillaries is, on average, around 0.7. Accordingly, the colloid osmotic pressure of pulmonary interstitial fluid when plasma protein concentration is normal is about 14 mmHg. To slightly complicate matters, we have noted that bronchial artery-supplied capillaries also filter fluid to the pulmonary interstitium and so contribute to pulmonary lymph flow [14]. Moreover, substantial heterogeneity in pulmonary endothelial permeability is induced by supernatants of lipopolysaccharide-stimulated leukocytes derived from patients with early sepsis and provide insights into some of the mechanisms that induce lung vascular injury [15]. Altered ion transport and shifts in the homeostasis of sphingolipids, angiopoietins and prostaglandins are all potentially important players in the regulation of pulmonary microvascular endothelial permeability [16, 17]. Many standard physiology textbooks show a balance sheet of representative Starling forces which conveniently declare a small net filtration pressure favouring filtration. Example values would be: • • •

Colloid osmotic pressures plasma 28, interstitium 14 mmHg; hydrostatic pressures capillary 7, interstitium -8 mmHg; net filtration pressure = 1 mmHg.

This approach is, of course, dangerously simplistic in that it suggests a small increase in plasma colloid osmotic pressure, perhaps by albumin infusion, will bring about sustained absorption of interstitial fluid, while a small decrease will increase the extravascular lung water. Recall that the steady-state Starling principle takes into account the fact that the perivascular interstitial colloid osmotic pressure rises as filtration rate across continuous capillaries falls, so preserving steadystate filtration at any plasma colloid osmotic pressure. What varies between tissues is the time taken to restore the steady-state equilibrium. Some are very quick; some are more prolonged. In animal

152

Chapter 7

studies it has been observed that pulmonary lymph flow, normally quite low, increases substantially when isotonic salt solution is infused to lower plasma colloid osmotic pressure. However, oedema did not manifest so long as pulmonary artery pressure was kept normal by titrated withdrawal of blood [18]. We also have to account for the observation that acute pulmonary oedema once established is only partially cleared by increased pulmonary lymph flow [19]. It would appear that the pulmonary microcirculation can sustain ‘transient’ transendothelial absorption of interstitial fluid to the plasma for a longer period than other tissues before steady-state is re-established. Alveolar walls are extremely fragile, and the alveolar epithelium covering the alveolar surfaces is so weak that it can be ruptured by any positive pressure in the interstitial spaces greater than alveolar air pressure (greater than 0 mmHg), which accounts for transfer of excess fluid from the interstitial spaces into the alveoli. It has long been appreciated that lymphatic vessels pump fluid from the pulmonary interstitial spaces. Except for a small amount that evaporates in the alveoli, filtered fluid is normally pumped back to the circulation through the pulmonary lymphatic system. However, lymph flow response to acute pulmonary oedema can be supplemented by a longer ‘transient’ period of transendothelial absorption. Pulmonary interstitium The pulmonary extracellular matrix is composed of a threedimensional fibre mesh that is filled with various macromolecules, among which are the glycosaminoglycans (GAGs) [20]. The proteoglycan (PG) families may be distinguished on their glycosaminoglycan (GAG) composition, molecular weight and function. Versican contains chondroitin sulphate. Versican is a large molecule (>1000 kDa) found in association with lung fibroblasts and blood vessels in regions not occupied by the major fibrous proteins collagen and elastin. It is localized mainly in the interstitium, creating aggregates with hyaluronic acid. Perlecan and glypican contain heparan sulphate. Perlecan is the largest PG in the lung and is a typical component of vascular basement membrane where it provides a filtration barrier interacting with collagen IV. Together they limit the flow of macromolecules or cells between the two tissue compartments. Perlecan also regulates the

Pulmonary circulation

153

interaction of the basic fibroblast growth factor (FGF) with its receptor and modulates tissue metabolism. Glypican occurs in the cell surface. Syndecan contains chondroitin and heparan sulphates. Syndecan is found in the cell surface. Decorin contains dermatan sulphate. Decorin is found both in the interstitium and in the epithelial basement membrane where it is linked with collagen fibrils. Pulmonary lymphatics Pulmonary parenchymal collecting lymphatics direct lymph to the nodes lying at the bifurcations of the larger bronchi, then on to the tracheobronchial nodes and finally to the bronchomediastinal lymph trunk on each side. The bronchomediastinal lymph trunks typically drain directly into the junction of the internal jugular and subclavian veins on each side, but may drain, on the right, into the right lymph trunk and, on the left, into the thoracic duct. After acute lung injury from smoke inhalation, lung lymph flow is markedly elevated. The lymph drainage from the airway’s bronchial artery-supplied blood may play an important role in this response. Bronchial blood flow is markedly increased after inhalation injury with airway oedema. The increases in lung lymph flow and extravascular lung water are markedly reduced by occlusion of the bronchial artery [21]. Pulmonary lymphatics are obviously important in neoplastic and inflammatory diseases. Understanding the role of lymphatics in human lung disease may contribute to the development of novel therapeutic targets [22].

Negative pulmonary interstitial pressure: keeping the alveoli dry Alveolar epithelium is a discontinuous barrier between the pulmonary interstitial spaces and the alveoli. Fenestrations between the alveolar epithelial cells admit even large protein molecules, as well as solvent and smaller solutes. In health, alveoli do not flood because the pulmonary capillaries and the pulmonary lymphatic system normally

154

Chapter 7

maintain a more negative pressure in the interstitial spaces. Alveolar fluid is normally absorbed into the lung interstitium through the small openings between the alveolar epithelial cells, and carried by afferent lymph vessels to pulmonary lymph nodes. The lymph nodes absorb some of the prenodal lymph to the plasma passing through their microcirculation and conduct the remainder to the thoracic duct as efferent lymph.

The distribution of pulmonary oedema Hydrostatic pressure gradients from the alveoli to the hilar interstitial space determine the perfusion of microvascular filtrate through the pulmonary interstitium. The interstitial pressure gradient leads fluid into the connective tissue surrounding the pulmonary artery, airways, and veins. The valved pulmonary lymphatics are able to pump lymph from there. Propulsion is determined by the intrinsic contractility of lymphatic vessels and by the cycling pressures of inspiration and expiration. While pulmonary microvascular fluid filtration rate exceeds the lymph pumping capacity, fluid accumulates primarily in the hilar regions and in sheaths around the large pulmonary vessels, where interstitial compliance is high. Perivascular cuffs and perihilar fluid are therefore characteristic of pulmonary oedema on chest radiographs. The capacity to increase pulmonary lymph flow is limited. Beyond maximal capacity, further increase in solvent filtration leads to accumulation of interstitial fluid. Eventually, interstitial fluid accumulation compresses lymph vessels causing a spiral of slowing lymph flow and accelerating interstitial fluid volume accumulation. Thus, pulmonary oedema can manifest suddenly rather than gradually; a catastrophic disequilibrium rather than a steady rise in extravascular lung water. Figure 7.1 illustrates the clinical course of a patient who suffered intra-operative rectal perforation and haemorrhage. He was admitted to the Intensive Care Unit in septic shock and intermittent positive pressure ventilation was continued. The trial drug administered on Day 1 was Dazoxiben, a thromboxane synthase inhibitor, and plasma thromboxane B2 levels became unmeasurably low. Disseminated intravascular coagulation developed on Day 2. Extravascular lung water volume suddenly doubled late in Day 3 and was apparent on a chest radiograph taken shortly after, fulfilling criteria for acute respiratory distress syndrome. Positive end expiratory pressure did not improve arterial oxygenation as expressed

Pulmonary circulation

155

by the alveolar to arterial oxygen tension difference.

Figure 7.1

Pleural fluid The normal pleural space is approximately 18 to 20 μm in width, although it widens at its most dependent areas. It has been shown that the pleural membranes do not touch each other and that the pleural space is a real, not a potential, space. Normal pleural fluid is mostly filtered from systemic capillaries fed by the bronchial arteries and draining into pulmonary veins, because the systemic filtration pressure is greater than in capillaries fed by the pulmonary artery. The protein concentration of normal pleural liquid is low, which affirms that Staverman’s reflection co-efficient of the microvessels that produce the filtrate is higher than in pulmonary capillaries. Pleural membranes offer little resistance to liquid or protein movement, and intrapleural pressure is normally lower than the interstitial pressure. The entry of pleural liquid from the visceral pleura is slow and compatible with known Starling forces and interstitial fluid flow rates. Though transporters and aquaporins have been identified in the pleural mesothelial layer, they have not been shown to have a role in reabsorption of effusions. Pleural solvent and solutes leave the pleural space by bulk flow, not by diffusion or active transport. The parietal pleural lymphatics connect to the pleural space via stomas, holes of 8 to 10 μm in diameter that are formed by discontinuities in the

156

Chapter 7

mesothelial layer where mesothelium joins to the underlying lymphatic endothelium. The stomas are large enough to admit cellular elements. These lymphatics have been shown to be the major route of exit of liquid from the pleural space. Decreased rate of pleural fluid clearance is most commonly due to disorders of lymphatic function. Examples include obstruction of the parietal pleural stomas, inhibition of lymphatic contractility, infiltration of draining parasternal lymph nodes, or elevation of the systemic venous pressure into which the lymph drains. Accumulation of pleural fluid creates pleural effusions. Transudates form by leakage of liquid across an intact capillary barrier, due to increases in hydrostatic pressures or decreases in colloid osmotic pressures across that barrier. Transudates generally indicate that the pleural membranes are not themselves diseased. Exudates generally form from leakage of liquid and protein across an altered capillary barrier with increased permeability. The protein ratio, lactate dehydrogenase ratio, and absolute pleural lactate dehydrogenase concentration constitute Light's criteria [23]. Congestive heart failure is the commonest cause of significant pleural effusion, and is a transudate formed from leakage of oedema across normal pulmonary capillaries into the pulmonary interstitium. This interstitial edema can then move toward the pleural space and across the leaky visceral pleura into the pleural space. Other transudates include those associated with the nephrotic syndrome (altered osmotic pressure) or atelectasis (altered hydrostatic pressure). Hepatic hydrothorax and effusions from peritoneal dialysis develop when fluid flows from the higher pressure peritoneal space into the lower pressure pleural space across ruptured herniations in the central diaphragmatic tendon [24]. As the surface area of the central diaphragmatic tendon is greater on the right than on the left, such effusions are most frequently right sided. The pleural space provides a pathway for pulmonary oedema clearance and can function as an additional safety factor protecting against the development of alveolar oedema. Carbon particles can be injected into the pleural space as a visible marker of lymphatic drainage pathways, and the black carbon can be seen to have been taken up into lymphatics on the parietal side, not the visceral side. The visceral pleura has extensive lymphatics, but they do not connect to the pleural space.

Pulmonary circulation

157

References 1. Dock DS, Kraus WL, McGuire LB, Hyland JW, Haynes FW, Dexter L. The pulmonary blood volume in man. J Clin Invest. 1961;40:317328. 2. De Freitas FM, Faraco EZ, NedelL N, De Azevedo DF, Zaduchliver J. Determination of pulmonary blood volume by single intravenous injection of one indicator in patients with normal and high pulmonary vascular pressures. Circulation. 1964;30:370-380. 3. De Freitas FM, Faraco EZ, De Azevedo DF, Zaduchliver J, Lewin I. Behaviour of normal pulmonary circulation during changes of total blood volume in man. J Clin Invest. 1965;44:366-378. 4. Curtis AM, Edelberg J, Jonas R et al. Endothelial microparticles: sophisticated vesicles modulating vascular function. Vasc Med. 2013;18:204-214. 5. Delabranche X, Boisramé-Helms J, Asfar P et al. Microparticles are new biomarkers of septic shock-induced disseminated intravascular coagulopathy. Intensive Care Med. 2013;39:1695-1703. 6. Schmidt EP, Yang Y, Janssen WJ et al. The pulmonary endothelial glycocalyx regulates neutrophil adhesion and lung injury during experimental sepsis. Nat Med. 2012;18:1217-1223. 7. Collins SR, Blank RS, Deatherage LS, Dull RO. Special article: the endothelial glycocalyx: emerging concepts in pulmonary edema and acute lung injury. Anesth Analg. 2013;117:664-674. 8. Neutrophil priming: pathophysiological consequences and underlying mechanisms. [editorial]. Clin Sci (Lond) 1998;94(5):461. 9. Summers C, Singh NR, White JF et al. Pulmonary retention of primed neutrophils: a novel protective host response, which is impaired in the acute respiratory distress syndrome. Thorax. 2014;69:623-629. 10. Millar FR, Summers C, Griffiths MJ, Toshner MR, Proudfoot AG. The pulmonary endothelium in acute respiratory distress syndrome: insights and therapeutic opportunities. Thorax. 2016;71:462-473. 11. Summers C, Chilvers ER, Peters AM. Mathematical modeling supports the presence of neutrophil depriming in vivo. Physiol Rep. 2014;2:e00241. 12. Yuan K, Shamskhou EA, Orcholski ME et al. Loss of Endothelial Derived WNT5A is Associated with Reduced Pericyte Recruitment and Small Vessel Loss in Pulmonary Arterial Hypertension. Circulation. 2018 13. Rowley JE, Johnson JR. Pericytes in chronic lung disease. Int Arch Allergy Immunol. 2014;164:178-188.

158

Chapter 7

14. Wagner EM, Blosser S, Mitzner W. Bronchial vascular contribution to lung lymph flow. J Appl Physiol (1985). 1998;85:2190-2195. 15. Leligdowicz A, Chun LF, Jauregui A et al. Human pulmonary endothelial cell permeability after exposure to LPS-stimulated leukocyte supernatants derived from patients with early sepsis. Am J Physiol Lung Cell Mol Physiol. 2018;315:L638-L644. 16. Simmons S, Erfinanda L, Bartz C, Kuebler WM. Novel mechanisms regulating endothelial barrier function in the pulmonary microcirculation. J Physiol. 2019;597:997-1021. 17. Gutbier B, Schönrock SM, Ehrler C et al. Sphingosine Kinase 1 Regulates Inflammation and Contributes to Acute Lung Injury in Pneumococcal Pneumonia via the Sphingosine-1-Phosphate Receptor 2. Crit Care Med. 2018;46:e258-e267. 18. Hara N, Nagashima A, Yoshida T, Furukawa T, Inokuchi K. Effect of decreased plasma colloid osmotic pressure on development of pulmonary edema in dogs. Jpn J Surg. 1981;11:203-208. 19. Mackersie RC, Christensen J, Lewis FR. The role of pulmonary lymphatics in the clearance of hydrostatic pulmonary edema. J Surg Res. 1987;43:495-504. 20. Souza-Fernandes AB, Pelosi P, Rocco PR. Bench-to-bedside review: the role of glycosaminoglycans in respiratory disease. Crit Care. 2006;10:237. 21. Traber DL, Lentz CW, Traber LD, Herndon DN. Lymph and blood flow responses in central airways. Am Rev Respir Dis. 1992;146:S15-8. 22. El-Chemaly S, Levine SJ, Moss J. Lymphatics in lung disease. Ann N Y Acad Sci. 2008;1131:195-202. 23. Light RW, Macgregor MI, Luchsinger PC, Ball WC. Pleural effusions: the diagnostic separation of transudates and exudates. Ann Intern Med. 1972;77:507-513. 24. Kinasewitz GT, Keddissi JI. Hepatic hydrothorax. Curr Opin Pulm Med. 2003;9:261-265.

CHAPTER 8 PLASMA VOLUME

Chapter summary The Starling principle guides clinicians in their aim to achieve optimal plasma volume without causing oedema. The steady-state Starling principle enables us to examine the kinetics of extracellular fluid circulation from plasma, through the interstitial space, and returning to veins via the lymphatics. We can balance the distribution of extracellular fluid by manipulating the Starling forces. Hydrostatic pressure is the Starling force driving filtration of fluid from the capillaries. Microcirculatory optimisation is all about avoiding capillary hypertension while we maintain adequate tissue blood flow. Sigma delta ȫ, the effective colloid osmotic pressure difference across the endothelial glycocalyx, is the sum of Starling forces that opposes transendothelial filtration Jv but it does not in steady-state reverse the flow to reabsorption. Additional therapeutic options arise from understanding how interstitial fluid is pumped as lymph to the circulating blood volume.

Introduction If I may be allowed a personal reflection, I was privileged to know Alison Twigley and Ken Hillman when we all worked at Charing Cross Hospital in London. I was one of the many anaesthetists seduced by the simplicity and elegance of the body fluid compartments paradigm that they published in the essential reading Journal of the time, Anaesthesia, in 1985 [1]. Their box diagram of static pools of plasma, interstitial fluid and intracellular fluid and the manipulatable mechanisms that moved fluid between them was copied many times but rarely improved over the following years. Unfortunately, it came to be seen as an expression

160

Chapter 8

of physiological science rather than what it is; an over-simplified quantitative paradigm for the rational prescription of intravenous fluid and diuretics. There is now a generation of anaesthetists and critical care physicians who seem to believe that plasma volume is determined by the presence of albumin. Their mantra runs something like this: • • • •

If the patient loses blood, resuscitate with blood. If the patient loses plasma through burns or sepsis, resuscitate with albumin. If the patient is short of salt and water (isosmotic dehydration), resuscitate with an isotonic salt solution. If the patient is short of water (hyperosmotic dehydration), resuscitate with an electrolyte-free solution.

I know from participating in colloquia that this reasoning is still used by many Professors in the art of anaesthesia and critical care medicine, and the purpose of this monograph is to put the record straight. One of these Professors declared that our central purpose in fluid therapy is to achieve an optimal plasma hypervolume state so that the circulation of oxygenated blood may confer its healing powers. The challenge, he declared, is to keep infused fluid within the intravascular space. In this chapter I therefore look specifically at the physiological mechanisms that support and regulate plasma volume.

Defining and measuring plasma volume Plasma is the biological solution in which circulating red blood cells are suspended? Much of our knowledge about plasma volume is deduced from accompanying changes in the haematocrit or haemoglobin concentration. Carbon monoxide (CO) re-breathing can be used to measure haemoglobin mass, while sodium fluorescein labelling can be used to determine the red blood cell mass. Indirect estimates of percentage change of plasma volume (without knowledge of absolute values) are often made from serial concomitant haemoglobin and haematocrit concentrations. (%) = 100 x (Hb before/ Hb after) x (1 - Hct after/ 1 - Hct before) - 100.

(before = haemoglobin concentration or haematocrit at baseline, after = after intervention.)

Plasma Volume

161

Of course, plasma volume can be measured at baseline using one of the standard techniques described below and then any subsequent changes can be followed indirectly with this method. However, during a non-orthostatic stress (e.g., saline infusion in the supine position), changes in Hct and Hb may confidently be used for estimating changes in plasma volume. Whole body haematocrit is higher than measured from a venous sample, and a compensating adjustment is required; whole body/venous haematocrit correction factor of 0.91 has been found to be satisfactory. The total haemoglobin mass CO re-breathing method is reportedly easy to perform [2] and may be repeated 3 hours after oxygen washout of CO [3]. Plasma volume is the central volume of distribution of albumin? Several methods are available for directly determining plasma volume, all of which involve the principles of dilutional analysis employed after the administration of tracer molecules. The recommended gold standard method is human serum albumin radiolabelled with iodine (125I-HSA or 131I-HSA). A semi-automated system that can be used in a critical care environment is able to report total blood volume, plasma volume and red cell volume using the indicator dilution principle, microhaematocrit centrifugation and ideal height and weight. Once the radio iodine albumin tracer is injected a technician takes five blood samples over six minutes which undergo microhaematocrit centrifugation to extrapolate true blood volume at time zero. The microhaematocrit data along with the indicator data provide a normalized haematocrit number, more accurate than haematocrit or peripheral haematocrit measurements [4, 5]. The system also reports albumin transudation time, equivalent to the transcapillary escape rate for albumin. Evans blue dye binds to albumin, and its subsequent loss from the circulation is dependent on the rate of loss of albumin. After injection, indocyanine green is rapidly bound to plasma proteins before hepatic elimination. This method is safe in clinical practice, but the half-life of indocyanine green is short, and even minor inaccuracies in sampling time can lead to significant errors. The plasma concentration of dyes is calculated spectrophotometrically by determining its absorbance at a specific wavelength, but the method has limitations which can lead to overestimation of plasma volume. Furthermore, the early sampling required for indocyanine green estimations might not fully allow

162

Chapter 8

adequate mixing when cardiac output is reduced. Plasma volume is the central volume of distribution of Dextran 70? Dextran 70 has a relatively long half-life and can be coupled to fluorescent compounds, thereby negating the need for radiolabels. In one comparison the mean difference in plasma volumes when determined by 125I-HSA and dextran-70 was 6%. Smaller molecular weight dextrans such as Dextran 40 may penetrate the endothelial surface layer more quickly to give an estimate of intravascular fluid volume [6]. Plasma volume is the intravascular extracellular fluid? I find this concept most helpful because the vessels that contain blood can constrict or dilate either to accommodate fluid that is entering the circulation at a steady pressure, or to vary the intravascular pressure of fluid already present. It also has limitations. The sinusoidal tissues do not have a continuous capillary membrane separating the intravascular from extravascular fluids so the Space of Disse, for example, acts as an extravascular reservoir of plasma fluid. Figure 8.1 is a cartoon of the volume kinetics of extracellular fluid, and the intravascular space (discontinuous circle) delimits the plasma volume as the central extracellular fluid volume. This diagram enables us to list the ways in which plasma volume can be regulated. Going clockwise:

Plasma Volume

163

Figure 8.1

• •

• •

• •

Enteral (or parenteral) fluid intake increases plasma volume, the central volume (Vc) of extracellular fluid. Jv is determined by the Starling forces. Hydraulic conductance (or, if you prefer, resistance to fluid flow across the permeability barrier) is a significant factor. Jv reduces Vc and increases the tissue fluid volume (Vt). Tissue fluid losses include the so-called insensible losses by perspiration and the water saturated of exhaled air. Qlymph is actively pumped to move fluid from the lowest pressured body spaces to the venous system. It increases Vc with protein-rich fluid and reduces Vt. Lymphatic pump failure is an underappreciated cause of interstitial fluid (Vt) accumulation. Haemorrhage and, more normally, urine output drain Vc. Vasodilation and vasoconstriction can accommodate volume changes or vary intravascular pressure.

Volume Sensors and neurohumoral response Cardiac volume and pressures are sensed, and regulated by variation of the plasma volume through salt and water balance as well as vascular

164

Chapter 8

tone. Low-pressure volume receptors are mostly located in the cardiac atria. Stretch of these receptors results in the local secretion of atrial natriuretic peptide and an alteration in neural signalling to the hypothalamus and medulla oblongata. Arginine-vasopressin synthesis and release and sympathetic nervous system discharge occur. High-pressure receptors are located in the left ventricle, aortic arch and carotid sinus. Decrease in arterial pressure results in the inhibition of baroreceptor discharge. Activation of the sympathetic nervous system and the renin-angiotensin-aldosterone systems counteract the drop in blood pressure. Vagal tone is reduced so that heart rate can rise, and there is non-osmotic release of arginine-vasopressin. B-type natriuretic peptide production by the ventricle facilitates natriuresis and diuresis. The renal juxtaglomerular apparatus senses a reduction in renal perfusion pressure or a reduction in tubular sodium load and enhances renin release. Renin release is also facilitated by sympathetic nervous system activation and inhibited by atrial natriuretic peptide. Efferent responses Sodium and water retention mechanisms have to be counterbalanced by natriuretic responses. The existence of a third “natriuretic factor” for regulating sodium balance was first suggested by Hugh De Wardener and his colleagues in 1961 [7]. Sodium retention Notice that kidneys normally reabsorb almost all the sodium from the glomerular filtrate. A reduction in renal tubular reabsorption of sodium from 99% to 98% will double sodium excretion. During reabsorption, sodium moves from the tubular lumen to the peritubular capillaries. It follows that reabsorption can be regulated either by physical factors, such as the Starling forces across the renal tubule, or by humoral agents such as aldosterone. Water retention Non-osmotic triggers of arginine vasopressin synthesis defend the blood volume by antidiuresis and by accelerated erythrocyte

Plasma Volume

165

production [8]. Natriuresis and diuresis The normal response to increasing blood volume in humans is a diuresis mediated by: • • • •

Renal vasodilation by selective inhibition of renal sympathetic nervous activity Inhibition of arginine vasopressin secretion Inhibition of renin angiotensin aldosterone system activity Natriuretic peptides secretion

This is also the physiological response to long-term weightlessness as experienced by astronauts: with no peripheral pooling of venous blood in the legs, plasma volume becomes contracted. On return to gravity, they may experience orthostatic intolerance until their plasma volume is restored by salt and water retention. In laboratory experiments atrial natriuretic peptide’s renal effect is supplemented by a peripheral endothelial action, increased vascular permeability to proteins [9]. Increased transcapillary escape rate for albumin should augment plasma volume reduction and limit a rise in plasma albumin concentration. Natriuresis and diuresis after the infusion of atrial natriuretic peptide do result in increased plasma albumin concentration in humans, showing that the dominant influence of atrial natriuretic peptide is on the transendothelial solvent filtration rate rather than the transcapillary escape rate of albumin.

Disorders of plasma volume Obesity Obesity is complicated by increased signalling through leptin receptors, which can promote activation of both the sympathetic nervous system and the renin angiotensin aldosterone system, and can directly stimulate the secretion of aldosterone [10]. Obese people therefore experience greater renal tubular sodium reabsorption and have a raised plasma volume proportional to their body mass index. This predisposes them to hypertension and heart failure.

166

Chapter 8

Hypertension Expanded plasma volume is also found in patients with primary hyperaldosteronism. Plasma volume, as measured by radio-iodinated albumin dilution, is typically diminished in essential hypertension [11]. The decrease has been found to be greater with higher diastolic pressure, but only evident in the patients with diastolic blood pressures over 105 mmHg. Heart failure Heart failure with a preserved ejection fraction (HFpEF) is the commoner form of heart failure, and the vast majority of afflicted individuals are overweight or obese [12]. The underlying pathophysiologies are sodium retention and systemic inflammation. The adipose tissue of obese people undergoes a biological transformation to an inflammatory state, which contributes to vascular rigidity (loss of compliance), endothelial dysfunction and reduced capillary density. The latter is sometimes called microcirculatory rarefaction. The raised state of inflammation may directly affect the heart, lungs, and kidneys, leading to the co-morbidities that are characteristic of HFpEF. Ventricular dilation in response to blood volume (plasma volume) expansion may be impaired when adipose tissue inflammation leads to dysfunction and fibrosis of the underlying myocardium. Ventricular compliance is limited, the ventricles becoming stiff. Cardiac volumes are increased by blood volume expansion, but only modestly: they are not normal or small as is the case in hypertrophic cardiomyopathy, yet they are insufficient to produce a circulatory flow rate appropriate to the elevated blood volume. Blood accumulates on the venous side of the circulation, leading to the signs and symptoms of heart failure (particularly congestion). It is also believed that the systemic inflammation of obesity-related HFpEF can cause changes in mitochondrial function and in the mass and composition of skeletal muscle. These abnormalities, acting in concert with the venous and arterial abnormalities that accompany inflammation, can contribute to exercise intolerance. Non-obese patients with HFpEF have much less plasma volume expansion than subjects with obese HFpEF, but higher N-terminal proB-type natriuretic peptide levels. Pulmonary capillary wedge pressure

Plasma Volume

167

is independent of plasma volume in most non-obese HFpEF patients, and pericardial restraint appears to be less of a feature [13]. Pharmacological diuresis is often prescribed for the correction of fluid overload and pulmonary oedema, but the treatment can actually increase plasma volume when initiated [14]. It is therefore proposed that the venous capacitance effect of furosemide lowers capillary pressure and transcapillary solvent filtration. At the same time intravascular volume is replenished by protein-rich lymph flow at a rate equal to or in excess of the volume removed by diuresis [15, 16]. With chronic diuresis, sustained reduction of extracellular fluid volume will result in proportionately reduced plasma volume. Drug-refractory congestion is the major cause of hospitalisations and determinant of fatality. Technological advances have provided ultrafiltration devices that can be used in lower-acuity hospital settings. Safety and efficacy data are still being assessed, but ultrafiltration performed at fixed rates after onset of therapy-induced increased serum creatinine was not superior to standard care and resulted in more complications [17]. Pregnancy and preeclampsia Plasma volume increases in the early weeks of pregnancy, continuing into the third trimester with an eventual maximum increase of 1 to 1.2 l. The plasma volume expansion with gestational hypertension and in growth-restricted pregnancies is only 0.8 l [18, 19]. Iatrogenic fluid overload While fluid overload has long been discussed as a feature of heart failure and renal failure, excessive intravenous fluid administration has only more recently become a topic of concern. There is a widelyaccepted view that fluid overload is an almost inevitable sequela of successful fluid resuscitation and that subsequent “deresuscitation” is indicated [20]. Fluid overload seems to be a sine qua non of the acute respiratory distress syndrome [21] and may worsen obstructive sleep apnoea [22]. It sometimes happens that patients with compromised pulmonary capillary permeability and acute respiratory distress syndrome are judged to be clinically hypovolaemic and treated with intravenous

168

Chapter 8

fluids. Whilst greater volumes of crystalloid were needed to achieve a haemodynamic improvement in a clinical report from Amsterdam, colloid conferred no advantage in terms of lung water change or pulmonary signs and symptoms [23]. This apparent contradiction of the traditional Starling principle is made explicable by application of the steady-state Starling principle [24].

Albumin and plasma volume Albumin has a dazzling range of physiological properties which are examined in the next chapter, but it cannot be considered an essential or vital protein because people who do not make albumin have only mild signs and symptoms at all ages. With appropriate fluid and electrolyte supervision, hospitalised patients with malnutritional hypoalbuminaemia can have adequate circulating blood volume without being oedematous. There are critical care experts who teach that plasma volume is reduced in hypoalbuminaemia while the facts are otherwise. NASA scientists have shown that after two weeks at zero gravity plasma volume is reduced in spite of elevated plasma colloid osmotic pressure [25]. Back on Earth, hypoproteinaemic patients with nephrotic syndrome and hypoalbuminaemia have normal or elevated plasma volume [26]. In Chapter Two we looked at the long history of misunderstandings about colloid osmotic pressure and plasma volume which underpin the market for albumin therapy and artificial biophysical colloid osmotic pressure therapies. Resuscitation from hypovolaemia requires less colloid solution than crystalloid solution If blood volume is reduced so far that capillary pressure falls, transendothelial solvent filtration from the plasma approaches zero: with very acute reduction of capillary pressure there is a transient reversal of transendothelial solvent flow from the interstitium to plasma. This phenomenon is known as autotransfusion. At the same time efferent lymph flow increases, injecting a high-protein solution into the central veins. Clinicians feel the need to restore circulating blood volume fully, though they lack evidence that to do so improves the patient’s prospect of survival.

Plasma Volume

•

•

169

When they infuse a solution that preserves or raises colloid osmotic pressure, they dilute erythrocytes within the freeflowing plasma volume. The solvent is restrained from hydrating the intravascular gel phase by the colloid osmotic pressure difference due to the infused colloid molecules which are excluded from the gel phase. When they infuse a crystalloid solution, they dilute the whole of the intravascular space including the gel phase. While there is no filtration of intravascular solvent to the interstitium while capillary pressure is low, the whole infused volume will increase the intravascular volume but with a lesser effect on haemodilution.

There is a range of reduced capillary pressure, or more precisely the capillary-interstitial pressure difference, over which there is minimal transendothelial solvent filtration. Graphically plotted as flow versus pressure we see a flat relationship, the foot of the hockey stick curve as described by Michel. The colloid delusion It has been widely observed in clinical trials that colloid resuscitation is associated with lower haematocrit than crystalloid resuscitation, and accordingly a greater risk of exposure to red blood cell transfusion. Sometimes attributed to a colloid-related coagulopathy and increased blood loss, the major factor is the different central volumes of distribution of the infused resuscitation fluids. This difference is also interpreted as demonstrating better plasma volume expansion with a colloid: I have called this the colloid delusion. What matters in resuscitation is the intravascular volume, or more specifically the effective blood volume. While sodium solutions of higher tonicity bring about vasoconstriction, histamine release associated with infused colloid solutions could have enough vasodilation and capillary permeability effect to offset the colloid osmotic pressure advantage [27]. I explore the concepts of context sensitivity and volume equivalence in Chapter 11. Iatrogenic hypervolaemia is more persistent after intravenous colloid plasma volume expansion than after intravenous crystalloid plasma volume expansion When the blood volume rises above normal and the capillary-interstitial pressure difference rises above normal, the transendothelial solvent

170

Chapter 8

filtration rate rises. Graphically plotted as flow versus pressure, we see dependence of flow on pressure, the handle of the hockey stick or J curve. A very overfilled intravascular volume empties rapidly, while a slightly overfilled intravascular volume empties more slowly. At the inflection point (also called the J point) and below, the intravascular volume is balanced by its inflow and outflow rates. This is the steadystate. After a colloid solution has been used to overfill the blood volume, the outflow rate is retarded by the transendothelial balance of Starling forces and it takes longer to return to steady-state. So-called haemodynamic optimisation is a clinical example. It is important to challenge suggestions by some experts that prolonged anaemia reported after colloid therapy is a sign of higher plasma volume, still less an indication of adequate effective blood volume.

A revised Twigley-Hillman diagram The Twigley-Hillman diagram has the advantage of being recognisable and comprehensible. I therefore offer here a revised Twigley-Hillman diagram, Figure 8.2. It represents: • • •

•

•

Plasma. The free-flowing intravascular plasma volume is normally around 3 l of the 4 l intravascular fluid. Red blood cell volume is intracellular fluid, and is approximately 2 l, but is not shown separately on this diagram. Circulating erythrocytes are repelled by the outermost, rather porous gel phase of the endothelial surface layer while the inner fibre matrix of the membrane-bound glycocalyx performs the small pore function of excluding larger molecules, including albumin. It is impossible to attribute a precise volume to this phase, but it is about 1 l in health. Aqueous and gel fluid phases of the triphasic interstitium are distinguished. In health the aqueous phase may be as little as 1%, but increases greatly in oedema. The third phase is structural, mostly collagen Type I fibres, which appear to channel the interstitial flow of solvent and solutes. Sodium hyaluronan is a non-osmotic buffer of sodium ions in the gel phase, and albumin is excluded. Lymph is aqueous interstitial fluid that has entered the lymphatic vasculature.

Plasma Volume

•

171

Intracellular fluid.

Figure 8.2

The revised Twigley-Hillman diagram of total body water compartmentalisation emphasises the extracellular fluid (ECF) circulation (arrows) which occurs in most tissues most of the time. The intravascular space normally contains about 5 litres of blood and 1 litre endothelial surface layer (ESL) from which circulating red blood cells are excluded. The intravascular extracellular fluid is free-flowing aqueous (plasma) and gel phase (ESL). ESL contains the fibre matrix molecules of glycocalyx which arise from the endothelial cell surface. The triphasic interstitial space has a structural collagen fibrous phase and around 14 litres of fluid. There is an aqueous phase, a gel phase within which glycosaminoglycans, such as hyaluronic acid, have the capacity to store sodium without raising tissue osmolality, and lymph. The intracellular fluid (ICF) volume, normally about 23 litres, is sensitive to acute changes in ECF osmolality. However, cell volume regulatory mechanisms exist to preserve the steady-state intracellular fluid volume and enable subjects to tolerate chronic hypotonicity.

172

Chapter 8

Water and solutes enter and leave the body across epithelial barriers. In clinical practice we can infuse fluids directly into, or haemofilter water out of the free-flowing plasma. In sinusoidal tissues (liver, spleen and bone marrow) the ESL is discontinuous and there are windows (fenestrations) through the endothelium that exclude red blood cells but admit albumin to the interstitial space. Consequently, there is no trans-endothelial colloid osmotic pressure difference to oppose filtration. In non-sinusoidal tissues the continuous ESL is almost impermeable to albumin so that filtered fluid in the immediate subglycocalyx space (the protected region) has a very low colloid osmotic pressure compared to plasma or the general interstitial fluid. The transendothelial colloid osmotic pressure difference opposing filtration is therefore high. Michel and Weinbaum correctly hypothesised that if transendothelial water movement across non-sinusoidal capillaries and venules is transiently reversed by a sudden drop in the hydrostatic pressure difference, interstitial albumin rapidly enters the protected region, diminishing the transendothelial colloid osmotic pressure difference. The net water movement therefore quickly return to steady-state filtration. This is the Michel-Weinbaum “No Absorption Rule”. Water is absorbed from afferent lymph into lymph node capillaries and venules. Efferent lymph therefore has high protein and lipid content and returns to the central veins via the thoracic duct. Acceleration of protein-rich efferent lymph to the circulating blood volume is an important compensatory response to haemorrhagic shock in humans.

Therapeutic options The revised Twigley-Hillman diagram shows that the steady-state distribution of extracellular fluid between intravascular and the interstitium depends on the balance of two flow rates, the transendothelial solvent filtration rate Jv and the afferent lymph flow rate Qlymph. Recall that around 4 litres per day of the total afferent lymph fluid is absorbed into the blood stream within lymph nodes, the remainder entering central veins as efferent (protein-rich) lymph. The clinician can therefore apply therapeutic strategies that affect these flow rates.

Plasma Volume

173

There are pathways via epithelial cell membranes for fluid output (kidneys, skin, lung etc) and input (enteral route). They can of course be supplemented by haemodiafiltration and parenteral fluid therapies. Transendothelial solvent filtration rate Jv The most important Starling force here is the capillary pressure, which is normally autoregulated but can be manipulated by volume infusions or pharmacological adjustments of vascular tone. Of particular note is the effect of alpha-adrenergic agonists at a non-hypertensive dose to lower capillary pressure by mild arteriolar constriction. By reducing Jv, alpha adrenergic agonists support the plasma volume and minimise accumulation of interstitial fluid as oedema. Alternatively, plasma colloid osmotic pressure can be raised by infusing hyperoncotic albumin solutions in order to slow filtration, especially in euvolaemic subjects, but at the subsequent steady-state there will be more albumin in the interstitium and an increased probability of oedema that is slow to resolve. Transendothelial fluid absorption when capillary pressure is very low (e.g. hypovolaemic shock) is limited and transient (autotransfusion), and the infusion of exogenous (non-native) albumin as so-called low volume resuscitation has only a limited volume benefit over crystalloids. Crystalloids have other advantages over colloid resuscitation fluids and should be given first in clinical practice. Afferent lymph flow rate Qlymph A new appreciation of the importance of Qlymph in fluid physiology is an important reason to study the steady-state Starling principle. Measures such as early ambulation after surgery or injury, active and passive limb movements, and postural variation are all effective ways to encourage Qlymph. Furosemide has been shown to reduce oedema, in part by increasing lymph flow. Recall that lymph flow is not a simple drainage system, it relies on active pumping. Opioids and other drugs that inhibit lymphatic smooth muscle contractility should be used as sparingly as possible. Sympathetic nervous system blockade by axial anaesthesia dramatically reduces Qlymph. Conversely, alpha-adrenergic agonist therapy enhances Qlymph in addition to its anti-oedema effect as a reducer of capillary pressure by arteriolar constriction.

174

Chapter 8

Urine output Diuretics reduce plasma volume and extracellular fluid volume, while intracellular fluid volume is largely unaffected. A notable exception is the use of osmotic diuretics to reduce raised intracranial pressure, with the intention of shrinking acutely-swollen brain cells. It may be a consideration that infused crystalloid solutions produce more diuresis than colloids, making them a better choice for the thankless task of relieving stress-induced oliguria in hospitalised patients. In a trial, no concordance was seen between cardiovascular (macrocirculatory) response and diuresis (microcirculatory) response [28].

References 1. Twigley AJ, Hillman KM. The end of the crystalloid era? A new approach to peri-operative fluid administration. Anaesthesia. 1985;40:860-871. 2. Keiser S, Meinild-Lundby AK, Steiner T et al. Detection of blood volumes and haemoglobin mass by means of CO re-breathing and indocyanine green and sodium fluorescein injections. Scand J Clin Lab Invest. 2017;77:164-174. 3. Plumb JOM, Kumar S, Otto J et al. Replicating measurements of total hemoglobin mass (tHb-mass) within a single day: precision of measurement; feasibility and safety of using oxygen to expedite carbon monoxide clearance. Physiol Rep. 2018;6:e13829. 4. Takanishi DM, Yu M, Lurie F et al. Peripheral blood hematocrit in critically ill surgical patients: an imprecise surrogate of true red blood cell volume. Anesth Analg. 2008;106:1808-1812. 5. Yu M, Pei K, Moran S et al. A prospective randomized trial using blood volume analysis in addition to pulmonary artery catheter, compared with pulmonary artery catheter alone, to guide shock resuscitation in critically ill surgical patients. Shock. 2011;35:220228. 6. Manzone TA, Dam HQ, Soltis D, Sagar VV. Blood volume analysis: a new technique and new clinical interest reinvigorate a classic study. J Nucl Med Technol. 2007;35:55-63; quiz 77, 79. 7. De Wardener HE, MILLS IH, CLAPHAM WF, HAYTER CJ. Studies on the efferent mechanism of the sodium diuresis which follows the

Plasma Volume

175

administration of intravenous saline in the dog. Clin Sci. 1961;21:249-258. 8. Mayer B, Németh K, Krepuska M et al. Vasopressin stimulates the proliferation and differentiation of red blood cell precursors and improves recovery from anemia. Sci Transl Med. 2017;9 9. Chen W, Gassner B, Börner S et al. Atrial natriuretic peptide enhances microvascular albumin permeability by the caveolaemediated transcellular pathway. Cardiovasc Res. 2012;93:141-151. 10. Packer M. Leptin-Aldosterone-Neprilysin Axis: Identification of Its Distinctive Role in the Pathogenesis of the Three Phenotypes of Heart Failure in People With Obesity. Circulation. 2018;137:16141631. 11. Tarazi RC, Frohlich ED, Dustan HP. Plasma volume in men with essential hypertension. N Engl J Med. 1968;278:762-765. 12. Packer M, Kitzman DW. Obesity-Related Heart Failure With a Preserved Ejection Fraction: The Mechanistic Rationale for Combining Inhibitors of Aldosterone, Neprilysin, and SodiumGlucose Cotransporter-2. JACC Heart Fail. 2018;6:633-639. 13. Obokata M, Reddy YNV, Pislaru SV, Melenovsky V, Borlaug BA. Evidence Supporting the Existence of a Distinct Obese Phenotype of Heart Failure With Preserved Ejection Fraction. Circulation. 2017;136:6-19. 14. Schuster CJ, Weil MH, Besso J, Carpio M, Henning RJ. Blood volume following diuresis induced by furosemide. Am J Med. 1984;76:585592. 15. Henriksen JH, Schlichting P. Increased extravasation and lymphatic return rate of albumin during diuretic treatment of ascites in patients with liver cirrhosis. Scand J Clin Lab Invest. 1981;41:589599. 16. Henriksen JH, Parving HH, Lassen NA, Winkler K. Increased transvascular escape rate and lymph drainage of albumin in pigs during intravenous diuretic medication. Relations to treatment in man and transport mechanisms. Scand J Clin Lab Invest. 1982;42:423-429. 17. Costanzo MR, Ronco C, Abraham WT et al. Extracorporeal Ultrafiltration for Fluid Overload in Heart Failure: Current Status and Prospects for Further Research. J Am Coll Cardiol. 2017;69:2428-2445. 18. de Haas S, Ghossein-Doha C, van Kuijk SM, van Drongelen J, Spaanderman ME. Physiological adaptation of maternal plasma

176

Chapter 8

volume during pregnancy: a systematic review and meta-analysis. Ultrasound Obstet Gynecol. 2017;49:177-187. 19. Pretorius T, van Rensburg G, Dyer RA, Biccard BM. The influence of fluid management on outcomes in preeclampsia: a systematic review and meta-analysis. Int J Obstet Anesth. 2018;34:85-95. 20. Silversides JA, Fitzgerald E, Manickavasagam US et al. Deresuscitation of Patients With Iatrogenic Fluid Overload Is Associated With Reduced Mortality in Critical Illness. Crit Care Med. 2018;46:16001607. 21. Sweeney RM, McAuley DF. Acute respiratory distress syndrome. Lancet. 2016;388:2416-2430. 22. Lam T, Singh M, Yadollahi A, Chung F. Is Perioperative Fluid and Salt Balance a Contributing Factor in Postoperative Worsening of Obstructive Sleep Apnea. Anesth Analg. 2016;122:1335-1339. 23. van der Heijden M, Verheij J, van Nieuw Amerongen GP, Groeneveld AB. Crystalloid or colloid fluid loading and pulmonary permeability, edema, and injury in septic and nonseptic critically ill patients with hypovolemia. Crit Care Med. 2009;37:1275-1281. 24. Woodcock TM, Woodcock TE. Revised Starling equation predicts pulmonary edema formation during fluid loading in the critically ill with presumed hypovolemia.[letter]. Crit Care Med 2012;40(9): 2741-2; author reply 2742. 25. Hsieh ST, Ballard RE, Murthy G, Hargens AR, Convertino VA. Plasma colloid osmotic pressure increases in humans during simulated microgravity. Aviat Space Environ Med. 1998;69:23-26. 26. Geers AB, Koomans HA, Boer P, Dorhout Mees EJ. Plasma and blood volumes in patients with the nephrotic syndrome. Nephron. 1984;38:170-173. 27. Celik I, Duda D, Stinner B, Kimura K, Gajek H, Lorenz W. Early and late histamine release induced by albumin, hetastarch and polygeline: some unexpected findings. Inflamm Res. 2003;52:408416. 28. Smorenberg A, Groeneveld AB. Diuretic response to colloid and crystalloid fluid loading in critically ill patients. J Nephrol. 2015;28:89-95.

CHAPTER 9 ALBUMIN AND OTHER CIRCULATING PROTEINS

Chapter summary Albumin is the major contributor to the Starling force we call colloid osmotic pressure, and so it receives most of our attention in this chapter. The fact that soluble proteins circulate around the body is under-appreciated. Measuring albumin in blood is surprisingly complicated, particularly in patients with oxidative stress in whom the degree of reported hypoalbuminaemia does not account for the measured colloid osmotic pressure. We consider the case for therapy by infusion of human albumin solutions, and briefly consider the traditional concept of post-injury metabolic phases. Is it really relevant to modern clinical practice?

Soluble extracellular proteins The group of proteins that we label plasma proteins are soluble, nonfixed molecules that shuttle between plasma and the interstitium. They are therefore more correctly called soluble extracellular proteins, and they exert Starling forces (colloid osmotic pressure) from both sides of the endothelial glycocalyx. In most tissues and at most times the colloid osmotic pressure difference opposes filtration of solvent from plasma to the interstitium. The proteins are; • • • ်

Albumin, normally around 55% by weight of the total protein. Fibrinogen, a soluble glycoprotein, is about 7%. The globulins are many, and have specific physiological roles to play: Ƚ1-Globulins (Ƚ1-antitrypsin, Ƚ1-antichymotrypsin, Ƚ1lipoprotein, orosomucoid, etc) account for about 5% of the total protein

178

် ် ်

Chapter 9

Ƚ2-Globulins (haptoglobin, ceruloplasmin, protein C, Ƚ2antiplasmin, Ƚ2-lipoprotein, cortisol-binding protein, angiotensinogen etc) are about 9% Ⱦ-Globulins (Ⱦ1-transferin, Ⱦ-lipoprotein, etc) 13% Gamma Globulins (antibodies, etc) 11.0%

Albumin The most abundant soluble extracellular protein molecule in the great majority of humans is albumin, a single polypeptide chain containing 585 amino acids that is unusual among plasma proteins in that it does not contain a carbohydrate moiety. Though it constitutes only 3% of total body protein, around 20% of daily dietary protein requirement is expended on albumin synthesis. Plasma albumin contributes a substantial anionic charge of around 17 mEq l-1, so that proteins have to be taken into account as an anion in quantitative acid base analysis [1]. The negative charge also accounts for the Gibbs-Donnan effect of holding ionised sodium in its field and so enhancing the colloidal osmotic effect of albumin by around 50% [2]. In all, albumin provides around 80% of the colloidal osmotic pressure of plasma. Albumin molecules are susceptible to modifications, which include oxidized, glycated, deamidated and N/C-terminal truncated forms. Ischaemia and the generation of radical oxygen species can alter the ability of the N-terminal region of the albumin molecule to bind transitional metallic ions, such as cobalt, copper, and nickel, and this modification results in the formation of ischaemia-modified albumin. Human nonmercaptalbumin 1 is a form of oxidised albumin found in cirrhotic patients that is capable of triggering a cytokine storm in peripheral blood mononuclear cells [3].

Circulation of albumin Extracellular fluid is in a constant state of circulation between intravascular and extravascular compartments and has to traverse the endothelial glycocalyx, which is semi-permeable to albumin molecules. The presence of albumin within the endothelial glycocalyx is an important determinant of its filter function: the albumin effect. A consequence of the endothelial barrier is to create a higher concentration of albumin upstream of the filter. Plasma is upstream of

Albumin and Other Circulating Proteins

179

the filtration process in most tissues and at most times. Glycocalyx filtrate is almost protein-free in health, but larger solutes like albumin can also diffuse across the endothelial barrier or be transported through endothelial cells by pinocytosis. The concentration of albumin in interstitial fluid is always less than in plasma, but as the interstitial fluid volume is around 4x greater than the intravascular extracellular fluid most of the total body albumin is at any one time within the interstitium. As a rule of thumb, we may estimate the extracellular pool of albumin in an adult human to be around 360 g, of which one third (120 g) is in the intravascular compartment and two thirds (240 g) in the interstitium. Albumin synthesis occurs exclusively in liver cells (hepatocytes) and is sensitive to the oncotic pressure of hepatic interstitial fluid in Disse’s space, which is essentially extravascular plasma. In an adult, the albumin synthesis and clearance rates (turnover) are around 9-15 g per day. The maximal synthesis rate is at least twice the basal rate, so that subtotal liver resection need not compromise the albumin synthesis rate if the remaining hepatocytes are healthy. It is presumed that the albumin synthesis rate is normally responsive to changes in plasma osmolality, but there are clear examples of patient groups with hypoalbuminaemia without compensatory synthesis response. The limited defence of plasma osmolality has led to the teaching aphorism that “the osmostat is readily reset”. Albumin synthesis is stimulated by insulin, thyroxine and cortisol, and inhibited by certain interleukins, most notably IL-6. In most disease states albumin synthesis is found to be normal or raised. Clearance pathways for albumin have been difficult to ascertain, but are mostly catabolic with renal and gastrointestinal losses estimated to be about 5% and 10% respectively. The rate of albumin uptake into cells by caveolation varies with atrial natriuretic peptide blood level, an effect that could be important in plasma volume regulation. Increased renal clearance causes hypoalbuminaemia in nephrosis, while increased intestinal clearance occurs in protein-losing enteropathy and perhaps in other intestinal pathologies. The half-life of extracellular albumin in health is around 17 days and total body turnover time about 25 days, but it is essential to appreciate that the extracellular fluid circulation causes up to 5% of plasma albumin (and about 10% of the plasma solvent) to leave the intravascular compartment every hour, to be returned to central veins in lymph. If there is such a thing as an average albumin molecule, it circulates around the plasma interstitium – lymphatic - plasma pathway once or twice a day,

180

Chapter 9

increasing to three or four times a day when the capillaries are ‘leaky’ to albumin. Intravenously administered albumin has a measured redistribution T1/2 of 12 to 16 h in health, but the volume kinetics of an albumin infusion show the duration of plasma expansion to be much shorter [4]. In clinical conditions in which the transcapillary escape rate for albumin is doubled, for example by major trauma, surgery or sepsis, T1/2 falls to 6-8 h and the plasma expansion effect of an albumin solution is transient. Exogenous albumin will, in the longer term, exacerbate the abnormal distribution of extracellular fluid in trauma, surgery or sepsis by increasing the amount in the interstitial compartment. The volume of distribution of albumin of course varies with the extracellular fluid volume, or state of hydration. Hyperalbuminaemia is an indicator of severe primary dehydration, while the clinical diagnosis of fluid overload is always associated with hypoalbuminaemia. Albumin binding in the tissues also affects the volume of distribution calculation. Albondin, the native albumin transendothelial transporter is an important example of albumin binding [5]. Albumin is distributed throughout the porous zone of the endothelial surface layer that excludes erythrocytes, but it then accumulates at the interface between the porous zone and the inner endothelial glycocalyx layer. This unstirred layer of albumin significantly modifies the observed permeability properties of the microvessel wall [6].

Clinical measurement of albumin Plasma protein concentrations increase with stasis during venepuncture: blood for total protein and albumin measurement should be collected with a minimum of stasis. Clinical laboratories usually measure albumin in plasma or serum by dye-binding methods (colourimetry), for many years with bromocresol green but now moving to bromocresol purple which is held to be more specific and accurate. Albumin binds these dyes with high affinity and the respective complexes absorb light at 628 nm and 600 nm wavelengths. The response is linear up to at least 60 g l-1. The gold standard method is immunochemistry which reveals that bromocresol green overestimates albumin concentration with a mean positive bias. In a general patient population, albumin by green dye is about 5.5 g l-1 higher than albumin by purple dye. The bias increases to 6-8 g l-1 in

Albumin and Other Circulating Proteins

181

hypoalbuminaemic patients, so common in critical care practice. To give an example, a green dye measured albumin 16 g l-1 may be as less than 10 g l-1 by purple dye method. For this reason, there are many clinicians who can claim never to have met a patient with very low plasma albumin! A variety of molecules interfere with albumin measurement: high levels of bilirubin, free haemoglobin, and triglycerides cause positive interference. Penicillin and a uraemic toxin can cause negative interference which manifests as artefactually-lowered albumin values. In the rare, inherited condition, analbuminaemia, plasma albumin is typically 0.25 g l-1 or less. Patients experience sporadic, mild, oedema but are otherwise well. There is a compensatory increase in globulins and other plasma proteins that raises the colloidal osmotic pressure to about 50% of normal. As oedema formation is not inevitable at such low colloid osmotic pressure other factors such as central venous pressure and venous valve competence are evidently able to compensate. Analbuminemics prove that either the functions of albumin are not critical to health or that these functions can be subsumed by other serum proteins. In bisalbuminaemia, another rare, inherited condition, plasma albumin is normal but two species of albumin are present and appear as separate bands on zone electrophoresis of serum. In clinical states associated with systemic oxidative stress and inflammation such as chronic kidney disease, oxidative modifications of serum albumin impair its quantification, resulting in apparent hypoalbuminaemia. In one Report, hypoalbuminaemic patients with the highest oxidative stress had colloid osmotic pressure values that were higher than expected. The contribution to colloid osmotic pressure by other prevalent plasma proteins, such as fibrinogen and immunoglobulins was negligible [7]. Clinical laboratories measure Total Protein in plasma or serum and calculate Globulin concentration as the difference between Total Protein and Albumin. The above considerations therefore are relevant to the accuracy of clinical interpretation of Globulins. Measurements of serum albumin concentration have no diagnostic value for individual conditions except that a low concentration in acute

182

Chapter 9

illness is related to poor prognosis [8]. For this reason, plasma or serum albumin concentration is often included in severity of illness scores such as APACHE [9], Child-Pugh and ALBI scores [10]. Serum albumin declines in so many disease states that it has been called a negative acute phase protein.

Clinical measurement using radiolabelled albumin Radiolabelled albumin is the gold standard for examining the disposition of albumin, but there are major limitations for such studies in trauma, surgery and acute inflammatory states [11]. Radio-iodinated albumin was used to measure transcapillary escape rate in Fleck’s classic paper which showed the median escape rate of albumin for septic patients studied within 24 h of ICU admission was 11.1% (healthy subjects 5.0% per hour). There was no suggestion of an association between the transcapillary escape rate of albumin and tissue oedema [12]. Spiess et al used 125I-labelled albumin to study albumin kinetics in hypoalbuminaemic patients receiving total parenteral nutrition. To assess volume of distribution, time to equilibration, half-lives and catabolic rate blood samples were collected for up to 16 days. Data from this study suggest that the elimination half-life of exogenous albumin is decreased in more severely ill compared to more stable patient populations [13]. As part of their investigations into tissue protein synthesis rates in critically ill patients, Essen et al used 131I-labelled albumin to estimate plasma volume with blood samples collected at 0, 20, 30, 35, and 40 minutes [14]. A similar method was used by Ernest, Belzberg and Dodek in their experiments on septic and cardiac surgical patients, who reported much greater apparent plasma volume expansion after albumin infusion than after an isotonic salt solution [15]. This did not express itself in improved haemodynamic measurements. Albumin labelled with technetium (99mTc-albumin) has been used to investigate extravasation values in patients with renal disorders and lends support to the general hypothesis that oedema is more likely due to disordered capillary permeability than to hypoalbuminaemia [16].

Albumin and Other Circulating Proteins

183

Albumin as a binding protein There are 4 discrete binding sites on an albumin molecule. Drugs competing for the same site may displace one another, for example phenytoin and warfarin. Diazepam binds at a different site and so does not compete with phenytoin or warfarin. Although it has a strong negative charge, albumin binds both positively and negatively charged compounds including hydrophobic organic anions such as bilirubin and long-chain fatty acids and the divalent cations calcium, magnesium and copper. The role of albumin in transporting sphingosine 1-phosphate from erythrocytes to the endothelial permeability autoregulatory system could be an important property [17]. Examples of other compounds bound by albumin are antibiotics, bile acids, copper, zinc, and compounds with specific serum binders such as vitamin D and thyroxin. Albumin binding reduces the free concentration of compounds, thus limiting their biologic activity, distribution, and rate of clearance. While this can be utilised to prolong the duration of action of certain drugs, hypoalbuminaemia can be largely ignored in practical pharmacokinetics.

Albumin as an anti-oxidant Albumin has one exposed free thiol-containing cysteine at position 34, making it the main circulating scavenger of reactive oxygen species. Administration of albumin to patients with sepsis syndrome leads to a sustained increase in plasma thiols. The cysteine antioxidant effect is potentially one of the few beneficial effects of albumin administration in patients with sepsis syndrome [18]. However, trials of intravenous N-acetylcysteine as a therapeutic anti-oxidant in sepsis have been subjected to meta-analysis which casts doubts on its safety and utility as an adjuvant therapy [19]. Another potential approach to boosting the thiol anti-oxidant effect is Vitamin C supplement to increase red cell glutathione [20].

Albumin and capillary permeability In studies of isolated microvessels the absence of plasma proteins from the perfusate increases vascular permeability. This has been called the albumin effect. The understandable interpretation of this observation

184

Chapter 9

had been that albumin contributes to permeability barrier efficiency by electrostatic binding to the endothelial glycocalyx. In an elegant series of experiments, researchers at University of California tested their alternative hypothesis that the albumin effect could be due to the decreased delivery of albumin-bound sphingosine-1-phosphate (S1P) from red blood cells to the microvascular endothelium. They concluded that the albumin effect is mostly due to S1P release and transport from red blood cells [21]. The extent to which hypoalbuminaemia in clinical practice impairs endothelial barrier function is not yet known, but there is no evidence that treating hypoalbuminaemia restores reduced microvascular permeability to albumin. Studies performed by Adam Fleck and his team in Glasgow in the 1980s established that increased microvascular permeability is the major cause of hypoalbuminaemia in critically-ill patients. The intravascular proportion of albumin decreases and the extravascular proportion increases. They measured the transcapillary escape rate of albumin to the tissues (TCERA) and found the normal TCERA to be about 5% of the plasma albumin per hour. TCERA typically doubled during cardiac surgery and values of 20% or more were recorded in septic shock [12]. Infusing albumin to hypoalbuminaemic patients in septic shock did not change the TCERA in a study at the Westminster Hospital, but at the time of measurement the TCERA was only 7% [22]. Lucas and colleagues in Detroit measured “plasma albumin leak” in patients after resuscitation from haemorrhagic shock and found that it was no greater than normal. They concluded that albumin leak is not responsible for reduced plasma colloid osmotic pressure, reduced intravascular albumin, and post-resuscitation weight gain. They hypothesised that low colloid osmotic pressure results from decreased re-entry of albumin into the plasma due to entrapment within the interstitium [23]. At the completion of pancreatectomy, researchers at the Karolinska Institute found a doubling of TCERA, and they observed that plasma albumin remained low after TCERA had returned to normal [24]. The magnitude of the early post-operative change in plasma albumin concentration is consistently found to correlate with inflammatory markers such as C-reactive protein and with adverse outcomes such as complications and length of stay. In one study serum albumin less than 10 g l-1 on post-operative day 1 was associated with a threefold increased incidence of overall postoperative complications and may thus be used to identify patients at risk [25].

Albumin and Other Circulating Proteins

185

In a small series of patients undergoing oesophageal or pancreatic surgery without human albumin infusions, plasma albumin trends were monitored and a fall of around 40% was observed over the first 4-5 hours. Plasma albumin concentration stabilised for the remainder of the three-day study period. Calculations gave an estimate of 24 g albumin moving from plasma to interstitium [26]. In a series of patients undergoing liver transplant who were liberally infused with human albumin solution in order to protect the plasma albumin concentration, it was similarly found that albumin shift from the intravascular compartment lasted about as long as the surgery. Calculations gave an estimate of 40 g albumin leaving the intravascular space.

Albumin in malnutrition Kwashiorkor syndrome is seen in infants and children whose diet is protein deficient. Hypoalbuminaemia is caused by reduced albumin synthesis and responds to increase in dietary protein. There are mixed views on the immediate aetiology of oedema in kwashiorkor, particularly as oedema seems to resolve with nutritional support before hypoalbuminaemia is corrected. After correction for the contribution of oedema fluid to body weight, the plasma volume is usually normal. Hypoalbuminaemia is less frequent in generalised malnutrition states such as anorexia nervosa and voluntary nearstarvation. There are 22 proteinogenic amino acids amongst which arginine and the branched-chain amino acids leucine, isoleucine and valine appear to be particularly rate-limiting. Protein released from tissue turnover and muscle breakdown appears to be sufficient to maintain serum albumin during caloric starvation. In any case, albumin is low in tyrosine and essential amino acids and contributes little to nutritional balance. The possibility that prior malnutrition increases surgical complication rates and risk of mortality led to several trials of preoperative nutritional therapies including parenteral nutrition, but so far with no proven benefit. The Academy of Nutrition and Dietetics have concluded that serum albumin concentration does not consistently or predictably change with weight loss, calorie restriction, or nitrogen balance [20]. Hypoalbuminaemia in chronically ill patients does not seem to be altered by nutritional modifications.

186

Chapter 9

Albumin in hepatic disease The hypoalbuminaemia of liver disease is characterised by reduced albumin synthesis in the absence of dietary protein deficiency and with normal albumin clearance. Cirrhotic patients usually have a normal plasma volume, but plasma volume expansion by albumin infusion is sometimes used in advanced cirrhosis to offset a reduced effective arterial blood volume associated with altered vascular capacitance [27]. It has been demonstrated that human albumin solution could also have a beneficial anti-inflammatory effect in decompensated cirrhosis of the liver [28]. Of course, renal or gastro-intestinal complications and co-morbidities will aggravate hypoalbuminaemia.

Albumin in pregnancy In pregnancy the total body albumin rises while plasma albumin concentration and plasma colloid osmotic pressure fall. Capillary permeability is not measurably reduced, but there is an increase in interstitial fluid volume. Preeclampsia is an expression of widespread endothelial dysfunction with increased capillary permeability to albumin and reduced plasma volume [29].

Albumin in extra-hepatic disease The hypoalbuminaemia of most extra-hepatic disorders is associated with normal or increased albumin synthesis, so redistribution of extracellular albumin to the interstitium and increased albumin clearance are the likely underlying mechanisms. Protein-losing nephropathies are a special case, with losses up to 12 g per day. Plasma volume can be measured with 131I-albumin in patients with nephrotic syndrome: plasma volume and blood volume are found to be normal or increased in most patients [30]. The gallium-transferrin pulmonary leak index can be used as an index of pulmonary permeability, and it has been found to be inversely related to plasma albumin and plasma transferrin concentrations in both septic and non-septic intensive care patients with acute lung injury [31].

Albumin and Other Circulating Proteins

187

The case for albumin therapy? Albumin has a long history of clinical use in biophysical colloid osmotic pressure therapy. It is used in greater volume than any other biopharmaceutical solution, and worldwide manufacturing is of the order of 100s of tonnes per annum [32]. Clinicians rely on the original Starling principle as a reason to transfuse plasma or albumin to preferentially resuscitate the intravascular volume. In some clinical trials the control group receive no other intravenous fluid, in others the control is an isotonic saline solution. It is important to distinguish the J curve context in which a trial is conducted. Induced hypervolaemia with albumin (patients above their J point) will bring longer-lasting plasma volume expansion compared to a similar volume of isotonic salt solution, while resuscitation from hypovolaemia (patients below their J point) is typically achieved with a volume equivalence ratio of 1:1.5. There is concern that hyperoncotic albumin infusions for resuscitation from hypovolaemia could put patients at risk of renal injury while achieving no benefit [33]. Though randomised controlled trials are few, there are data to support the use of human albumin solution in patients with cirrhosis of the liver for the prevention of circulatory dysfunction during large volume paracentesis or spontaneous bacterial peritonitis. Human albumin solution is also used for the treatment of hepato-renal syndrome and hypervolaemic hyponatraemia. It is likely that infused albumin counters harmful effects of the dysfunctional albumins produced by the cirrhotic liver [34]. Longer-term albumin therapy for decompensated cirrhosis prolongs overall survival and might act as a disease-modifying treatment [35, 36]. Renal transplant recipients were pushed into hypervolaemic heart failure (central venous pressure 12-15 mmHg) in the belief this would ensure graft perfusion and early function. 20% human albumin solution did not reduce the volume of 0.9% Sodium Chloride administered and was not associated with better patient outcomes [37]. Preoperative correction of hypoalbuminaemia reduced the postoperative incidence of acute kidney injury in a single-centre controlled trial involving cardiac surgical patients, but there was no reduction in the incidence of severe acute kidney injury, renal replacement therapy, morbidity or mortality [38]. A pilot study of patients with subarachnoid haemorrhage found that higher dosages of 25% human albumin

188

Chapter 9

solution were associated with a lower incidence of transcranial Doppler vasospasm, delayed cerebral ischaemia and cerebral infarction [39]. There are to date no randomised controlled trial data supporting the use of albumin in non-cirrhotic patients with sepsis or other forms of hypovolaemia [40, 41]. To investigate if oedema fluid can be mobilised from the tissues to the blood and excreted as urine, researchers at the Karolinska Institute investigated the haemodilution that follows an acute increase in the plasma colloid osmotic pressure in fifteen healthy euvolaemic volunteers. They increased the plasma colloid osmotic pressure by about 8%, the plasma albumin concentration by about 16% and found that the volume of distribution of red blood cells (their surrogate for plasma volume) was increased by about 16%. Maximal haemodilution was reached within 20 minutes post-infusion but was then halved by 4 hours. On average, they claimed, every ml of infused 20% albumin recruited 3.4 ml of fluid into the red cell dilution volume during the study period [4]. Red cell dilution studies of hyperoncotic human albumin solution transfusion in critically-ill patients have also been interpreted as showing osmotic absorption of fluid from the extravascular to intravascular compartment, but without information from an indicator of the whole intravascular volume, such as Dextran 40, such a conclusion is not justified. An acute increase in circulating plasma colloid osmotic pressure would be expected to draw water from the non-circulating gel phase of the intravascular volume associated with the glycocalyx and the interstitial fluid of the liver. Studies reporting red cell dilution data that do not take into account all the contributors to intravascular volume should be interpreted with caution. Note that autotransfusion by increased osmotic pressure is of limited relevance in clinical resuscitation situations where autotransfusion due to reduced hydrostatic pressure will already have occurred.

Acute Phase Proteins The anti-proteases and pro-coagulants such as C-reactive protein, fibrinogen, alpha 1-antitrypsin and complement C3 are increased in the so-called stress response. The plasma concentrations of constitutive

Albumin and Other Circulating Proteins

189

proteins such as albumin and transferrin are initially decreased, leading to the description of these proteins as negative acute phase proteins. This terminology reflects the venerable ebb and flow phases (catabolism followed by anabolism) following injury, described by Cuthbertson in 1942. One of the more recent proposals for post-injury metabolic staging is: • • •

ischaemia/reperfusion phenotype the leukocytic phenotype the angiogenic phenotype [42].

It remains debatable whether a concept of "metabolic staging" in guidance and decision making for the support of severely injured patients could be of benefit [43].

References 1. Margarson MP, Soni N. Serum albumin: touchstone or totem. Anaesthesia. 1998;53:789-803. 2. Levitt DG, Levitt MD. Human serum albumin homeostasis: a new look at the roles of synthesis, catabolism, renal and gastrointestinal excretion, and the clinical value of serum albumin measurements. Int J Gen Med. 2016;9:229-255. 3. Alcaraz-Quiles J, Casulleras M, Oettl K et al. Oxidized Albumin Triggers a Cytokine Storm in Leukocytes Through P38 MitogenActivated Protein Kinase: Role in Systemic Inflammation in Decompensated Cirrhosis. Hepatology. 2018;68:1937-1952. 4. Zdolsek M, Hahn RG, Zdolsek JH. Recruitment of extravascular fluid by hyperoncotic albumin. Acta Anaesthesiol Scand. 2018;62:12551260. 5. Merlot AM, Kalinowski DS, Richardson DR. Unraveling the mysteries of serum albumin-more than just a serum protein. Front Physiol. 2014;5:299. 6. Curry FE, Michel CC. The endothelial glycocalyx: Barrier functions versus red cell hemodynamics: A model of steady state ultrafiltration through a bi-layer formed by a porous outer layer and more selective membrane-associated inner layer. Biorheology. 2019

190

Chapter 9

7. Michelis R, Sela S, Zeitun T, Geron R, Kristal B. Unexpected Normal Colloid Osmotic Pressure in Clinical States with Low Serum Albumin. PLoS One. 2016;11:e0159839. 8. Nicholson JP, Wolmarans MR, Park GR. The role of albumin in critical illness. Br J Anaesth. 2000;85:599-610. 9. Knaus WA. APACHE 1978-2001: the development of a quality assurance system based on prognosis: milestones and personal reflections. Arch Surg. 2002;137:37-41. 10. Wang YY, Zhong JH, Su ZY et al. Albumin-bilirubin versus Child-Pugh score as a predictor of outcome after liver resection for hepatocellular carcinoma. Br J Surg. 2016;103:725-734. 11. Erstad BL. Albumin disposition in critically Ill patients. J Clin Pharm Ther. 2018;43:746-751. 12. Fleck A, Raines G, Hawker F et al. Increased vascular permeability: a major cause of hypoalbuminaemia in disease and injury. Lancet. 1985;1:781-784. 13. Spiess A, Mikalunas V, Carlson S, Zimmer M, Craig RM. Albumin kinetics in hypoalbuminemic patients receiving total parenteral nutrition. JPEN J Parenter Enteral Nutr. 1996;20:424-428. 14. Essén P, McNurlan MA, Gamrin L et al. Tissue protein synthesis rates in critically ill patients. Crit Care Med. 1998;26:92-100. 15. Ernest D, Belzberg AS, Dodek PM. Distribution of normal saline and 5% albumin infusions in cardiac surgical patients. Crit Care Med. 2001;29:2299-2302. 16. Rostoker G, Behar A, Lagrue G. Vascular hyperpermeability in nephrotic edema. Nephron. 2000;85:194-200. 17. Zhang L, Zeng M, Fan J, Tarbell JM, Curry FR, Fu BM. Sphingosine-1phosphate Maintains Normal Vascular Permeability by Preserving Endothelial Surface Glycocalyx in Intact Microvessels. Microcirculation. 2016;23:301-310. 18. Quinlan GJ, Margarson MP, Mumby S, Evans TW, Gutteridge JM. Administration of albumin to patients with sepsis syndrome: a possible beneficial role in plasma thiol repletion. Clin Sci (Lond). 1998;95:459-465. 19. Szakmany T, Hauser B, Radermacher P. N-acetylcysteine for sepsis and systemic inflammatory response in adults. Cochrane Database Syst Rev. 2012CD006616. 20. White JV, Guenter P, Jensen G et al. Consensus statement: Academy of Nutrition and Dietetics and American Society for Parenteral and Enteral Nutrition: characteristics recommended for the identification

Albumin and Other Circulating Proteins

191

and documentation of adult malnutrition (undernutrition). JPEN J Parenter Enteral Nutr. 2012;36:275-283. 21. Adamson RH, Clark JF, Radeva M, Kheirolomoom A, Ferrara KW, Curry FE. Albumin modulates S1P delivery from red blood cells in perfused microvessels: mechanism of the protein effect. Am J Physiol Heart Circ Physiol. 2014;306:H1011-7. 22. Margarson MP, Soni NC. Effects of albumin supplementation on microvascular permeability in septic patients. J Appl Physiol (1985). 2002;92:2139-2145. 23. Lucas CE, Benishek DJ, Ledgerwood AM. Reduced oncotic pressure after shock: a proposed mechanism. Arch Surg. 1982;117:675-679. 24. Komáromi A, Estenberg U, Hammarqvist F, Rooyackers O, Wernerman J, Norberg Å. Simultaneous assessment of the synthesis rate and transcapillary escape rate of albumin in inflammation and surgery. Crit Care. 2016;20:370. 25. Labgaa I, Joliat GR, Kefleyesus A et al. Is postoperative decrease of serum albumin an early predictor of complications after major abdominal surgery? A prospective cohort study in a European centre. BMJ Open. 2017;7:e013966. 26. Norberg Å, Rooyackers O, Segersvärd R, Wernerman J. Leakage of albumin in major abdominal surgery. Crit Care. 2016;20:113. 27. Brinch K, Møller S, Bendtsen F, Becker U, Henriksen JH. Plasma volume expansion by albumin in cirrhosis. Relation to blood volume distribution, arterial compliance and severity of disease. J Hepatol. 2003;39:24-31. 28. Fernández J, Clària J, Amorós A et al. Effects of Albumin Treatment on Systemic and Portal Hemodynamics and Systemic Inflammation in Patients With Decompensated Cirrhosis. Gastroenterology. 2019 29. Neal CR, Hunter AJ, Harper SJ, Soothill PW, Bates DO. Plasma from women with severe pre-eclampsia increases microvascular permeability in an animal model in vivo. Clin Sci (Lond). 2004;107:399-405. 30. Geers AB, Koomans HA, Boer P, Dorhout Mees EJ. Plasma and blood volumes in patients with the nephrotic syndrome. Nephron. 1984;38:170-173. 31. Aman J, van der Heijden M, van Lingen A et al. Plasma protein levels are markers of pulmonary vascular permeability and degree of lung injury in critically ill patients with or at risk for acute lung injury/acute respiratory distress syndrome. Crit Care Med. 2011;39:89-97.

192

Chapter 9

32. Matejtschuk P, Dash CH, Gascoigne EW. Production of human albumin solution: a continually developing colloid. Br J Anaesth. 2000;85:887-895. 33. Udeh CI, You J, Wanek MR et al. Acute kidney injury in postoperative shock: is hyperoncotic albumin administration an unrecognized resuscitation risk factor. Perioper Med (Lond). 2018;7:29. 34. Valerio C, Theocharidou E, Davenport A, Agarwal B. Human albumin solution for patients with cirrhosis and acute on chronic liver failure: Beyond simple volume expansion. World J Hepatol. 2016;8:345-354. 35. Caraceni P, Riggio O, Angeli P et al. Long-term albumin administration in decompensated cirrhosis (ANSWER): an openlabel randomised trial. Lancet. 2018;391:2417-2429. 36. Di Pascoli M, Fasolato S, Piano S, Bolognesi M, Angeli P. Long-term administration of human albumin improves survival in patients with cirrhosis and refractory ascites. Liver Int. 2019;39:98-105. 37. Shah RB, Shah VR, Butala BP, Parikh GP. Effect of intraoperative human albumin on early graft function in renal transplantation. Saudi J Kidney Dis Transpl. 2014;25:1148-1153. 38. Lee EH, Kim WJ, Kim JY et al. Effect of Exogenous Albumin on the Incidence of Postoperative Acute Kidney Injury in Patients Undergoing Off-pump Coronary Artery Bypass Surgery with a Preoperative Albumin Level of Less Than 4.0 g/dl. Anesthesiology. 2016;124:1001-1011. 39. Suarez JI, Martin RH, Calvillo E, Bershad EM, Venkatasubba Rao CP. Effect of human albumin on TCD vasospasm, DCI, and cerebral infarction in subarachnoid hemorrhage: the ALISAH study. Acta Neurochir Suppl. 2015;120:287-290. 40. Albumin Reviewers (Alderson P BF, Li Wan Po A, Li L, Blackhall K, Roberts I, Schierhout G. Human albumin solution for resuscitation and volume expansion in critically ill patients. Cochrane Database Syst Rev. 2011CD001208. 41. Patel A, Laffan MA, Waheed U, Brett SJ. Randomised trials of human albumin for adults with sepsis: systematic review and meta-analysis with trial sequential analysis of all-cause mortality. BMJ. 2014;349:g4561. 42. Aller MA, Arias JI, Alonso-Poza A, Arias J. A review of metabolic staging in severely injured patients. Scand J Trauma Resusc Emerg Med. 2010;18:27. 43. “Metabolic staging” after major trauma - a guide for clinical decision making [editorial]. Scand J Trauma Resusc Emerg Med 2010;18:34.

CHAPTER 10 A HAEMODYNAMIC PARADIGM

Chapter summary The circulation of blood is in a complex dynamic equilibrium that defies a simple narrative explanation, which explains why there are several textbook narrative explanations and not all are simple. In this Chapter I review some of the more influential viewpoints and attempt to draw them into a perspective that respects the steady-state Starling principle.

Historic The most venerable statement about the circulation of blood is that of the discoverer William Harvey, who in 1628 wrote: “It is absolutely necessary to conclude that the blood is in a state of ceaseless motion; that this is the function which the heart performs by means of its pulse; and that this is the sole and only end of the motion and contraction of the heart.”

Thomas Young may have been The Last Man Who Knew Everything. In 1809 he wrote: “When the quantity of the blood transmitted by the heart is smaller than in health, the arteries must be contracted . . . and the veins must of course become distended . . . until the blood, which is accumulated in the veins, has sufficient power to urge the heart to a greater exertion . . .”

After careful experimentation, Patterson, Piper and Ernest Starling described The Regulation of the Heart Beat in 1914. In their own words, the rule we know as Starling’s Law of the Heart is stated as follows:

194

Chapter 10

“Within physiological limits, the energy of myocardial contraction is directly proportional to end-diastolic ventricular myocardial fibre length.”

Otto Frank had made complimentary observations on frog hearts in the 18th century, and it is now customary to describe the ability of the heart to change its energy of contraction and therefore stroke volume in response to changes in the volume of blood presented to the diastolic heart as the Frank-Starling mechanism. In canine studies on venous and capillary pressures Bayliss and Starling had reported the asystolic vascular pressure throughout the vascular loop to be 5 to 10 mmHg. The vascular pressure that exists during asystole and after redistribution of blood is now called the mean circulatory filling pressure (Pmcf). It is said to reflect the “fullness” of the circulatory system, but also provides an index of change in venous smooth muscle tone while the blood volume is unchanged. To be more anatomically precise, the Pmcf is an estimate of the distending pressure in the venules and small veins, which contain most of the blood in the body and comprise most of the vascular compliance [1]. Arthur Guyton applied the principles of engineering and systems analysis to develop a permissive heart concept of cardiovascular physiology incorporating these principles. He developed experimental models in his laboratory in Oxford, Mississippi, to construct cardiac output curves (FrankStarling curves) that could be superimposed on venous return curves to create the famous Guyton diagrams [2]. The points at which the curves intersect solve a pair of simultaneous equations defining the interaction of heart and circulation at any one point in time. At zero cardiac output (and so zero venous return) the right atrial pressure (Pra) becomes equal to what Guyton called the mean systemic pressure (Pms). In this view the pressure difference for venous return at all levels of cardiac output is claimed to be between the upstream Pms and the downstream Pra; when the pressure difference becomes zero, so does venous return. George Brengellman critically appraised Guyton’s experiments and challenged the implication that there is in effect a venous reservoir with its pressure fixed at Pms for any given intravascular volume and vascular compliance state. Modern controversies about cardiovascular regulation largely revolve around such criticisms of the Guyton diagrams and the interpretation of his laboratory findings. Here are some examples.

A haemodynamic paradigm

195

In a Point-Counterpoint debate Sheldon Magder argued the case that “The heart cannot pump out more than the flow that is determined by the drainage characteristics of the circuit. The heart provides the ‘restorative force’ and, as per Guyton, plays a ‘permissive’ role.” [3]

Brengellman rejected what he called the (PmsΫPra)/venous resistance concept advocated by Magder and counter-argued that the work manifested in pressure-volume changes for the ventricles is what makes the blood go round [4]. It is sometimes argued that Guyton mis-characterized the Starling Law of the Heart by swapping the axis parameters for Starling’s data and making it appear that central venous pressure (CVP) determined the circulatory rate of blood, rather than the other way around. He made CVP appear to be a cause of blood flow rather than the effect of it, and the widespread practice of infusing fluid to elevate CVP in order to distend the right ventricle was given undeserved legitimacy [5, 6]. Researchers in Paris performed passive leg raising volume challenges on patients with septic shock on a higher or a lower dose of norepinephrine [7]. By varying positive pressure pulmonary ventilation to raise or lower central venous pressure and measuring the resultant cardiac output they extrapolated data points to the hydrostatic pressure at zero blood flow, Guyton’s Pms. They calculated the inverse of the slope of the regression line and called it “resistance to venous return”. They concluded: “In septic shock patients, decreasing the dose of norepinephrine decreased the mean systemic pressure and, to a lesser extent, the resistance to venous return. As a result, venous return decreased.”

Correspondents argued that this was just “Good old physiology in a modern jacket” [8] and that an increase in unstressed blood volume with decreased norepinephrine infusion rate would explain the results [9]. Carl Rothe observed that the venules and small veins change little in diameter when the circulation of blood ceases. He therefore proposed that they are at a pivotal pressure for haemodynamics. He suggests that when the cardiac pump fails, arterial pressure declines while central venous pressure rises, and venular pressure is the pivot point [1]. Rothe goes on to develop the argument that this pivot pressure is the Pmcf.

196

Chapter 10

The currently-prevailing paradigm In current clinical textbooks and journals the prevailing paradigm is that volume resuscitation protocols are used to increase Pms, with the intention of increasing the Pms - Pra pressure difference for venous return, thus increasing cardiac output. Pms has been described as the upstream pressure that drives blood into the heart [10]. I shall argue in this chapter that Pms, or more specifically the venular filling pressure, is the upstream pressure that ensures an adequate venous excess volume for complete right ventricular filling during diastole. Either way, there is increasing acceptance that CVP is not a helpful clinical index of ventricular preload. Changes in vascular capacitance or compliance induced by reflexes, hormones, or drugs have physiological effects similar to a rapid change in blood volume, and thus also strongly influence venous return and cardiac output [1]. A cardiac surgeon can of course measure the no-flow circulatory pressure directly after clamping the aortic root and the great veins at point of entry to right atrium. Less invasive techniques have been developed for the determination of a zero-flow pressure value. There is a patented software for the bedside estimation of mean systemic filling pressure analogue (Pmsa) which applies a simplified electric circuit analogy and takes input data from changing central venous pressure, arterial pressure and cardiac output during positive-pressure ventilation [11]. The software program is called NavigatorTM. Evidence so far supports the expectation that Pmsa rises with passive leg raising or fluid challenging, and it is suggested that it’s measurement is of greater clinical value than the dichotomous fluid responder or nonresponder approach [12–14]. The inspiratory-hold method for estimating Pms in sedated and positive-pressure ventilated patients applies four 12 second inspiratory holds at airway pressures of 5, 15, 25 and 35 cm H2O. Central venous pressure and cardiac output are measured during the plateau phase, and the zero blood flow intercept of the pressure and flow data is taken to be an estimate of Pms [15]. The cardiac surgeon Robert Anderson described a technique for clinical determination of the mean circulatory pressure in his Textbook [16]: “A useful approximation of the mean cardiovascular pressure can be made without the need of stopping the heart and letting the pressure

A haemodynamic paradigm

197

equalize in the entire cardiovascular system. The assumption is made that the same pressure would be found by isolating a representative sample of the arteries, capillaries, and veins at ventricular diastole, and letting the pressure equilibrate in that sample. By instantaneously interrupting arterial inflow to the arm and venous outflow from it, the pressure will fall in the arteries and rise in the veins until they are equal. This equalized pressure, which will occur within 30 seconds, approximates that found when circulation is stopped in the entire body.”

In his description of the method, Anderson believed it to be important to use a narrow arm cuff to minimise displacement of blood by the inflating cuff. More data was provided about the precision of this technique (the Transient stop-flow arm arterial-venous equilibrium pressure) which is claimed to approximate Pms. The coefficient of error was about 5%, and the least significant change 14%. Performing two or three measurements and taking their average improved precision further [17]. Using only a radial artery catheter to determine stop-flow arm equilibrium pressure seems to be feasible [18]. In a comparison of three methods to assess Pms, it was concluded that the equilibrium pressure in the arm during stop-flow (Parm) and inspiratory-hold manoeuvre-derived Pms values are interchangeable in mechanically ventilated postoperative cardiac surgery patients [13]. Typically, the euvolaemic sedated and ventilated patient has a Pms of around 20 mmHg. Navigator gave lower estimates of Pmsa around 14 mmHg, but with a consistent offset of about -6 mmHg all three techniques seemed to track the effective circulatory volume. The venular filling pressure can be simply, directly and non-invasively measured by coextensive bioimpedance plethysmography which is discussed later in this chapter. If Rothe was correct that the venules are the site of the Pmcf and are at the circulatory pivot point then the preceding indirect and invasive techniques are redundant. Moreover, it is likely that they underestimate the Pms.

Cardiomythology There are problems with the preceding account. Venous return, considered as a flow, is a consequence and not a cause of cardiac output: it cannot in itself be controlling the heart. J. Rodney Levick (St Georges School of Medicine, London) explains:

198

Chapter 10

“Students (and some professionals!) sometimes get into an awful muddle by trying to explain the control of cardiac output in terms of ‘venous return’ rather than filling pressure. ‘Venous return’ is the flow of blood into the right atrium. When the circulation is in a steady-state the venous return equals the cardiac output because the circulation is a closed system of tubes; any inequality can only be transient. Under steady-state conditions the venous return is simply the cardiac output observed in veins rather than arteries. The popular notion that the venous return controls the cardiac output is an unhelpful, literally circular viewpoint, because venous return depends on the cardiac output.” [19]

Roger Carpenter at The University of Cambridge complained: “Most of us who have to teach cardiovascular control will probably share this concern and probably also agree about the origin of the muddle; the conventional explanation of the circulation that assigns fictitious roles to vaguely defined variables. The main offender is ‘venous return’.” [20]

Mean circulatory filling pressure The second great myth is Pmcf in its various guises. One certainly can divide the asystolic intravascular pressure by the estimated blood volume and call it vascular compliance, but to presume that that number has any relevance to understanding the complex dynamic equilibrium of the heart and its vascular loops is a leap of faith. We have already noted that peripheral venules and small veins contain around half of the blood in the body while it is flowing, and they account for most of the total vascular compliance. It makes more sense to have a paradigm that is based on the individual pressure-volume characteristics of the component segments of the vascular loop in life than to try to read something into their redistributed average in death. Ideally, we would wish to use a simple non-invasive technique for measuring venular and small vein compliance, the venular filling pressure and the central venous pressure in order to estimate the post-capillary pressure gradient. Whatever the finer points of physiological philosophy, the clinical practitioner needs a working paradigm built on definable and quantifiable physiological parameters that guides her unerringly to make rational therapeutic decisions.

A haemodynamic paradigm

199

Venous excess The late Roger Carpenter was a Cambridge University neurophysiologist who used saccadic movements of the eyes to study human decision making. The single most frequent decision that we make is that of where to look, a decision that we make two or three times a second throughout our waking lives. Professor Carpenter also taught cardiovascular physiology to undergraduates and brought fresh insights to how we can most clearly teach the principles of autoregulation of the circulation of blood [20]. We have known since Starling’s experiments that the output of the heart is greatly influenced by extracardiac factors, and in particular factors associated with the venules, small veins and large veins. The Frank-Starling mechanism serves to prevent accumulation of blood within the venous system by ensuring that the ventricle expels the blood delivered to it. At every stroke there will be a degree of variance between the volumes of blood ejected and returned, but over a moving average these volumes have to be equalised. A Guyton diagram illustrates how, for each pair of venous return and cardiac output curves, there is an intersection at a central venous pressure that simultaneously impedes venous return and drives cardiac output to be equal. At steady-state, the net variance every minute is zero. If the next stroke ejects 75 ml from the left ventricle, while in the same time 74 ml of blood enters the venous system, the venous volume is decreased by 1 ml. Alternatively, if 76 ml of blood enters the venous system, the venous volume is increased by 1 ml. There has to be a regulatory mechanism to balance the variances, a feedback control to stabilise the circulation of blood. If a positive (e.g. by infusion of fluid) or negative (e.g. by phlebotomy) venous accumulation rate continues for 60 heart beats over a full minute the venous accumulation rate is +60 or -60 ml min-1. The integral of the venous accumulation rate is a volume, the venous excess. This volume has two roles: it is the volume available in the large and central veins to allow the relaxing right ventricle (assisted by a contracting right atrium) to fill with near-zero resistance, and it is the volume that distends the venules and small veins to account for the pressure difference that maintains near-constant post-capillary blood flow to the central veins. The Carpenterian model also features the concept of arterial excess, a much smaller volume that stretches the aorta, large arteries and small arteries balancing the volume ejected from the ventricle at each stroke and the volume that leaves the

200

Chapter 10

arteriolar segment to perfuse the capillary beds. The Carpenterian model distinguishes the venous vascular loop segments as ‘collective’ and the arterial vascular loop segments as ‘distributive’. The venous and arterial sides of the peripheral vascular loop have different pressure/ volume characteristics and relatively independent control mechanisms. While arterial-side regulation is predominantly about resistance to flow and maintenance of a perfusion pressure serving the capillary bed, venous-side regulation concerns capacitance of volume. Their account of the arterial-side control mechanism is traditional and familiar. It preserves homeostasis by negative feedback through the baroreceptor reflexes. Arterial blood pressure is the error signal that represents a mismatch between the supply of blood (by the cardiac output) and the demand for blood by tissues. Arterial excess and peripheral resistance are the balancing factors for arterial pressure. For example, when there is a rise in demand from muscle activity, tissue metabolites cause peripheral resistance to fall and the volume of blood leaving the arterial side to increase. The negative arterial accumulation rate means that arterial excess volume drops. For both of these reasons, arterial pressure will fall, and the baroreceptors will respond by increasing heart rate and by increasing peripheral resistance in vascular beds that are not experiencing increased demand. The physiology purist may argue that there are many details of this system that are complex and not wholly understood, but the paradigm of an underlying negative feedback control system explains how arterial pressure is controlled and why it has to be controlled: a drop in blood pressure indicates that supply is not keeping up with demand. Reddi and Carpenter’s venous-side regulation model attempts to avoid the confusion we risk when we teach nebulous concepts like venous return and mean systemic pressure. Venous excess is the venous-side error signal that represents a mismatch between the inflow of blood from the microcirculation and the outflow of blood by efficient ventricular contraction and relaxation. Notice again that while a key arterial-side regulator is resistance to flow, the equivalent venous-side regulator is capacitance for volume which assists flow. Both are associated with vascular tone; on the venous-side, increased tone reduces compliance more than it increases resistance, and on the arterial side increased tone increases peripheral resistance more than

A haemodynamic paradigm

201

it reduces compliance. Venous resistance and arterial compliance exist, but are relatively insignificant. To illustrate by the previous example, as the arterial-side control mechanism increases cardiac output in response to increased demand, the venous excess increases and assists the increased circulatory rate. The central venous pressure increases if the heart is failing to match output and input, but otherwise an appropriately increased circulatory rate to match increased demand is sustained at near-normal arterial and central venous pressures. Now let us consider autoregulation in response to a moderate blood volume reduction; a phlebotomy or transient haemorrhage of around 500 ml. On the venous side, the venous accumulation rate becomes negative, reducing the venous excess volume. This causes reduced central venous pressure, reduced stroke volume and reduced cardiac output. The arterial-side mechanism then comes into play, with reduced arterial excess leading to reduced arterial pressure which is defended by reflex increases in heart rate and peripheral resistance. The baroreflex arousal also reduces venous capacitance (the volume/ pressure ratio) and stabilises the venous excess volume. A new steadystate with protected arterial and central venous pressures is achieved. In a Carpenterian view, volume resuscitation protocols are used to restore the volume we call venous excess and which is accessible by non-invasive clinical examination. A danger of the Carpenterian paradigm is that it supports confidence in central venous/right atrial pressure targets as determinants of stroke volume in the non-failing heart; the therapeutic focus must rather be on venous excess volume. In the words of Michael Pinsky: “The primary role of the right ventricle is to deliver all the blood it receives per beat into the pulmonary circulation without causing right atrial pressure to rise.” [21]

The failing heart is altogether different. A quantitative or even semiquantitative measure of congestion in heart failure is very desirable, and the jugular venous pulse height (JVP) is the most useful physical finding for determining a patient's volume status. Not only does an elevated JVP detect systemic congestion, but there is good sensitivity (70%) and specificity (79%) between JVP and elevated leftsided filling pressure [22]. We have spoken of autoregulation “within the physiological limits”. Acute extreme disturbances to the cardiovascular equilibrium produce

202

Chapter 10

disequilibrium states which are outwith physiological limits and demand therapeutic intervention. As Shakespeare’s Claudius puts it in the Danish play: “Diseases desperate grown, by desperate appliance are relieved, Or not at all.”

Myocardial energy excess Can we be more specific about the limits of the “physiological” range? Frank and Starling had talked about preload (the stretch of ventricular muscles) determining the energy of myocardial contraction. The concept of myocardial energy excess was explained by Robert M. Anderson, a cardiac surgeon in Arizona, in 1983 [23]. He constructed a simple hydraulic model of the human circulation to examine the underlying engineering principles of cardiovascular physiology. There is a 15 minute video in which Dr Anderson demonstrates his model in a tutorial style, a transcript of the video, and a Textbook [16] which you can download in paperless formats at no cost from http://cardiacoutput.info. The electronic editions include a tribute to Dr Anderson who deceased in 2010. With few peer-reviewed papers it is unsurprising that Anderson’s contribution is overlooked by citationfocussed academics. Simple hydraulic engineering does not use the concepts of venous return or cardiac output. Anderson more specifically referred to the circulatory rate of blood which we may call Qt. He acknowledged that the model lacks many of the properties of fully-integrated homeostatic physiology with, for example, baroreflexes, neurohumoral adaptations and tapered/branching arterial and venous side vasculature, but as he observed: “The close correlation of model findings to physiological observations in man makes the model useful in understanding how the cardiovascular system works, and helps to anticipate cause and effect in human physiology.”

Anderson argued that a valid clinical paradigm of cardiovascular function must explain observed facts. He listed a number of examples that challenge previous paradigms and are revealed in cardiac surgical practice.

A haemodynamic paradigm

•

• • • • • •

203

There is volume equilibrium between the systemic and pulmonary circuits even when there are massive congenital shunts between the two, and equilibrium remains after closure of the shunts. The empty heart continues to beat strongly in the absence of any diastolic filling or stretching of the ventricles. Increasing pacemaker rate above that necessary to prevent failure does not increase Qt. Slowing heart rate that is above that necessary to prevent failure and reducing ventricular contractility by beta blockade does not reduce Qt. Ventricular transmural pressures measured during heart catheterization are always above zero. After heart transplantation, without nerve supply to the heart or artificial pacing, Qt and pulmonary/systemic blood volume balance remain normal. An increase in arterial resistance does not reduce Qt in the nonfailing cardiovascular system.

Anderson listed Ten Unique Characteristics of the healthy intact cardiovascular system, which I summarise here. 1) The system is circular. 2) The system is elastic. 3) The system is filled with blood at a mean cardiovascular pressure which is independent of the pumping action of the heart. 4) The right and left ventricles are in series with interposed systemic and pulmonary vascular beds. 5) The heart fills passively: it is a non-sucking pump. 6) The rate of blood circulation is regulated by extra-cardiac vascular factors. 7) Flow from the heart is intermittent, while the flow to it is continuous. 8) Excess expenditure of energy by the heart is needed for the circulation rate imposed by vascular regulators which we shall call myocardial energy excess. 9) Ventricular capacity is in excess of the diastolic filling volume: the ventricular capacity excess. 10) Venous-side vascular resistance and venous compliance characteristics determine Qt while arterial-side vascular

204

Chapter 10

resistance regulates the arterial pressure. Anderson’s term for the combined effects of venous-side vascular resistance and venous compliance characteristics is venous impedance. As a consequence, we may assert in Andersonian terms that “in health the circulatory rate of blood varies directly with the mean cardiovascular pressure and inversely with venous impedance.”

I must re-iterate that Anderson’s venous impedance must not be confused with venous resistance (pressure-flow characteristic) alone; venous compliance (pressure-volume characteristic) is an important contributor to Anderson’s venous impedance concept. Anderson delineated physiology from pathophysiology on the basis of myocardial energy excess. Normal physiology principles apply only when there is a positive myocardial energy excess. Myocardial energy failure occurs when the pump does not provide enough energy to prevent ventricles filling completely before end diastole. Note that it is important not to confuse myocardial energy failure with other definitions of heart failure. In bradycardia, Qt is limited to the maximum ventricular volume multiplied by the heart rate. At around 50-60 beats per minute the accumulation of myocardial energy excess can begin and ventricular capacity excess appears. Further increases in rate or contractility cause no further increase in Qt. It is only during the bradycardic state of myocardial energy failure that heart rate is a direct determinant of Qt. In tachyarrhythmic heart failure the maximum ventricular volume is fixed by the very short diastolic filling time and Qt is low. Qt can only be improved by slowing the heart rate so that ventricular filling is no longer limited by diastolic time. Another state of myocardial energy failure is ventricular contractility impairment. Ventricles are filled to their maximum (i.e. they have no ventricular capacity excess) when impaired ventricular contractility cannot deliver a myocardial energy excess after overcoming ventricular outflow impedance. Stroke volume is fixed at its maximum, and heart rate is the only regulator of Qt. Exemplar clinical conditions include myocarditis, infarction and aortic outflow obstruction. Anderson gives a thoughtprovoking definition of heart failure: “Heart failure exists whenever heart function is inadequate to produce the circulatory flow rate which would be allowed by the mean

A haemodynamic paradigm

205

cardiovascular pressure and inlet impedance. “

We may now say that in heart failure Qt varies directly with heart rate and inversely with the state of ventricular contractility relative to the ventricular outflow impedance. Qt therefore has two different sets of determinants, of which only one set can be operative at any one time. An important challenge to Anderson’s hydraulic engineering account is the response of Qt to tissue oxygen demand as explained above by Carpenter’s paradigm. If, normally, cardiac output is not dependent on arterial-side resistance or heart rate, how do we explain increased Qt in response to reduced tissue oxygen tension? The answer comes from Braunwald’s canine experiments published in 1962. Systemic hypoxia actually results in striking venoconstriction, not vasodilation, mediated by chemoreceptor reflexes and so increasing Carpenter’s venous excess and Anderson’s mean circulatory pressure. The baroreceptor reflex secondarily reduces arteriolar resistance in order to preserve arterial pressure. Thus, Anderson’s and Carpenter’s paradigms can be reconciled.

Effective versus ineffective blood volumes? Concepts such as effective arterial blood volume or effective circulating blood volume, and stressed blood volumes are used to convey the idea that the active redistribution of blood between vascular loops and the several segments of the vascular loop with their unique pressure/volume characteristics is an important regulator of cardiovascular function. Within physiological limits, the ability to redistribute the intravascular volume between segments is claimed to be the primary means by which venous return and cardiac output are regulated in response to varying metabolic demands. The following thought experiment is based on one entertained by Bayliss and Starling in 1894. If you could start with an entirely avolaemic cardiovascular system (no blood and no venous return), then start infusing blood, venous, arterial and intracardiac pressures would remain the same as interstitial pressure until around two thirds of the normal blood volume was present. Beyond that point, the Pmsf would increase as the blood volume further increases. The now-distended vessels would stretch with increasing intravascular volume, instead of just altering their shape.

206

Chapter 10

The blood volume has exceeded what we could call the unstressed blood volume and further volume infusion boosts the stressed blood volume, causing the Pms to increase. It is now possible for a distended ventricle to contract, generating a stroke volume and cardiac output. These in turn generate a venous return proportional to the Pms, or more specifically to the difference between Pms and Pra so that an autoregulated circulation of blood can be established. Only small further volumes of fluid should be needed to substantially vary the stressed volume and Pms. Once the circulation of blood is established, there is an unstressed volume within larger veins (transmural pressure difference close to zero) while there is a stressed volume within the higher pressured segments of the peripheral vascular loop. Of course, rather than a binary stressed - unstressed division there are in reality at least seven segments of the peripheral vascular loop, plus portal circulations and the pulmonary vascular loop. Blood within the liver is believed to be a major unstressed reservoir. The distinction of stressed to unstressed volume is a poorly-defined and hypothetical concept because in reality all blood moves through stressed and unstressed vascular loop segments. In a Point: Counterpoint debate, Rothe argued that active venoconstriction is important in maintaining or raising enddiastolic volume and stroke volume [24] while Hainsworth challenged the view that the autonomic control of veins provides a major mechanism for maintaining cardiovascular homeostasis in humans [25]. The first reported human experiments were on anaesthetised cardiac surgical patients on cardiopulmonary bypass [26]. Two peripheral vascular compartments were distinguished by their emptying time constants after a step change in central venous pressure. When the experiment was performed with an epinephrine (adrenaline) infusion there was redistribution of blood volume from the longer to the shorter time constant compartments. Separate experiments suggested that the compartments were mesenteric vascular loops (longer time constant) and muscle bed (shorter time constant) vascular loops. The peripheral vascular actions of epinephrine infusion were said to account fully for the effect of epinephrine on cardiac output. Increased stressed volume may contribute to the hydrostatic pulmonary oedema associated with massive intracranial haemorrhage and chaotic autonomic nervous system discharge. Unstressed volume is said to increase in the vasoplegia of severe sepsis, contributing to relative hypovolaemia.

A haemodynamic paradigm

207

Describing pressure-volume characteristics If the vascular loops were rigid (capacitance and compliance = 0) there could be no circulatory flow. In the absence of a compliant circuit the pump can neither accept fluid from nor deliver its content to the vascular loop. Arterial elastance allows the energy supplied by the pump to move fluid in the direction allowed by the valves, distending the compliant segments which then release stored energy to propel the fluid past the arteriolar resistance while the pump refills. The pressure and volume within the arterial segments of the vascular loop is pulsatile. The capillary bed drains into the venules which are distended and under pressure. The heart creates circulatory flow by lowering the right atrial pressure and allowing the recoil pressure in veins and venules (Pmcf) to energise the flow of blood back to the heart [27]. Compliance (distensibility) is the ratio of increasing segmental intravascular volume to increasing transmural pressure difference. It describes the slope of the volume/ pressure curve, or dV/dP, at any particular volume or pressure. Blood vessels that assist flow by storing potential energy in their stretched walls operate on the steep part of their sigmoid compliance curve. The arterial-side segments and the venules and small veins store and release potential energy to assist blood flow. The reciprocal of compliance is elastance, dP/dV, a term often used when describing the pressure volume characteristics of the aorta and large arteries and their assistive effect on blood flow. Vascular capacitance (volume at a given venous pressure) is a term borrowed from electricity theory. Capacitors store electrons when they are charged, and can discharge them to generate a current (flow of electrons). The cardiovascular analogy is that some segments of the vascular loops effectively store blood at low transmural pressure difference, while others pressurise the blood within them and so generate flow. The mathematical reciprocal of electrical capacitance is electrical elastance, but electrical engineers rarely use the term. The physiologist Hainsworth, for instance, uses the term capacitance to express the state of active vasoconstriction (mainly within veins) which affects cardiac output. Capacitance changes, he argues, participate in cardiovascular reflexes, but passive volume changes resulting from changes in transmural pressure difference are likely to be at least as important [28] [25]. Capacitance is typically used to describe the pressure/ volume characteristics of large veins and central veins which

208

Chapter 10

in health operate at the low pressure end of their compliance curve and do not contribute flow-assisting energy to their contained blood. They accommodate substantial changes in the venous excess volume at very low pressures and apply minimal resistance to flow.

Arterial pressure-volume characteristics in practice Clinicians have easier access to flow, stroke volume and pressure measurements on the arterial side of the vascular loops, so these are the more investigated and are used to guide treatment. The pressurevolume characteristics of arterial system feature both steady and pulsatile components. The classic definition of arterial compliance is the change in arterial blood volume due to a given change in arterial blood pressure C = ȟV/ȟP [29]. The relationship between vascular compliance, pressure, and flow rate is Q=C(dP/dt). Mathematically, arterial compliance is the reciprocal of elastance, which is mostly accounted for by large arteries and especially the thoracic aorta. We have noted above that venous capacitance is a major factor in the regulation of cardiac output. Arterial compliance is essential for efficient cardiac output. If the artery-arteriolar - capillary tubes conducting blood were entirely rigid, there could be no cardiac output. Arterial compliance is an important cardiovascular risk factor. Compliance diminishes with age and menopause. Arterial compliance can be measured by ultrasound as a pressure change (carotid artery) and volume change (outflow into aorta) relationship. Pulse contour analysis is another non-invasive method that allows easy measurement of arterial elastance to identify patients at risk for cardiovascular events. Low compliance is often used as an indication of arterial stiffness. An increase in the age and also in the systolic blood pressure is accompanied by decreased arterial compliance. A mathematician would logically state that advancing age and increasing hypertension are therefore accompanied by increasing arterial elastance, but the physiologist appreciates that stiffer arteries may have less elastic recoil to assist cardiac output. The mathematical concept of complianceelastance to describe the pressure-volume characteristics of blood vessels can therefore be confusing! Changes in mean arterial pressure to cardiac output define arterial resistance, but resistance is only one component of vasomotor tone. Compliance/ elastance is the other component.

A haemodynamic paradigm

209

Effective arterial elastance is the ratio of left ventricular end-systolic pressure and stroke volume. It is a simple and convenient way to characterize the arterial load from pressure-volume data measured in the left ventricle. Effective arterial elastance lumps the steady and pulsatile components of the arterial load into a single number, but it does not provide any information on their relative contribution. Additional information is required for an unequivocal characterization. Effective arterial elastance alone is not a measure of arterial stiffness, because resistance and heart rate also contribute to the parameters used to calculate it [30]. Dynamic arterial elastance is estimated by the pulse pressure variation to stroke volume variation relation in critical care practice. Dynamic arterial elastance is only one component of vasomotor tone, however, and increases in pulse pressure may not be proportional to increases in mean arterial pressure. The use of dynamic arterial elastance for clinical decision-making needs to be validated separately for different devices and types of patients [31]. Dynamic predictors of fluid responsiveness have become extremely fashionable in human and veterinary practice [32]. To test fluid responsiveness, the clinician can artefactually provoke an increase in venous excess volume and measure the consequent change in stroke volume or its derivatives and heart rate. Passive leg raising is intended to displace blood from the leg veins to the venae cavae. The reverse process is tilt table testing which reduces the venous excess by changing the subject’s posture from supine to head up and can be used to investigate the cause of syncope in patients. Remarkably, it has been shown that circadian rhythm affects cardiovascular responses to postural (orthostatic) stress, resulting in greater susceptibility to presyncope at night [33]. We should therefore expect anaesthesia to modify a subject’s responses to changed venous excess. Imposing a positive intrathoracic pressure by mechanical ventilation (inspiration) initially reduces the volume of blood within the superior vena cava (venous excess), right atrium, right ventricle and the pulmonary vascular loop. Simultaneously, the volume of blood within the inferior vena cava, the left atrium, left ventricle and intrathoracic aorta (arterial excess) is increased causing the next left ventricular stroke volume and associated aortic pressure to rise. If the positive pressure inspiration continues for longer than the pulmonary transit time, as in the Valsalva manoeuvre, blood available to the left atrium and ventricle is restricted

210

Chapter 10

and left ventricular stroke volume falls beat by beat. Releasing the raised intrathoracic pressure (expiration) restores blood volume to the superior vena cava (venous excess), right atrium and right ventricle with stabilisation of the blood circulatory rate. Stroke volume variation induced by mechanical ventilation will be exaggerated when the venous excess volume is low, and attenuated when the venous excess volume is high. Dynamic predictors of fluid responsiveness that have been described using mechanical ventilation-induced changes in intrathoracic pressure include stroke volume variation (SVV). In the absence of stroke volume measurement, surrogates such as pulse pressure variation, systolic pressure variation or even plethysmographic variation index are substituted. In post-surgical human practice, the minimum volume needed for an effective fluid challenge seems to be about 4 ml kg-1 infused over 5 minutes [34]. The haemodynamic effect of a fluid challenge (a disequilibrium) may be as short as ten minutes before steady-state returns [35]. For a clinician’s overview of the basic concepts of fluid responsiveness, see a review by the late Johan Groeneveld [36] who paraphrased Shakespeare’s Hamlet when he wrote “To fill or not to fill, that is the question.”

Professor Groeneveld made important contributions to the European Society of Intensive Care Medicine’s narrative review and expert panel recommendations on Fluid administration for acute circulatory dysfunction using basic monitoring [37].

Systemic vascular waterfall The observed fact that the arterial pressure at zero flow is consistently found to be higher than the venous pressure at zero flow could be explained either by the existence of a vascular waterfall (Starling resistor) or by a discharge of vascular capacitance. Using dog leg models, Sheldon Magder showed that the Starling resistor mechanism was the more likely [38]. The presence of a systemic vascular waterfall has been demonstrated in humans. The experimental method is to apply an inspiratory hold to a sedated and positive-pressure ventilated patient. The drop in cardiac output is compared to the drop in arterial pressure and the rise in central venous pressure. Extrapolation of the central venous pressure to zero circulatory rate gives the mean

A haemodynamic paradigm

211

systemic filling pressure, while extrapolation of the arterial pressure to zero circulatory rate gives the critical closing pressure. The difference between the two was found to be of the order of 25 mmHg: this confirms that arterial-side and venous-side resistances within the vascular loop should be calculated separately. Calculations of systemic vascular resistance using the difference between arterial and central venous pressure will both overestimate actual vasomotor tone and may not accurately represent changes in vasomotor tone. In a study on Intensive Care patients, an intravenous volume boost increased the mean systemic filling pressure, but unexpectedly the intravenous volume boost also increased the critical closing pressure [39]. The critical closing pressure seems to be located at about the capillary filling pressure, and different tissue beds have their own critical closing pressures determined by the local arteriolar tone.

Pressure-volume and flow characteristics of the vascular loop Large arteries direct blood to their own organ or tissue, and each has a vascular loop. Each loop segment is a conductive unit with its own pressure-volume characteristic, and so each segment contributes in a unique way to the blood distribution or collection in the loop. It is possible to collect data on the pressure-volume characteristics of the different peripheral vascular loop segments in the upper arm (axillary/ brachial artery) or thigh (iliac/ femoral artery) by co-extensive bioimpedance plethysmography. In simple terms, this entails placing bioimpedance electrodes under a standard blood pressure cuff and plotting volume changes as pressure rises and then falls in the tissues encompassed by the cuff during a cycle of inflation and deflation. During inflation, blood is squeezed out of each segment of the peripheral vascular loop in turn from large veins, small veins, venules, capillaries, arterioles, small arteries and finally the large artery. There are ‘compliance inflection points’ in the rate of change of volume at certain pressure points related to transitions from one segment to the next. Data collected during cuff inflation are referred to as ‘depletion data’. Data are also collected during cuff deflation and called ‘replenishment data’. Mean and filling pressures can be ascertained for each capacitative segment of the peripheral vascular loop. For the pulsatile (arterial) segment diastolic, mean and systolic are determined

212

Chapter 10

by volume change rather than pulsatility and may be more accurate than traditional non-invasive arterial pressure measurement, especially when arterial pressure is low. • • • • •

Large veins – mean and filling pressure Small veins – mean and filling pressure Venules – mean and filling pressure Capillaries – mean and filling pressure Arteries – diastolic, mean and systolic pressure

As relative volumes are recorded for each segment, compliances can also be calculated. The potential for the diagnosis of vascular disorders by the analysis of “compliance stacks” (volume against pressure range for each segment of the peripheral vascular loop) is unexplored to date. It is likely that more research would have been done around the world using this technology, and more would be known about the extracardiac regulation of cardiac output, if it were not protected by US Patents. In particular, PN 6,749,567 protects co-extensive monitoring by volume and pressure applicators and sensors for physiologic parameters including central venous and arterial pressure among others. We will now consider the characteristics of the loop segments from lowest pressure to highest. Clinicians are familiar with the anatomy of several large veins which join the central venae cavae because they cannulate them for vascular access. The internal jugular veins or subclavian veins are foremost, but the external jugular veins of the neck, axillary vein, cephalic and basilic veins of the upper limb, and the posterior intercostal veins of the thorax can also be used to access the superior vena cava which leads to the right atrium of the heart. The femoral vein is sometimes used to access the inferior vena cava, though complications of access make it less desirable. There are of course many other large veins less apparent in surface anatomy. Some of the large veins run parallel to large arteries and are often named accordingly. Large veins in the legs are classified as superficial or deep, these being linked by short connecting veins. The large veins have the lowest intravascular pressures within the peripheral vascular loop and, at or above the level of the heart, are mostly unstretched. As we can see by raising or lowering an arm, and observing the emptying or filling of the veins on the back of the hand, the large vein pressure is a little higher than right atrial and close to the central venous pressure. When

A haemodynamic paradigm

213

there is a continuous column of blood between them, peripheral venous pressure approximates central venous. Veins in the limbs feature oneway valves which have two cusps or leaflets with edges that meet to prevent backwards flow, thereby assisting the efficient return of blood to the heart, especially against gravitational effects. The valves are a specialised section of the tunica intima, the thin endothelial inner lining. The tunica media of veins typically contains 2-3 layers of muscle cells, while the tunica adventitia contains longitudinal collagen fibres and vasa vasorum. The chief role of the large veins is to provide a minimal-resistance conduit to the central veins and right atrium so that the relaxing right ventricle has easy access to sufficient venous excess volume with which to fill. Veins have smooth muscle cells within their tunica media, and smooth muscle tone sets the degree of vasoconstriction that reduces large vein capacitance but has no significant effect on resistance. Many unnamed small veins (typically less than 1mm in diameter) form irregular networks and deliver blood to the large veins. They are at a higher transmural pressure difference than the large veins and are mostly distended in health. Constriction of small veins contributes to reduced venous capacitance and displaces small vein volume towards the large veins, but the increase in venous resistance is relatively insignificant. Venules are typically 10 to 100 Ɋm in diameter and deliver blood to the small veins. The immediate post-capillary venules are uninvested, but there are collecting venules that are invested with pericytes, and muscular venules invested with smooth muscle. Muscular venules are innervated and respond to a variety of constrictors which have two effects; to increase resistance to capillary blood exiting the capillary beds (post-capillary resistance), and to displace some of the blood volume contained in venules towards the small veins and large veins (capacitance effect). A vasomotor role of pericytes that invest postcapillary and collecting venules is being investigated. Small venules (less than 30 nm) are more often associated with pericytes, rather than invested with smooth muscle. Nonetheless they display active tone, and show venular constriction in response to reduced oxygen tension or norepinephrine. Though these vessels lack innervation, it is possible that pericyte-mediated venular contraction contributes to haemodynamic responses.

214

Chapter 10

We have seen in the preceding narrative that the disposition of venous blood between the segments of the venous side of the vascular loop is dependent on capacitance; general venoconstriction shifts the balance in favour of the large veins where the venous excess volume available to the right atrium and the diastolic right ventricle is manifested as the fullness of the jugular veins, and where the fullness of the inferior vena cava can be assessed on ultrasound examination. Venular resistance to blood flow is lower than arteriolar resistance because there are more venules of a given size than there are arterioles. Changes in venular resistance play an important role in determining capillary and post-capillary venular hydrostatic pressure. Venules and small veins serve as capacitance (volume storage) vessels and normally contain most of the total blood volume. At the point where capillaries enter venules, the venular filling pressure is around 18 mmHg above heart level. Venular pressure is uniform across a tissue because of the inter-arcading arrangement of the venular bed. With no significant resistors after the muscular venules, the pressure gradient associated with flow between the capillary bed and the central veins is around 15 mmHg. Capillaries and small non-invested venules form the microcirculation and are at the centre of the pulmonary and systemic vascular loops. The arterial side distributes blood to, and the venous side collects blood from, capillary beds. The rate at which solute and smaller solvents move from plasma to the interstitium critically depends on tightlyregulated capillary pressure, or more precisely the gradient of luminal to interstitial hydrostatic pressure differences from the arteriolar outflow (capillary filling pressure) to the capillary outflow (venular filling pressure). Too high, and hyperfiltration will cause oedema. Too low, and filtration will dry up and venular filling pressure will become inadequate to maintain the venous excess, with consequent reduced central venous pressure and stroke volume (hypovolaemia). There are several interdependent mechanisms by which average capillary hydrostatic pressure can be maintained within one or two mmHg despite wide variations in arterial pressure [40]. Capillaries that are open to the active circulation have hydrostatic pressures higher than the colloid osmotic pressure of the plasma along their length: capillaries in which flow has been interrupted have no arteriolar to venular gradient. Their hydrostatic pressure is close to that of the collecting venule into which they drain, and that pressure is usually

A haemodynamic paradigm

215

close to the colloid osmotic pressure of plasma. Capillary fluid shift mechanism is the term used for the regulatory concept that an increase in plasma volume increases capillary pressure, which raises the filtration rate and so tends to lower the plasma volume thereby restoring a steady-state of plasma volume and capillary pressure. Traditional physiology teaching often includes the presence of precapillary sphincters to control the distribution of blood flow through a capillary bed, but current thinking is that sphincters are a particular feature of mesenteric capillary beds. Arterioles are at the lower pressure end of the distributive arterialside of the vascular loop. They feature just a few layers of smooth muscle cells within the tunica media, but they are the site of precapillary resistance and a major contributor to peripheral resistance. Recall that just a small change in luminal radius brings about a large change in vascular resistance in this segment of the vascular loop. Arteriolar tone is in a visible state of readjustment three or four times a minute, a phenomenon called arteriolar vasomotion. Vasomotion is the spontaneous oscillation of vasomotor tone. Vasomotion was first observed in 1852 and has remained somewhat enigmatic. By causing small frequent Starling force disequilibria it may facilitate short bursts of absorption during otherwise sustained filtration [41]. Vasomotion is preserved in patients with sepsis [42] and may increase in response to tissue hypoxia [43]. Small arteries feed the arterioles. The muscle layer of the tunica media is sandwiched by elastic layers. The tunica adventitia is thick, and formed of collagen and elastin. Large arteries and the aorta have fewer muscle fibres and feature broad, fenestrated sheets of collagen and elastin in the tunica media. The tunica adventitia contains vasa vasorum. The arterial excess volume largely resides in the large and small arteries. There are several special case vascular loops. Portal circulations take blood leaving the first capillary bed and distribute it to a second capillary bed before it is collected and returned to the large and central veins. The largest is the hepatic portal system, often simply referred to as the portal system, which supplies the liver with blood that has passed through capillaries of the small intestine, right side of the colon and spleen. Blood volume within the liver and spleen is a significant contributor to total venous-side capacitance, but (at least

216

Chapter 10

in humans) splenectomy and hemi-hepatectomy are well tolerated haemodynamically.

Haemodynamic computation models The power of modern computers is being harnessed to develop increasingly realistic simulations of cardiovascular performance. Just as an aspiring pilot can learn to fly with the aid of a flight simulator, so aspiring critical care clinicians can learn about the principles of haemodynamic manipulation from a simulator. The human machine is, however, much more complex than any flying machine, and the ability of a computer to predict an individual patient’s response to a surgical or pharmaceutical intervention is still too limited for clinical application. For an example of progress in anticipating a patient’s interaction with a mechanical ventricular assistance device, see [44]. A research Team at the CARIM School for Cardiovascular Diseases, Maastricht University Medical Center in The Netherlands is working on a mathematical model of the human heart and circulation that they call CircAdapt. Circadapt enables real-time simulation of cardiovascular system dynamics in a wide variety of physiological and pathophysiological situations. They claim that their model uniquely includes structural adaptation of the vascular and myocardial tissues to the mechanical load generated by the model itself. The model is disseminated as freeware at http://www.circadapt.org/home. HARVI is a commercial enterprise from the USA by Daniel Burkhoff which is available for purchase at the iTunes store. Applying experimentally measured changes in arterial pressure and heart rate during infusion of normal saline in healthy volunteers, researchers at Eindhoven University of Technology have taken a first step towards a model predictive clinical decision support tool for fluid administration in a clinical setting [45]. Do they match up to Anderson’s seven-point test?

A haemodynamic paradigm

217

Shock It is tempting to say that shock is a life-threatening condition of haemodynamic failure. For instance, the Third International Consensus Definitions for Sepsis and Septic Shock ran an iterative Delphic process to decide that shock is the presence of profound circulatory, cellular, and metabolic abnormalities that are associated with a greater risk of mortality [46]: “Patients with septic shock can be clinically identified by a vasopressor requirement to maintain a mean arterial pressure of 65 mm Hg or greater and serum lactate level greater than 2 mmol/l (>18 mg/dl) in the absence of hypovolemia. This combination is associated with hospital mortality rates greater than 40%.”

Classically, shock comes in four flavours which, on the basis of their presumed pathophysiology, are labelled: • • • •

Distributive Cardiogenic Obstructive Hypovolaemic

In a true closed loop system like the cardiovascular system there is no start or end of the loop. Everything in the loop depends upon everything else in the loop. It is impossible to know which elements are leading and which ones are following. Central venous and arterial are just two of the pressure points along the pathway of flow. A Guytonian view puts the heart in control of blood flow by virtue of the energy that it adds to the blood. Indeed, the heart normally has an excess of energy that it can provide to support a hyperdynamic circulation when called upon.

Refractory shock Refractory shock has been defined as an inadequate hemodynamic response to high doses of standard vasopressor medications, with a short-term mortality exceeding 50%. When norepinephrine proves inadequate to restore arterial pressure, vasopressin or epinephrine can be added. Adjunctive therapies, such as hydrocortisone, thiamine, and Vitamin C should be considered when combination vasopressor

218

Chapter 10

therapy is needed. Angiotensin II or methylene blue can increase arterial pressure and reduce the need for high doses of catecholamine vasopressors in severe or refractory vasodilatory shock [47].

Shock in Children The predominant feature of septic shock in adults is said to be distributive vasomotor paralysis (reduced systemic vascular resistance). Myocardial dysfunction manifests as a decreased ejection fraction, though Qt is usually maintained or increased by tachycardia and ventricular dilation. Septic shock in children is more typically associated with hypovolaemia, and it is observed that children respond well to fluid volume expansion. Low Qt, not low systemic vascular resistance, is associated with mortality in paediatric septic shock. Attainment of a therapeutic goal of cardiac index 3.3–6.0ௗl/min/m2 is associated with higher survival rate. Moreover, a reduction in oxygen delivery rather than a defect in oxygen extraction seems to be the major determinant of oxygen consumption in children. Attainment of the therapeutic goal of oxygen delivery greater than 200ௗml/min/m2 may also be associated with improved outcome [48].

Assessing shock and the response to treatment Capillary refill time (CRT) is defined as the time taken for blanching to disperse from a capillary bed after pressure has been applied. First described in 1947, it has since become widely adopted as part of the rapid structured circulatory assessment of ill children [49]. Capillary blood flow is affected by; •

• •

Pressure and volume characteristics of arterioles and their associated microvessels. Functional capillary density or the number of capillaries in a given area which are filled with flowing red blood cells is a measurable characteristic. Arteriolar tone. Blood constituents. For example, the size and rigidity of red blood cells and the state of plasma viscosity can affect flow through narrow capillaries.

A haemodynamic paradigm

219

Meta-analysis suggests that an abnormal CRT is a valid ‘red flag’ for critical illness in children, but low sensitivity means that a normal capillary refill time should not reassure a concerned clinician [50]. CRT has also been used in adult medicine, though studies have failed to reveal a strong correlation between systemic haemodynamic and microvascular measures [51, 52]. In an experiment on healthy blood donors, orthostatic vital signs were found to be more sensitive and specific than CRT in revealing the 450 ml blood loss. CRT does not appear to be a useful test for detecting mild-to-moderate hypovolaemia in adults [53]. A dissociation between macrocirculatory function and microcirculatory function has been hypothesised. The Andromeda shock study compared normalization of capillary refill time versus normalization (or a decrease of more than 20%) of lactate measured every 2 hours. The common protocol started with fluid responsiveness assessment and fluid loading in responders, followed by a vasopressor and an inodilator test if necessary. Thereafter, a randomly allocated resuscitation endpoint was applied. The CRT-guided approach was not inferior to the lactate level–guided approach, but did not reduce allcause 28-day mortality [54].

Heart failure; an appropriate name for venous congestion? We have considered above a role for pericyte-invested venules as a motor of venous return. The disease we label congestive heart failure could, at least in part, be a disease of pericyte and venular insufficiency. Here we look at the signs and symptoms of extracellular fluid accumulation that result in increased cardiac filling pressures [55]. The gold standard for diagnosing congestion in heart failure is direct measurement of right atrial pressure and pulmonary capillary wedge pressure. Signs and symptoms with reasonable specificity (positive rules in the diagnosis) include; • • •

Jugular venous pulse height Jugular venous reflux Hepatomegaly

220

• • •

Chapter 10

Dyspnoea Rales on auscultation. Echocardiography ် Inspiratory diameter of the inferior vena cava < 12 mm ် Deceleration time < 130 ms ် Pulmonary vein Systolic/Diastolic velocity < 1

Signs and symptoms with reasonable sensitivity (negative rules out the diagnosis) include; • • • • •

Bilateral leg oedema Exertional dyspnoea Orthopnoea S3 heart sound Echocardiography ် Inspiratory diameter of the inferior vena cava < 12 mm ် Mitral inflow E-wave velocity > 50 (cm/s) ် Lateral E/e’ > 12 ် Deceleration time < 130 ms ် Pulmonary vein Systolic/Diastolic velocity < 1 ် Diffuse B-lines on lung ultrasound

In the most severe congestion echocardiography shows stop-start blood flow in the inferior vena cava. The goal of treatment is to return the expanded extracellular fluid volume (and therefore the plasma volume) to euvolaemia. Clinical assessments include signs and symptoms listed above, plus: • • •

Improving six minute walking test Falling natriuretic peptide plasma concentration (B-type natriuretic peptide or N-terminal pro B-type natriuretic peptide) Radiographic assessment of cardiomegaly and pulmonary oedema

Non-invasive measurement of central venous pressure by coextensive bioimpedance plethysmography should be a low-cost way of following the progress of heart failure patients but has yet to be evaluated.

A haemodynamic paradigm

221

Re-evaluating central venous pressure monitoring There is a fashion for disparaging the clinical utility of central venous pressure monitoring, but the following points are offered in its favour: • •

•

•

Central venous pressure is a clinical surrogate for right atrial pressure Pra and will be low during hypovolaemia. Central venous pressure will stabilise at its ‘normal’ autoregulated level once euvolaemia is achieved. The zero reference level is the right atrium, which is approximated by the sternal manubrium. From that reference point the normal central venous pressure is about 0. Central venous pressure rises above normal when the heart cannot shift the venous return and central venous volume accumulates. Recall that heart failure exists whenever heart function is inadequate to produce the circulatory flow rate which would be allowed by the mean cardiovascular pressure and inlet impedance. Fluid overloading in an attempt to achieve supranormal central venous pressure decreases rather than increases venous return and cardiac output. In the extreme, fluid overloading causes cardiovascular collapse as the pressure difference for venous return falls.

A better haemodynamic paradigm? I will develop my account of the glycocalyx model haemodynamic paradigm in the capillary bed, or to be more precise the network of capillaries and uninvested collecting venules. I have described above the vital need to regulate the capillary and venular transmural pressures for the optimal control of filtrate flow (Jv) from plasma to the interstitium. There are, of course, tissues where fluid is transferred from the interstitium to the plasma, such as intestinal mucosa and lymph nodes, but most of the capillary beds are in a state of sustained fluid filtration most of the time. The capillary filling pressure is close to diastolic arterial pressure when arterioles are dilated as in sepsis and post-surgical or post-traumatic vasoplegia. The mean capillary pressure is close to what Carl Rothe described as the pivot pressure for the circulation of blood. We may also note that the velocity of blood within the peripheral vascular loop

222

Chapter 10

is at its lowest in the capillary bed, barely 1 cm s-1, and the total crosssectional area is greatest, around 1,000 cm2. Blood leaving the capillary bed enters the venular segment of the vascular loop. The venular filling pressure is the highest pressure in the venous blood collecting system and the pressure drop from entry to exit of this segment is higher than other segments of the collecting system. Funnelling can be said to occur as the cross-sectional area at entry falls from around 1000 cm2 to just 100 cm2 at exit. Acceleration of blood is highest in the venules, as blood leaves to enter the small vein segments at around 10 cm s-1. Venules are unsurprisingly near full distension throughout their length and on the flatter part of their pressure-volume capacitance curve. There is energy stored as elastic recoil contributing to the acceleration of blood from venules to small veins, as well as kinetic energy contributed by the heart to blood flow into the venules. This vascular loop segment has been called the Second Heart. During exercise, intermittent external compression of venules and small veins and deep veins by somatic muscles, especially in the legs, is another source of energy accelerating blood towards the large veins. Blood entering the small veins finds itself in a less-distended vasculature at lower transmural pressure and with less acceleration but still increasing velocity. Elastic recoil of small veins adds to the energy associated with blood flow, but the pressure-volume curve is getting steeper; compliance here is greater than in the venules. Blood entering the large veins and central veins is close to its terminal velocity of around 15 cm s-1. There is almost no elastic recoil potential energy store available, but the effects of the respiratory pump of rising and falling intrathoracic pressure in a valved system are evident. The large veins impose very little resistance to blood flow and can be thought of as a passive conduit to the central veins. However, their relatively large volume at relatively low pressure enables a substantial fall in volume during active venoconstriction, supporting the effective blood volume when there is reduced total blood volume. In Carpenterian terms, the capacitance change by venoconstriction supports the venous excess. A very important practical point is that in profound hypovolaemia the flow of blood is significantly aided by muscle activity in the limbs and ventilatory activity in the thorax. The induction of anaesthesia, and in particular neuromuscular blockade,

A haemodynamic paradigm

223

will reduce circulatory flow and raising thoracic pressure by assisted breathing will also have adverse consequences. The anaesthetist faced with a hypovolaemic patient will, whenever possible, “Fill First.” The right ventricle delivers all the blood it receives per beat into the pulmonary circulation, via the venous excess, at an auto-regulated central venous (right atrial) pressure. Cardiac output is a function of metabolic demand of the body, so the right ventricle and pulmonary vascular loop must pass on enough blood to the pulmonary vein/ left atrium to allow unimpeded left ventricular filling during diastole. This is assisted by the facts that the heart rate is the same for each ventricle, the state of contractility is linked and there is a shared interventricular wall. Ventricular interdependency is complicated by the fact that very increased right ventricular diastolic volume displaces the interventricular wall and can compromise the left ventricular diastolic volume. Right ventricular dysfunction or pulmonary embolism may critically limit pulmonary venous accumulation of blood and pulmonary venous excess leading to reduced left atrial pressure and reduced left ventricular output. The pulmonary vascular loop is short, with normally low vascular impedance and low arterial pressure. Right heart catheterisation is used to measure pulmonary artery pressure, pulmonary artery blood flow (cardiac output) and pulmonary venous pressure after balloon occlusion of a small pulmonary artery and equalisation of the resultant still column of blood to the venous confluence.

References 1. Rothe CF. Mean circulatory filling pressure: its meaning and measurement. J Appl Physiol (1985). 1993;74:499-509. 2. Guyton AC. Determination of cardiac output by equating venous return curves with cardiac response curves. Physiol Rev. 1955;35:123-129. 3. Magder S. Point: the classical Guyton view that mean systemic pressure, right atrial pressure, and venous resistance govern venous return is/is not correct. J Appl Physiol (1985). 2006;101:1523-1525. 4. Brengelmann GL. Counterpoint: the classical Guyton view that mean systemic pressure, right atrial pressure, and venous resistance

224

Chapter 10

govern venous return is not correct. J Appl Physiol (1985). 2006;101:1525-6; discussion 1526. 5. Henderson WR, Griesdale DE, Walley KR, Sheel AW. Clinical review: Guyton--the role of mean circulatory filling pressure and right atrial pressure in controlling cardiac output. Crit Care. 2010;14:243. 6. Magder S. Right Atrial Pressure in the Critically Ill: How to Measure, What Is the Value, What Are the Limitations. Chest. 2017;151:908916. 7. Persichini R, Silva S, Teboul JL et al. Effects of norepinephrine on mean systemic pressure and venous return in human septic shock. Crit Care Med. 2012;40:3146-3153. 8. Good old physiology in a modern jacket. [editorial]. Crit Care Med 2012;40(12):3309. 9. Neto AS. Decrease in mean systemic filling pressure, increase in unstressed blood volume, or both[letter]. Crit Care Med 2013;41(2):e19. 10. Berlin DA, Bakker J. Starling curves and central venous pressure. Crit Care. 2015;19:55. 11. Parkin WG, Leaning MS. Therapeutic control of the circulation. J Clin Monit Comput. 2008;22:391-400. 12. Cooke K, Sharvill R, Sondergaard S, Aneman A. Volume responsiveness assessed by passive leg raising and a fluid challenge: a critical review focused on mean systemic filling pressure. Anaesthesia. 2018;73:313-322. 13. Maas JJ, Pinsky MR, Geerts BF, de Wilde RB, Jansen JR. Estimation of mean systemic filling pressure in postoperative cardiac surgery patients with three methods. Intensive Care Med. 2012;38:14521460. 14. Gupta K, Sondergaard S, Parkin G, Leaning M, Aneman A. Applying mean systemic filling pressure to assess the response to fluid boluses in cardiac post-surgical patients. Intensive Care Med. 2015;41:265-272. 15. Maas JJ, Geerts BF, van den Berg PC, Pinsky MR, Jansen JR. Assessment of venous return curve and mean systemic filling pressure in postoperative cardiac surgery patients. Crit Care Med. 2009;37:912-918. 16. Anderson RM. The gross physiology of the cardiovascular system. Racquet Press; 1993 17. Aya HD, Rhodes A, Fletcher N, Grounds RM, Cecconi M. Transient stop-flow arm arterial-venous equilibrium pressure measurement:

A haemodynamic paradigm

225

determination of precision of the technique. J Clin Monit Comput. 2016;30:55-61. 18. Yastrebov K, Aneman A, Slama M et al. The stop-flow arm equilibrium pressure in preoperative patients: Stressed volume and correlations with echocardiography. Acta Anaesthesiol Scand. 2019 19. Levick RJ. An Introduction to Cardiovascular Physiology. Hodder Education Publishers; 2010 20. Reddi BA, Carpenter RH. Venous excess: a new approach to cardiovascular control and its teaching. J Appl Physiol (1985). 2005;98:356-364. 21. Pinsky MR. The right ventricle: interaction with the pulmonary circulation. Crit Care. 2016;20:266. 22. Ponikowski P, Voors AA, Anker SD et al. 2016 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure: The Task Force for the diagnosis and treatment of acute and chronic heart failure of the European Society of Cardiology (ESC)Developed with the special contribution of the Heart Failure Association (HFA) of the ESC. Eur Heart J. 2016;37:2129-2200. 23. Anderson RM, Larson DF, Lundell DC. The interrelationship of factors controlling cardiac output. Med Hypotheses. 1983;10:77-95. 24. Rothe CF. Point: active venoconstriction is/is not important in maintaining or raising end-diastolic volume and stroke volume during exercise and orthostasis. J Appl Physiol (1985). 2006; 101:1262-4; discussion 1265. 25. Hainsworth R, Drinkhill MJ. Counterpoint: active venoconstriction is not important in maintaining or raising end-diastolic volume and stroke volume during exercise and orthostasis. J Appl Physiol (1985). 2006;101:1264-5; discussion 1265. 26. Lichtenstein SV, El-Dalati H, Panos A, Rice TW, Salerno TA. Systemic vascular effects of epinephrine administration in man. Journal of Surgical Research. 1987;42:166-178. 27. Magder S. Volume and its relationship to cardiac output and venous return. Crit Care. 2016;20:271. 28. Hainsworth R. The Importance of Vascular Capacitance in Cardiovascular Control. News Physiol Sci. 1990;5:250-254. 29. Spencer MP, Denison AB. The aortic flow pulse as related to differential pressure. Circ Res. 1956;4:476-484. 30. Segers P, Stergiopulos N, Westerhof N. Relation of effective arterial elastance to arterial system properties. Am J Physiol Heart Circ Physiol. 2002;282:H1041-6.

226

Chapter 10

31. Pinsky MR. Defining the boundaries of bedside pulse contour analysis: dynamic arterial elastance. Crit Care. 2011;15:120. 32. Drozdzynska MJ, Chang YM, Stanzani G, Pelligand L. Evaluation of the dynamic predictors of fluid responsiveness in dogs receiving goal-directed fluid therapy. Vet Anaesth Analg. 2018;45:22-30. 33. Hu K, Scheer FA, Laker M, Smales C, Shea SA. Endogenous circadian rhythm in vasovagal response to head-up tilt. Circulation. 2011;123:961-970. 34. Aya HD, Rhodes A, Chis Ster I, Fletcher N, Grounds RM, Cecconi M. Hemodynamic Effect of Different Doses of Fluids for a Fluid Challenge: A Quasi-Randomized Controlled Study. Crit Care Med. 2017;45:e161-e168. 35. Aya HD, Ster IC, Fletcher N, Grounds RM, Rhodes A, Cecconi M. Pharmacodynamic Analysis of a Fluid Challenge. Crit Care Med. 2016;44:880-891. 36. Cherpanath TG, Geerts BF, Lagrand WK, Schultz MJ, Groeneveld AB. Basic concepts of fluid responsiveness. Neth Heart J. 2013;21:530536. 37. Cecconi M, Hernandez G, Dunser M et al. Fluid administration for acute circulatory dysfunction using basic monitoring: narrative review and expert panel recommendations from an ESICM task force. Intensive Care Med. 2018 38. Magder S. Starling resistor versus compliance. Which explains the zero-flow pressure of a dynamic arterial pressure-flow relation. Circ Res. 1990;67:209-220. 39. Maas JJ, de Wilde RB, Aarts LP, Pinsky MR, Jansen JR. Determination of vascular waterfall phenomenon by bedside measurement of mean systemic filling pressure and critical closing pressure in the intensive care unit. Anesth Analg. 2012;114:803-810. 40. Zweifach BW. Microcirculatory homeostasis 1930-1990: insight into microcirculatory readjustments provided by studies on the peripheral circulatory insufficiency of the shock syndrome. Microcirculation. 1995;2:245-251. 41. Levick JR. Capillary filtration-absorption balance reconsidered in light of dynamic extravascular factors. Exp Physiol. 1991;76:825857. 42. van Ierssel SH, Van Craenenbroeck EM, Hoymans VY, Vrints CJ, Conraads VM, Jorens PG. Endothelium dependent vasomotion and in vitro markers of endothelial repair in patients with severe sepsis: an observational study. PLoS One. 2013;8:e69499.

A haemodynamic paradigm

227

43. Pajk W, Stadlbauer KH, Kleinsasser A et al. The impact of endotoxin on jejunal tissue oxygenation. Microcirculation. 2017;24 44. Sack KL, Dabiri Y, Franz T, Solomon SD, Burkhoff D, Guccione JM. Investigating the Role of Interventricular Interdependence in Development of Right Heart Dysfunction During LVAD Support: A Patient-Specific Methods-Based Approach. Front Physiol. 2018;9:520. 45. Rosalina TT, Bouwman RA, van Sambeek MRHM, van de Vosse FN, Bovendeerd PHM. A mathematical model to investigate the effects of intravenous fluid administration and fluid loss. J Biomech. 2019 46. Singer M, Deutschman CS, Seymour CW et al. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA. 2016;315:801-810. 47. Jentzer JC, Vallabhajosyula S, Khanna AK, Chawla LS, Busse LW, Kashani KB. Management of Refractory Vasodilatory Shock. Chest. 2018;154:416-426. 48. Davis AL, Carcillo JA, Aneja RK et al. American College of Critical Care Medicine Clinical Practice Parameters for Hemodynamic Support of Pediatric and Neonatal Septic Shock. Crit Care Med. 2017;45:1061-1093. 49. King D, Morton R, Bevan C. How to use capillary refill time. Arch Dis Child Educ Pract Ed. 2014;99:111-116. 50. Fleming S, Gill P, Jones C et al. The Diagnostic Value of Capillary Refill Time for Detecting Serious Illness in Children: A Systematic Review and Meta-Analysis. PLoS One. 2015;10:e0138155. 51. Dubin A, Pozo MO, Casabella CA et al. Increasing arterial blood pressure with norepinephrine does not improve microcirculatory blood flow: a prospective study. Crit Care. 2009;13:R92. 52. Hernandez G, Bruhn A, Luengo C et al. Effects of dobutamine on systemic, regional and microcirculatory perfusion parameters in septic shock: a randomized, placebo-controlled, double-blind, crossover study. Intensive Care Med. 2013;39:1435-1443. 53. Schriger DL, Baraff LJ. Capillary refill--is it a useful predictor of hypovolemic states. Ann Emerg Med. 1991;20:601-605. 54. Hernández G, Ospina-Tascón GA, Damiani LP et al. Effect of a Resuscitation Strategy Targeting Peripheral Perfusion Status vs Serum Lactate Levels on 28-Day Mortality Among Patients With Septic Shock: The ANDROMEDA-SHOCK Randomized Clinical Trial. JAMA. 2019;321:654-664. 55. Mullens W, Damman K, Harjola VP et al. The use of diuretics in heart failure with congestion - a position statement from the Heart Failure

228

Chapter 10

Association of the European Society of Cardiology. Eur J Heart Fail. 2019;21:137-155.

CHAPTER 11 AN IMPROVED PARADIGM OF FLUID PHYSIOLOGY AND THERAPY

Chapter summary It is seven years since Tom Woodcock and Son first proposed an improved paradigm for prescribing intravenous fluid therapy based on the revised Starling principle. We called it The Glycocalyx Model Paradigm. If you've read this far, you're already well on your way to beating the odds of grasping it. In this final chapter I recap some issues addressed in earlier chapters and use a volume kinetic approach to explain the glycocalyx model paradigm. To be clear, I emphasise that this is not a statement of physiology, it is a paradigm for rational prescribers based on physiological principles.

A valid clinical paradigm of body water disposition and response to intravenous infusions must explain observed facts • •

• •

Infusion solutions with colloid osmotic pressure cause more red blood cell dilution than isotonic salt solutions that contain no larger molecules. An isotonic crystalloid solution infused into a healthy nonanaesthetised subject has only a very transient effect on red blood cell dilution, but under anaesthesia the duration of dilution is markedly prolonged. When using arterial pressure as a resuscitation end point, the effective resuscitation volume is around 50% higher with a crystalloid solution than with a colloid solution. After resuscitation treatment, patients who received colloids

230

• •

Chapter 11

tend to have higher central venous pressure and cardiac output but… Colloid-associated haemodilution means that oxygen delivery is not increased. Hypovolaemic hypotension is effectively prevented or reversed during venesection by isotonic crystalloid solutions given in a ratio of 3x the blood loss. The possibility of haemodynamicallystable volume exchange with less than 3 to 1 replacement has yet to be investigated.

Moore’s Volume Kinetics F. D. Moore investigated the “Clinical Kinetics of Blood Volume Support” in a healthy volunteer experiment reported from Harvard University in 1966 [1]. Moore fully appreciated that the capillaries of the body are highly heterogenous in their resistance to filtration of solvent and small solutes and their permeability to larger molecules like albumin. The fact had been described by Pappenheimer, Landis and Renkin in their classic works. He pointed out that kinetic analysis of the distribution of extracellular fluid between plasma and interstitial fluid involves three biologic processes: 1. The diffusive exchange of solvent and solutes. 2. Convective processes: Net effect of filtration-reabsorption in capillaries and venules, and lymph flow according to the (original) Starling principle. 3. Vesicular transport of macromolecules, at the time referred to as pinocytosis or cytopempsis. Net movement of water, he argued, results from an imbalance between ingress and egress among these three different fluid-flow mechanisms. He had previously demonstrated that pressor therapy causes water and salt to leave the circulation during the period of drug injection. In a study of haemorrhagic shock, he found that “the post-haemorrhagic state, while producing a net fluid flow from interstitial to plasma phases of body water, biases the distribution of an infused balanced salt solution toward the plasma volume”. He was particularly interested in reports, which he confirmed, that normal transcapillary refilling after haemorrhage in man does not involve any significant period of hypoalbuminaemia, which would be expected if refilling was solely due

An improved paradigm of Fluid Physiology and Therapy

231

to capillary reabsorption of solvent and small solutes.

The ratio of plasma volume to interstitial fluid volume “In the steady state there is a dynamic equilibrium between the volume of fluid in the interstitial phase and that found in the plasma. This partition of the extracellular fluid can be expressed as the PV:IF ratio and is maintained by a number of anatomic barriers and biochemical forces.”

Moore’s experimental method was to measure red cell volume by chromium-labelled erythrocytes, plasma volume by Evans blue dye, and the extracellular fluid volume by radiobromide dilution. He calculated a Whole Body Haematocrit from the red cell and plasma volumes. Corrections were made for the Donnan equilibrium and for bromide uptake by red blood cells. His measurement of the normal plasma volume to interstitial fluid volume ratio PV:IF became embedded in what practitioners considered to be a physiological approach to fluid therapy. The pre-haemorrhage PV:IF ratio was found to be around 0.23, suggesting that “one fifth of the entire extracellular volume is to be found in the plasma”. Notice, however, that the concept of PV:IF ratio is emphasised in Moore’s commentary in order to highlight the fact that it is not a fixed ratio, it changes during haemorrhage, during spontaneous plasma refill, and during the infusion of intravenous fluids.

The rate of plasma volume refill after blood loss Nine subjects had about 12% of their blood volume removed, which did not change heart rate or blood pressure in these fit young men. We would now attribute the cardiovascular stability to preservation of the venular filling pressure and venous excess volume by protection of the stressed blood volume at the expense of the unstressed. Over the following four hours, their plasma volume was almost completely restored at the expense of interstitial fluid volume. The average plasma volume refill rate was found to be around 1 ml min-1. Moore and his contemporaries believed this was due to net capillary reabsorption. It has since been demonstrated that thoracic duct lymph flow and protein return rate are significant contributors to blood volume and plasma

232

Chapter 11

protein restitution after haemorrhage [2]. In awake sheep with lymphatic cannulation, hypotensive hypovolaemia induced enhanced lymphatic pumping [3]. The interpretation today therefore is that transendothelial filtration rate Jv fell following reduction of capillary pressure to become 1 ml min-1 less than the enhanced lymph flow rate Qlymph. This would account for the absence of hypoalbuminaemia during post-haemorrhage plasma refill.

Volume kinetics of an infused isotonic salt solution The next stage of the experiment was to give these men 500 ml h-1 of isotonic salt solution over 4 hours. Moore hypothesised that the measured large vessel haematocrit would at all times be the same as whole body haematocrit if the “dispersal rate” of infused Ringers Lactate throughout the extracellular fluid space was the same as the infusion rate. His demonstration that the measured large vessel haematocrit was always less than the whole body haematocrit during Ringers Lactate infusion at 8.33 ml min-1 (500 ml h-1) was taken as proof of partial retention of infused isotonic salt solution in the plasma volume, at least transiently. From the volume kinetics of this crystalloid infusion (its rate of dispersal) he determined that the net vector of fluid exchange from plasma to interstitial fluid, and from interstitial fluid to plasma, is of the order of 5 ml min-1.

Crystalloid infusion enhances the movement of interstitial proteins to the plasma Moore observed that while two thirds of the infused 2 l isotonic salt solution was leaving the circulation, approximately 15-17 g protein (mostly albumin) was entering the plasma. The saline infusion, rather than producing a washout of plasma protein to the interstitium, appeared to restore interstitial fluid volume sufficiently to support protein flow to the plasma. He also presciently warned that “much larger saline infusions, by producing a more drastic protein dilution, might mask this effect completely, leading to the erroneous interpretation of washout”.

An improved paradigm of Fluid Physiology and Therapy

233

Moore’s four rate factors to consider Moore identified four rate factors whose understanding is essential to rational treatment of haemorrhage and in the use of salt solutions for blood volume replacement. 1. The rate constant for spontaneous plasma volume refill. We now attribute this to absorption of fluid from lymph in lymph nodes, efferent lymph flow via the thoracic duct, and (when capillary pressure is lower than plasma oncotic pressure) autotransfusion by transient reabsorption of interstitial fluid. 2. The dispersal of infused saline. This is the net effect of the rate constant for transendothelial fluid filtration Jv during the saline infusion and the rate constant for afferent lymph flow, as measured by haemodilution in many modern reports. The water in afferent lymph enters the plasma by absorption in lymph nodes and by efferent (thoracic duct) lymph flow. 3. The ingress of protein. The sinusoidal tissue capillaries pose no filtration barrier to the shift of albumin from, for example, the interstitial fluid of the liver. In addition, it is now appreciated that protein-rich lymph enters the central veins and that lymph flow rate increases during haemorrhage. 4. The renal excretion of water and salt. Balance between these four rate factors, Moore explained, determines plasma volume.

The development of volume kinetics In 1968 Adamson and Hillman reported a small but important experiment to examine the value of haematocrit as an index of blood volume deficit. Following pre-labelling of the albumin space by 131iodinated albumin, six male volunteers were phlebotomised 15% to 20% of their blood volume. Blood volume, as reflected by serial determinations of haematocrit values, was restored more slowly than predicted by previous studies. They noted no evidence of reabsorption of solvent and solutes from interstitial fluid to plasma, and the plasma protein concentration remained constant throughout the study. The new plasma proteins were from the interstitial space, and new protein formation contributed little to the restorative process. They concluded

234

Chapter 11

that haematocrit value is a relatively poor index of blood volume deficit following acute blood loss in normal man [4]. This knowledge allows researchers to use haematocrit changes to investigate acute plasma volume response to rapid infusion of intravenous fluid. Lucas and Ledgerwood at Wayne State University, Detroit, have been delivering quality fluid physiology research for decades. One of their earliest clinical reports was of a randomised trial of albuminsupplemented versus non-albumin fluid resuscitation following hypovolaemic shock. The post-resuscitation plasma volume was slightly higher in the albumin group (average 4.0 l versus 3.7 l in nonalbumin patients), while urine output and sodium clearance were less (average 2.5 and 1.1 ml min-1 compared to 3.6 and 2.6 ml min-1 in nonalbumin patients). Albumin patients also experienced prolonged ventilatory dependency, averaging 8 days compared to 3 days [5]. Lucas and Ledgerwood highlighted the potential clinical importance of interstitial fluid space resuscitation in a report on post-resuscitation hypertension in trauma patients in 1981 [6]. A later canine laboratory experiment, using the Wiggers model of haemorrhagic shock, showed that the shocked interstitial fluid space remained contracted during the first hour after resuscitation with hypertonic saline, hypertonic saline with 6% Dextran, and 6% Dextran. Only the non-colloid Ringer’s Lactate solution quickly restored lymph flow and the movement of interstitial proteins to the plasma. The restoration of plasma volume, arterial pressure and cardiac output was greatest with the inclusion of Dextran [7].

Hahn’s Volume Kinetics Robert Hahn introduced his novel approach to the investigation of in vivo volume kinetics in reports published in 1997 [8, 9]. Eight healthy male volunteers were given boluses of 25 ml kg-1 or 5 ml kg-1 6% Dextran in Sodium Chloride or 3 ml kg-1 7.5% Sodium Chloride, and the subsequent dilutional changes in blood haemoglobin or serum albumin were followed. These first experiments also measured blood water mass by drying a known volume of sampled blood. Results suggested that the size of the body fluid space expanded by each of these infusions was around 5.9 l, 2.6 l and 1.2 l respectively. Dilution was most pronounced after the higher volume Ringers Acetate experiments. A two-compartment model was found to fit the data adequately. Dilution

An improved paradigm of Fluid Physiology and Therapy

235

curves after Dextran infusion fitted a single compartment model. With good agreement between the blood water measurement approach and the estimation of plasma volume change by haemoglobin dilution, later experiments focused on the simpler haemoglobin dilution method. The authors expressed the hope that such volume kinetic data would one day enable the prescriber to calculate an infusion strategy that would achieve a desired boost in plasma volume over a desired period of time. Hahn’s accumulated published data have since been used to test a very simplified feedback control model that could obviate the need to incorporate the several physiological mechanisms involved in the PF:IF equilibrium in reproducing blood volume response to fluid infusion [10]. Drobin and Hahn reported their volume kinetic experiments in hypovolaemic volunteers in 1999 [11]. Their experimental method was to administer a large rapid bolus of Ringers Acetate (25 ml kg-1) at normovolaemia and after withdrawal of blood (450 ml or 900 ml). Unlike Moore’s study, their subjects did manifest hypotension during hypovolaemia. In the hypovolaemia experiments, no time was allowed for spontaneous plasma volume refill rate before the Ringers Acetate was given. In most experiments the volume kinetic profile fitted a twocompartment model; a central volume of distribution which presumably was indicative of the intravascular extracellular fluid, and a peripheral compartment which was approximately twice the volume of the central compartment. The authors remarked that the total body fluid space expanded by the infused fluid was substantially less than the extracellular fluid space into which an indicator like bromide is diluted. The concept of an infused isotonic salt solution expanding the whole of the extracellular space is thus untenable, and it has to be appreciated that a significant part of the extracellular space is of structurally fixed volume, unable to expand or constrict with changing kinetic influences. In terms of volume kinetics, the ratio of expandable plasma volume to expandable interstitial space is closer to 1:2, making crystalloids more effective contributors to plasma volume than suggested by Moore’s 1:4 figure which did not take account of nonexpandable interstitial fluid spaces.

Hahn’s rate factors Hahn’s model is a differential equation for the rate of change of the central volume of distribution Vc following intravenous infusion and

Chapter 11

236

applies the following factors: 1. 2. 3. 4.

The infusion rate R0 The urine output (clearance) Cl. … With allowance for ongoing and insensible losses Cl0 In the case of crystalloid infusions, a second “tissue” compartment Vt becomes evident, and the rate of distribution between Vc and Vt is the distribution clearance Cld.

Now if the distribution clearance is mostly determined by net absorption or net filtration in the capillary beds, a single rate representing flow in either direction is the logical approach [12]. But when we accept that water flow from the central to peripheral compartment (presumably intravascular to extravascular extracellular fluid) is mostly due to transendothelial filtration, and the redistribution of fluid back to the plasma is mostly attributable to absorption from lymph and efferent lymph flow, we see they deserve their own rate constants k12 and k21. By this adjustment, Hahn was able detect slowed redistribution of fluid back to the plasma (k21) in patients with appendicitis [13]. By focusing exclusively on water volume, haemodilution kinetics do not take into account protein movement between the compartments. By the nature of the experimental model, the plasma volume refill constant is not taken into consideration.

Applications of Hahn’s volume kinetics Robert Hahn’s experimental method of fluid bolus haemoglobin dilution has been used to investigate several aspects of volume kinetics in his own institution and in collaboration with others. The adequacy of the two-compartment model is frequently confirmed for crystalloid solutions, while a single compartment model describes haemoglobin dilution following Dextran or albumin. A study published in 2005 investigated the effects of two different anaesthetic techniques in patients undergoing thyroid surgery [14]. Induction of anaesthesia with either isoflurane or propofol and fentanyl brought about a fall in the patient’s haemoglobin and plasma albumin concentration, suggesting a plasma dilution of about 7%. Isoflurane-anaesthetised patients had lower arterial pressure and received more frequent rescue with ephedrine. Rapid infusion of around 2 l Ringers Acetate solution caused dilutional anaemia in both groups that was much more

An improved paradigm of Fluid Physiology and Therapy

237

pronounced than previously found in awake volunteers: that is, the crystalloid solution was a much more efficient haemoglobin diluter in the context of anaesthetised subjects undergoing thyroid surgery. The practical implication claimed was that, because of slowed elimination, only 7 ml min-1 (42 ml h-1) of intraoperative crystalloid is necessary to maintain a steady-state plasma expansion of 10%. In this trial, Hahn examined the difference in volume kinetics based on haemoglobin and albumin to indicate the magnitude and direction of translocation of albumin. His data showed that albumin entered the free-flowing plasma during the induction of anaesthesia. Although he noted a marked dilution of albumin during the experimental haemodilution, it was not as great as predicted and therefore implied a further net movement of albumin to the free-flowing plasma. The translocation of albumin in this direction had been observed by Moore during capillary refill after haemorrhage. The most likely contributor of this immediately-accessible albumin is the tissues supplied by capillaries with discontinuous glycocalyx, in particular the Space of Disse (interstitial fluid in the liver), whose volume rises and falls with venous pressure [15]. With discontinuous capillaries and no glycocalyx filter, hepatic interstitial fluid is essentially an extravascular pool of plasma. With falling venous pressure, it will redistribute to the intravascular volume. The slower but perhaps greater contribution (especially during intravenous fluid resuscitation) is efferent lymph flow, of which at least half comes from the liver. It is important to appreciate that the proteins entering the blood stream after haemorrhage are not those one would expect from transendothelial filtration across continuous capillaries [16]. While the ingress of albumin might be welcomed, there are other proteins that might have injurious effects on the pulmonary and systemic endothelium. Of course, some of the albumin appearing in free-flowing plasma could have translocated from the intravascular gel phase associated with glycocalyx. In this trial we see some shortcomings of Hahn’s method. It demonstrates the consequences of a severe disequilibrium state brought about by a potentially harmful rapid intravascular volume boost, but tells us nothing about the physiologic compensatory mechanisms at play or consequences of administering intravenous fluid at a rate that does not exceed what Moore called the dispersal rate of the solution.

238

Chapter 11

In a later retrospective analysis of his extensive volume kinetic data sets, Hahn has noted that the rate of elimination of crystalloid fluid decreased in proportion to mean arterial pressure but was otherwise independent of general anaesthesia and moderate-sized surgery [17]. This observation supports the hypothesis that context sensitivity is a characteristic of the haemodynamic profile, but does not point more specifically to the most responsible segment of the vascular loop. The glycocalyx model paradigm sees the capillaries and uninvested postcapillary venules as the logical site. Changes in the hydrostatic pressure difference will change the net filtration pressure and so the transendothelial filtration rate Jɋ. This effect has been called the capillary fluid shift mechanism. Arteriolar constriction will, to some extent, protect capillary pressure from high-resistance arterial hypertension, but when arterial hypotension is caused by hypovolaemia (which leads to low venular filling pressure), arteriolar constriction will act to keep arterial pressure up. At the same time capillary pressure falls and switches off filtration. If capillary pressure falls below the plasma colloid osmotic pressure, it brings about transient reabsorption, which is called autotransfusion.

Pharmacodynamics of a fluid challenge Hahn’s volume kinetic data are also complicated by the destabilising effect of the rapid infusion on haemodynamics. Context, as always, is important; the haemodynamic response to resuscitation from hypovolaemia is different from the response to iatrogenic hypervolaemia. In patients with little or no myocardial energy excess, venous congestion may be exacerbated. In an observational study of sedated and ventilated post-surgical patients the maximal increase in cardiac output was present just a minute after the end of a 250 ml fluid challenge. In patients classified as non-responders (no increase in cardiac output), central venous pressure rose instead. The associated rise in capillary pressure would have boosted Jv. In responders, the rise in cardiac output increases exposure of blood to the surface area of the microvascular endothelium, another way to boost Jv. In all patients the effect of a fluid challenge on haemodynamics was dissipated by 10 minutes [18]. Slower infusions of fluid stay in the circulation longer than rapid infusions and boluses.

An improved paradigm of Fluid Physiology and Therapy

239

Future volume kinetic research? Perioperative fluid therapy practices have evolved in recent times away from a desire to temporarily boost plasma volume and towards favouring avoidance of boluses and higher intravenous volume administration, making Hahn’s method increasingly irrelevant. Hahn’s method measures dilutional anaemia after a fluid bolus, but not resuscitative efficiency: the two are not related, and it is commonly observed that patients are more anaemic after colloid rescue from hypovolaemia than after crystalloid rescue. There remains a need for an approach to predicting steady-state volume kinetics.

The Glycocalyx Model Paradigm then and now: Terminology There have been some inconsistences in the terminology of microvascular physiology which demand some clarification. There have been many terms used to describe what I call the glycocalyx model. On just one page of J Rodney Levick’s An Introduction to Cardiovascular Physiology 5e we find: • • • •

The Michel-Weinbaum glycocalyx-cleft theory of fluid exchange The Michel-Weinbaum theory The glycocalyx-cleft model The Michel-Weinbaum glycocalyx-junctional break model of fluid exchange.

They are one and the same thing, and I am content to continue to use the term glycocalyx model paradigm to describe my paradigm for rational fluid therapy, while using the term Michel-Weinbaum model for the physiology discourse. The terms transcapillary and transendothelial tend to be used interchangeably. I feel it is important to recognise the contribution of venules to fluid exchange and to haemodynamics, so I prefer the term transendothelial in relation to solvent and solute movement between the plasma and interstitium when venules are involved. The volume flow across a semipermeable membrane is annotated as Jɋ in Kedem-Katchalsky equations, including the Starling equation [19],

240

Chapter 11

but given more biological descriptions in physiology discourse. To use the term filtration instead of volume flow implies a direction of flow, which is arguably appropriate when Levick speaks of the capillary filtration rate in his textbook. I prefer transendothelial solvent filtration. There has been some inconsistency in anatomic terminology. In 2018 Roy Curry identifies a quasiperiodic inner endothelial glycocalyx layer (EGL), less than 300 nm thick, and associated with the endothelial cell membrane. It is the primary molecular filter between circulating blood and the body tissues. Direct optical observations describe endothelial surface layers (ESLs) with porous outer layers that extend 1-2 Ɋm beyond the EGL. Changes in the thickness and distribution of thick ESLs in vessels with diameters larger than 50 Ɋm may not reflect functional changes in the inner endothelial glycocalyx layer [20]. In a separate analysis, Curry and Michel have worked out that albumin accumulates at the interface between the porous layer of the ESL and its selective inner layer, the EGL. The osmotic pressure of accumulated albumin significantly modifies the observed permeability properties of the microvessel wall. This is an example of the effective unstirred layer effect, and goes some way to explaining why the colloid osmotic pressure of free-flowing plasma is a step removed from the colloid osmotic pressure at the ESL to EGL interface. Future work to assess the clinical consequences of this observation are keenly anticipated [21]. The concept of a bilayered endothelial surface layer also points to a sodium regulatory role that has already brought about a further possible revision of the Starling principle [22].

The intravascular space In 2012 we introduced the concept that the intravascular space contains three compartments of interest to fluid prescribers. They are: • • •

Erythrocytes, which contain intracellular fluid. The intravascular aqueous phase within which blood cells circulates: the free-flowing plasma. The intravascular gel phase, which includes the luminal layer of the endothelial surface layer and which excludes erythrocytes.

An improved paradigm of Fluid Physiology and Therapy

241

We defined intravascular fluid volume to be the volume contained by the endothelial cells, and for measurement of which we suggested the Dextran 40 dilution volume. Dextran 40 is a smaller marker molecule which more easily permeates the gel phase of the endothelial surface layer. We proposed that the central volume of distribution of an infused isotonic salt solution approximates to the intravascular volume. The intravascular aqueous phase can be measured by Dextran 70 dilution. This larger molecular marker is largely excluded from the noncirculating endothelial glycocalyx layer and its dilution volume largely consists of the circulating plasma. Radio-labelled albumin or indocyanine green estimates of plasma volume are measuring the aqueous phase not limited by a glycocalyx layer; that is, the intravenous free-flowing plasma and some of the extravascular plasma in the sinusoidal tissues. There is a degree of exclusion of erythrocytes from plasma at the endothelial surface layer boundary, so the volume of distribution of red blood cells is rather less than the Dextran 70 dilution volume. Compaction of the endothelial surface layer can have significant effect on the balance of total intravascular fluid and red blood cell dilution volumes. Endothelial surface layer compaction (volume or thickness reduction) by inflammation or by infusing a hyperoncotic solution can increase the red cell dilution volume without increasing the intravascular volume. This point must be considered in red blood cell dilution studies that purport to indicate changes in intravascular volume.

The heterogeneity of capillaries We pointed out that there is no such thing as a standard capillary barrier separating plasma from interstitial fluid. • • •

Sinusoidal tissues (marrow, spleen, and liver) have discontinuous capillaries and their interstitial fluid is essentially part of the plasma volume. Open fenestrated glomerular capillaries produce the renal glomerular filtrate. Diaphragm fenestrated capillaries in specialized tissues absorb interstitial fluid to plasma.

242

•

Chapter 11

Continuous capillaries are the great majority in the pulmonary and systemic microvasculature and exhibit ‘no re-absorption’ under the Michel-Weinbaum (glycocalyx) model. In these capillaries the continuous endothelial glycocalyx layer is semipermeable to anionic proteins and their concentration in the intercellular clefts below the EGL is very low.

The Starling forces The important Starling forces are the transendothelial pressure difference and the plasma to subglycocalyx colloid osmotic pressure difference. In particular, notice that the colloid osmotic pressure of the general interstitial fluid is not a direct determinant of transendothelial solvent filtration Jv. Jv is much less than predicted by the Starling principle, and the major route for return to the circulation is as lymph. Raising plasma colloid osmotic pressure reduces Jv but does not cause absorption.

The J curve; a hockey stick? We described the cartoon of a plot of Jv against the transendothelial hydrostatic pressure difference as a J curve. At subnormal capillary pressure, Jv is regulated close to zero, a state of minimal filtration to the tissues. Autotransfusion (tissue fluid shift to the plasma) occurs in extreme hypovolaemia, is acute, transient, and limited to about 500 ml in humans. At supranormal capillary pressure, when the colloid osmotic pressure difference is maximal, Jv is proportional to transendothelial hydrostatic pressure difference and the curve rises up. The general principle of two different relationships between Jv and the transendothelial hydrostatic pressure difference has been confirmed by more rigorous mathematics than I am capable of, and described as a hockey stick curve [23]. An advantage of J curve is that includes a concept of the J point, the capillary pressure that is autoregulated by Jv. I would expect the J point to be a determinant of Rothe’s pivotal pressure for haemodynamics [24].

An improved paradigm of Fluid Physiology and Therapy

243

Figure 11.1

The effects of colloid and crystalloid infusions • • •

•

Infused colloid solution is initially distributed through the plasma volume, and infused isotonic salt solution through the intravascular volume At supranormal capillary pressure, infusion of colloid solution preserves plasma colloid osmotic pressure, raises capillary pressure, and increases Jv At supranormal capillary pressure, infusion of isotonic salt solution also raises capillary pressure, but it lowers colloid osmotic pressure and so increases Jv more than the same colloid solution volume At subnormal capillary pressure, infusion of colloid solution increases plasma volume and infusion of isotonic salt solution increases intravascular volume, but Jv remains close to zero in both cases

244

Chapter 11

The interstitium and lymphatics The interstitial fluid is for the most part a gelatinous (gel) phase due to the presence of glycosaminoglycans. Older practitioners may recall these molecules by their earlier name of mucopolysaccharides. The disorders caused by abnormalities of glycosaminoglycan turnover are still called the mucopolysaccharidoses. Acute reduction of transendothelial pressure difference by precapillary vasoconstriction or hypovolaemia can result in transient (non-steadystate) absorption of fluid to the plasma volume, equivalent to as much as a 500 ml autotransfusion in human physiology. This effect lasts only for a few minutes. Autotransfusion was not evident in experiments on volunteers who were moderately venesected (less than a litre) [4]. Absorption reverts to filtration as proteins diffuse into the subglycocalyx space from the interstitium, diminishing the colloid osmotic pressure difference that opposes filtration: this is the glycocalyx model. With less acute or extreme disturbance to the equilibrium, the same mechanism preserves filtration, albeit at just a few millilitres per minute, and the no re-absorption rule applies. The pressure at which Jv approaches zero will depend on capillary porosity, which is the net effect of the various capillaries’ hydraulic conductivities, area for fluid exchange, and the reflection coefficient of the macromolecules determining colloid osmotic pressure. The Jshaped curve describing Jv and mean capillary pressure will be shifted to the left with increased capillary porosity, with the inflection on the curve being the J point. Below the J point, any transfused fluid, whether colloid or crystalloid, will appear to be retained within the intravascular space until the transendothelial pressure difference reaches the level at which filtration recommences. The glycocalyx model and the no absorption rule explain why the colloid osmotic pressure properties of plasma or plasma substitutes add little or nothing to plasma volume resuscitation while transendothelial pressure difference is below the J point. Above the J point, the oncotic pressure difference opposing filtration is maximal and Jv becomes proportional to capillary pressure (or transendothelial pressure difference if the interstitial pressure is not constant). Porosity increases in inflammatory states, but recall that interstitial colloid osmotic pressure has no direct effect on Jv. A 10- to 20-fold

An improved paradigm of Fluid Physiology and Therapy

245

increase in Jv in the acute inflammatory response is actively regulated by integrins acting upon collagen fibrils in the extracellular matrix, exposing GAGs to take up water, and does not necessarily imply increased capillary porosity. The effects of fluid therapies on this mechanism, if any, are unknown. Changes which compact the endothelial surface layer releasing GAGs into the circulating plasma are associated with increased transendothelial protein flux, but compaction of the endothelial surface layer and increased porosity of the glycocalyx layer may be separate processes; the association may not be entirely causal. Although transfused macromolecules do not easily permeate an intact glycocalyx layer, they pass easily into the interstitial fluid through the sinusoidal capillaries in the bone marrow, spleen, and liver, equilibrating with interstitial macromolecules and returning to the venous system via lymphatics. An increase in the proportion of the cardiac output going to sinusoidal tissues will increase Jɋ and the transcapillary escape rate of albumin. There is no significant absorption of interstitial fluid to the plasma under a colloid osmotic pressure difference, so colloid therapy does not prevent or improve tissue oedema.

A two-compartment kinetic model In 2017 I introduced a two-compartment kinetic diagram of extracellular fluid distribution to explain the equilibrium of plasma volume, which owed much to the research model of Robert Hahn [25]. See Figure 11.2. The free-flowing aqueous plasma and red cells are within the intravascular compartment. Lighter shading indicates the gel phase of extracellular fluid, both intravascular and extravascular. There is an interstitial aqueous phase, but in health it is less than 1% of total and not represented in this diagram. Do remember that the proportion of interstitial aqueous phase increases greatly in oedema. Cartoon a shows a healthy intravascular gel phase (endothelial surface layer and glycocalyx) which concentrates the erythrocytes into the aqueous phase. Cartoon b shows reduced volume of the intravascular gel phase and shows that if the intravascular volume is unchanged, the free-flowing aqueous phase (plasma) is increased, causing haemodilution that is not an indicator of the intravascular volume. Many experts presume that the more reduced haematocrit after the

246

Chapter 11

infusion of colloid solutions compared to crystalloid solutions indicates a more expanded intravascular volume. I have called this the Colloid Delusion, with apologies to a colleague who tells me he finds that offensive. The clinical fact is that patients resuscitated with a crystalloid often have significantly higher post-resuscitation haematocrit and are less likely to be given red blood cells when they reach a reduced haematocrit transfusion trigger level. Figure 11.2 reminds us that the peripheral or tissue volume Vt can only expand in tissues that are not rigidly constrained. The volume of distribution of an infused fluid volume is therefore significantly less than the volume of distribution of a marker solute. The central volume Vc depends on the constituents of the infusate under consideration. The initial central volume of distribution of an albumin solution (as assessed by red blood cell dilution) is limited by the endothelial glycocalyx and the sinusoidal tissue interstitium which consists of extravascular plasma. The initial central volume of distribution of a colloid-free isotonic salt solution (as assessed by red blood cell dilution) is limited by the intercellular endothelial tight junctions (represented by the broken line circle), with just a small volume being lost through the occasional junction breaks. It will also include extravascular plasma in the sinusoidal tissues. Parenteral fluid administration goes directly to the central volume of distribution. One could argue that enterally-absorbed fluid crosses an epithelial layer to enter the enteric mucosal interstitium before entering the central volume of distribution, but for simplicity I have shown it entering through the intravascular gel phase. Haemorrhage could be seen as primarily involving the free-flowing plasma, but in practice the endothelial surface layer volume is also reduced. Urine output is the major exit from the central volume.

An improved paradigm of Fluid Physiology and Therapy

Figure 11.2

247

248

Chapter 11

Hahn’s two compartment model for non-colloid solutions shows fluid flux from central to tissue volumes as k12, while the return fluid flux is k21. Appreciating that k12 is mostly the transendothelial solvent filtration rate Jv and that k21 is mostly the afferent lymph flow Qlymph allows us to relabel the diagram accordingly. The diagram also shows that Qlymph is the sum of the fluid volume absorption rate from lymph to plasma in lymph nodes and the efferent lymph flow. The diagram encourages us to think of ways we can rebalance the central and tissue volumes by adjusting Jv or Qlymph rather than the traditional but flawed rationale of infusing a solution presumed to “stay in the circulation”. The pumping of efferent lymph to the central volume is an important compensatory response to haemorrhagic shock in humans.

A revised Twigley-Hillman diagram

Figure 11.3

The new Twigley-Hillman diagram emphasises the extracellular fluid

An improved paradigm of Fluid Physiology and Therapy

249

(ECF) circulation which occurs in most tissues most of the time. The intravascular space normally contains about 5 litres of blood and 1 litre endothelial surface layer (ESL) from which circulating red blood cells are excluded. The intravascular extracellular fluid is free-flowing aqueous (plasma) and gel phase ESL contains the fibre matrix molecules of the endothelial glycocalyx layer which arise from the endothelial cell surface, as well as the outer porous zone. The triphasic interstitial space has a structural collagen fibrous phase and around 14 litres of fluid; an aqueous phase, a gel phase within which glycosaminoglycans such as hyaluronic acid have the capacity to store sodium without raising tissue osmolality, and lymph. The intracellular fluid (ICF) volume, normally about 23 litres, is sensitive to acute changes in ECF osmolality. However, cell volume regulatory mechanisms exist to preserve the steady-state intracellular fluid volume and enable subjects to tolerate chronic hypotonicity. Water and solutes enter and leave the body across epithelial barriers. In clinical practice we can infuse fluids directly into, or haemofilter water out of the free-flowing plasma. In sinusoidal tissues (liver, spleen and bone marrow) the ESL is discontinuous and there are windows (fenestrations) through the endothelium that exclude red blood cells but admit albumin to the interstitial space so that there is no trans-endothelial colloid osmotic pressure (COP) difference to oppose filtration. In non-sinusoidal tissues the continuous endothelial glycocalyx layer is almost impermeable to albumin so that filtered fluid in the immediate subglycocalyx space, also known as the protected region, has a very low COP compared to plasma or the general interstitial fluid. The transendothelial COP difference opposing filtration is therefore high. Michel and Weinbaum correctly hypothesised that if transendothelial water movement across non-sinusoidal capillaries and venules is transiently reversed by a sudden drop in the hydrostatic pressure difference, interstitial albumin rapidly enters the protected region, diminishing the trans-endothelial COP difference. The net water movement therefore quickly return to steady-state filtration. This is the Michel-Weinbaum no-absorption Rule. Solvent is absorbed from afferent lymph into lymph node capillaries and venules. Efferent lymph therefore has high protein and lipid content and returns to the central veins via the thoracic duct.

250

Chapter 11

Acceleration of protein-rich efferent lymph to the circulating blood volume is an important compensatory response to haemorrhagic shock in humans.

An ideal extracellular fluid substitute? As the colloid debate has abated, the attention of clinical investigators has turned to the possible advantages of balanced saline solutions versus normal saline. We consider here the results of two of the larger recent trials. Both trials were cluster randomised, but negative trial results may reflect the relatively small quantity of fluid infused in the two groups: the median quantity in each trial was less than 2 litres. The glycocalyx model paradigm would predict that in such moderate quantities, a small difference in anion composition would have negligible consequence. If the rational prescriber can avoid fluid inundation, the choice of isotonic salt solution need only come down to availability and cost. The SPLIT trial, conducted in four intensive care units, showed no advantage in either group [26]. Some additional data identified an increased proportion of patients receiving blood or blood product with allocation to balanced salt solution compared to Normal Saline [27]. The SMART trial was a single-centre study involving 5 intensive care units and yielded similar results. There was no difference in mortality or kidney injury using balanced solution vs normal saline. A significant difference in favour of PlasmaLyte, a balanced salt solution which is fully isotonic, was found in days free from renal replacement therapy and in a composite outcome of renal complications and mortality [28]. Such secondary outcomes are difficult to accept as real differences. Despite the lack of definitive evidence, balanced solutions have theoretical advantages that should be compared with the risk of hyperchloraemic acidosis after large volume resuscitation with normal saline. Balanced solutions are not blameless, but they are less frequently associated with hyperchloraemia, and are probably the best choice as a first-line fluid therapy in patients likely to need higher resuscitation volumes.

An improved paradigm of Fluid Physiology and Therapy

251

Context sensitivity and Volume equivalence One of the colloid versus crystalloid controversies is apparently conflicting claims about volume equivalence of the two classes of resuscitation fluids. There is no consensus definition of the term volume equivalence. Twigley & Hillman’s reasoning was based on two premises; that Colloids are distributed through the plasma volume (circa 3-4 litres) and that Crystalloids are distributed through the total extracellular fluid volume (circa 15-16 litres). Their Conclusion was that the volume equivalence of colloid to crystalloid will be around 1:5. Their reasoning remains widely accepted and taught. The most obvious flaws in this reasoning lie in the truth of the premises. The first premise depends upon how you define and measure plasma volume. The second premise ignores the fact that the expandable portion of fluid in the interstitium is only around 6-8 litres, and typically it is twice the estimated size of the plasma volume of 3-4 litres [12]. These facts point instead to an expected volume equivalence closer to 1:3. The glycocalyx model paradigm takes as it’s premises that the central volume of distribution of colloids approximates the free-flowing plasma volume (2.5-3.5 litres) while the central volume of distribution of crystalloids approximates the whole of the intravascular fluid volume plus the sinusoidal tissues interstitial fluid (4-5 litres). The Glycocalyx Model Paradigm therefore concludes that the “volume equivalence ratio for resuscitation from hypovolaemia” is less than 1:2. Moreover the “volume equivalence ratio for haemodilution” is greater than 1:2. To the best of my knowledge, this important distinction had not previously been identified. I suggest that Glycocalyx Model Paradigm is superior for fluid prescribers because it explains the findings of randomised controlled clinical comparisons of colloid and crystalloid infusions. How then do we explain the widely-differing volume equivalence claims from other experiments? A key consideration here is context sensitivity, which depends on the balance of Starling forces pertaining at the time of volume infusion and the rate at which any given volume of any given intravenous fluid is administered. The FEAST trial was an alarm bell warning us that rapid infusions are associated with increased rate of death [29]. Fluid bolus therapy has been a cornerstone of fluid resuscitation, but is now under a shadow of doubt about its safety.

252

Chapter 11

In the randomised controlled clinical trials that enrol patients with suspected hypovolaemia, the truly hypovolaemic patients (hopefully the majority of participants) will have a hydrostatic pressure difference across the capillary/ venular barrier that is less than the normal (i.e. below the J point). The steady-state Starling principle describes the limitations applying to microvascular absorption of tissue fluids to the plasma volume. A trickle of filtration persists at all sub-J point οP values and substantial Jv is not resumed until οP is above the J point. Infused Colloids increase the free-flowing plasma volume while crystalloids increase the whole intravascular volume and sinusoidal tissue interstices. Colloids therefore have a greater haemodilution effect than crystalloids at similar pressure - volume states of the blood vessels. In an ideal experiment the ideal end point would be return of the transendothial hydrostatic pressure difference οP to the J point. This should minimise oedema while providing enough venular filling pressure (or mean systemic pressure) for homeostatic control of haemodynamics. However, other end points such as arterial pressure, stroke volume, blood volume or cardiac output, are often selected. All of these end points are achieved at a much higher resuscitation dose (i.e. volume) than is strictly necessary to restore the subject to a condition that is optimal for complication-free recovery. Let us consider some illustrative examples from the literature in the era of a revised Starling principle. Lobo et al performed their experiment on ten healthy, nonanaesthetised and spontaneously breathing male participants. We can presume that all subjects were at their autoregulated J point when the study fluid was infused, and that οP rose along with Jv after infusion. We would expect return to the J point to be quickest in the crystalloid group and retarded by higher colloid osmotic pressure in the colloid groups. This prediction was confirmed. Interestingly, though dilution studies pointed to prolonged hypervolaemia in the colloid studies, plasma concentrations of hormones regulating sodium and water balance did not change [30]. Jacob M. et al 2012 claimed rather dramatically that “the intravascular volume effect of Ringer’s lactate is below 20%.” They conducted an experiment on women under anaesthesia and controlled positive pressure ventilation. They removed about a litre of blood while simultaneously administering three times the volume of Ringers

An improved paradigm of Fluid Physiology and Therapy

253

Lactate solution, an amount empirically chosen as intermediate between the Twigley-Hillman physiology-predicted 1:5 “volume effect“ and the clinical trials consensus of 1:1.5. Blood pressure and heart rate remained steady throughout, which should have been interpreted as proving that 1:3 is perfectly safe and effective for a euvolaemic haemodilution protocol. Plasma volume was measured as the volume of distribution of indocyanine green (which binds to albumin) and red cell volume using sodium fluorescein-labelled autologous red cells. Blood volume was calculated as the sum of these two values. The crystalloid exchange fluid was fully effective at protecting the plasma volume during haemorrhage, while the red cell volume was unsurprisingly diminished. Then the researchers transfused the women with albumin, with the stated intention of achieving total blood volume the same as baseline. The logic of this is highly questionable; the experiment had shown that plasma volume preservation had been achieved and the participants showed no signs of hypovolaemia. The subsequent albumin infusion was therefore not indicated, and the headline conclusion unjustified. If blood volume restoration was the protocol objective, the logical transfusion was red blood cells [31]. Yates et al reported the first blinded comparison of colloid (hydroxyethyl starch) and crystalloid (Hartmann’s solution) in patients subjected to a fluid optimisation protocol while under anaesthesia and positive pressure ventilation [32]. The goal was to achieve stroke volume variation less than 10%. Hahn’s experiments show that patients under anaesthesia behave as though their delta P has been reduced below the J point: the rate at which fluid leaves their circulation after a bolus is substantially reduced. Such a model is therefore a good one for quantifying the volume equivalence ratio for resuscitation from hypovolaemia. They arrived at an overall ratio of 1:1.7 and declared the 1:5 prediction obsolete. We see from their results the haemodilution effect of the colloid, making the colloid was less effective for oxygen delivery than the crystalloid. Clinical outcomes were not different. Heming et al reported a subgroup analysis of patients from their CRISTAL clinical trial who were under pulmonary artery catheter monitoring. The median cumulative volume of fluid administered during the first 7 days in the ICU was higher in the crystalloids than in the colloids arm (3500 (2000–6000) vs 2500 (1000–4000) ml [33].

254

Chapter 11

This is clearly in line with glycocalyx model paradigm expectations. Thirty anaesthetised and ventilated pigs were bled until stroke volume was 50% of the baseline value, then resuscitated with the study fluid (hydroxyethyl starch or Ringerfundin) to baseline stroke volume [34]. Regrettably the researchers were not blinded to fluid type and, for reasons not explained, the animals randomised to starch were ventilated to lower PaCO2 and higher pHa. Shock was reversed in all animals, while lactate was at all times lower in the crystalloid-rescued pigs and the colloid-rescued pigs were more anaemic and tachycardic at study end. Time to restore haemodynamic stability was said to be shorter in the colloid-rescued group, which was claimed to be an advantage even though clinical trials have found more lives saved by slower and less aggressive shock reversal. The volume replacement ratio by this experimental design averaged 1:3. This value was claimed to be compatible with a so-called “Starling’s three compartment model” puzzlingly attributed to me. As noted above, Twigley and Hillman’s three-compartment model actually suggests a volume equivalence of 1:5. It was claimed that these pigs represented patients in the earliest stage of shock while the permeability barrier was still intact, yet the animals also had pulmonary oedema attributed by the researchers to prolonged and stressful pre-experiment preparation.

Understanding leaky capillaries There are rare cases in which there is whole-body failure of the microvascular permeability barrier presenting as life-threatening hypovolaemic shock and generalised oedema. There are many more clinical challenges of critical care that are attributed to what we call leaky capillaries. Pushed to elaborate, even the experts may recall ‘the reflection coefficient’, but few can elaborate further. The in vivo capillary filtration unit needs almost continuous maintenance and tonic regulation of its permeability [35]. The key regulating parts of the capillary filtration unit are; • the adherens junctions, • the continuity of the tight junctional strands (length and number of breaks), and • the balance of synthesis and degradation of glycocalyx layer

An improved paradigm of Fluid Physiology and Therapy

255

components. The small GTP-ase enzymes which maintain the near-impermeability of the capillary filtration unit require a continuous supply of intracellular cyclic adenosine monophosphate and a continuous supply of sphyngosine-1-phosphate (S1P), generated within circulating erythrocytes and carried to the endothelium by plasma albumin. Inflammation reduces activation of the small GTP-ases and so reduces the near-impermeability of the capillary filtration unit. Kedem-Katchalsky Equation 1 explains the transvascular solvent filtration rate Jv in terms which describe the hydrostatic and colloid osmotic pressure differences across the semipermeable microcirculation. Less often taught, but equally important to understanding the pathophysiology, is the second Kedem-Katchalsky equation explaining a solute transfer rate Js as the sum of the mass of that solute carried with the microvascular filtrate (convection) and the mass of that solute that permeates the microcirculation independently of flow (diffusion). In clinical considerations the solute of interest is albumin. Researchers who measure Js of albumin or another marker molecule in disease states often presume that Jv will be increased with Js and cause oedema. Figure 11.4 is a cartoon illustrating some of the ways in which inflammation affects Js and Jv.

256

Figure 11.4

Chapter 11

An improved paradigm of Fluid Physiology and Therapy

257

Figure 11.4 illustrates, from top to bottom, how continuous capillaries can leak in health and with endothelial dysfunction of increasing severity. Jv is the transendothelial solvent filtration rate, and Js is the transendothelial solute transfer rate, of which the major contributor is the permeability to albumin. The hydrostatic pressure difference across the microvasculature is a key regulator, and can increase Jv twenty-fold with no change in the barrier components. In microvasculature with a normal glycocalyx, Staverman’s reflection coefficient for albumin less than 1.0 explains increased albumin flux with no permeability change. The series on the left shows the deterioration of paracellular barrier function through glycocalyx dysfunction and disintegration, tight junction gaps surface area increasing, and adherens junction proteins disconnecting. At the extreme, a full inflammatory gap occurs: these are most often observed in post-capillary venules, and will leak just like a sinusoidal tissue capillary fenestration. The right column illustrates the deterioration of transcellular barrier function, using caveolation of albumin molecules as an example.

A glycocalyx model paradigm of haemodynamic control Optimal haemodynamic therapy would ensure adequate microcirculation with oxygen-carrying blood, and create an optimal pressure gradient for the recirculation of interstitial fluid from capillaries to veins. The heart is the primary pump of blood around the cardiovascular circuit, but not the only motivator. The important parameters of the distributive blood vessel system are arterial pressures and blood flow, largely regulated by resistance characteristics. The capillaries and collecting venules are the microcirculation, the blood-tissue interface where vital transfer of solutes occur, and the important parameters here are transmural pressure and stop-flow characteristics. The important parameters of the collective blood vessel system are venular and venous pressures and volume. There are measurable arterial capacitances and venous resistances, but they are of minor importance in a rational approach to haemodynamic therapeutics. The highest venous-side pressure is the venular filling pressure (possibly measurable as the mean systemic filling pressure or non-invasively by coextensive bio-impedance plethysmography), and the volume of

258

Chapter 11

blood available to the relaxing right ventricle is the venous excess. The venous excess can be evaluated by inspecting the distension of large veins. The RUSH ultrasound approach to the emergency assessment of a patient presenting in shock looks at “the tank”; the volume and collapsing or non-collapsing state of the inferior cava [36]. Note that as compliance of the large veins is high, changes in large and central venous volume bring about only small changes in large and central venous pressure, making central venous pressure (CVP) a poor indicator of the volume of blood available to satisfy the diastolic ventricle. The value of CVP monitoring lies in the early detection of right heart failure, when the ventricle no longer drinks its fill of the venous excess and comes to need a pressure contribution to its preload, or supraphysiological iatrogenic hypervolaemia (fluid overload). A very low CVP is found when venous excess is critically low (hypovolaemia), but there are better ways to make the diagnosis of hypovolaemia! The extra-cardiac pumps that support the venous excess volume include the contraction and relaxation of somatic muscles in the arms and legs, the compliance of venules and small veins, and the respiratory pump. The valved lymphatic pump makes contributions by returning interstitial fluid to veins and lowering interstitial pressure, thereby increasing transmural vascular pressure differences. The glycocalyx model haemodynamic paradigm expands the therapeutic options for resuscitation from arterial hypotension beyond ‘fill or squeeze’. For example, it invites further investigations into the possibilities of selective venoconstrictors and venodilators such as dihydroergotamine [37], alpha 2 agonists, and nitrates to support venous excess without ischaemic arteriolar constriction in the management of sepsis. The possibilities of drug-regulated compliance changes in the venular and small vein segments also need to be investigated.

Disequilibrium events We have noted that a rapid reduction of capillary hydrostatic pressure, whether by reduction of blood volume or increase of arteriolar resistance, is followed by a transient reversal of Jv from filtration to absorption. Michel and Weinbaum solved the mystery of how the steady-state is rapidly re-established at a new lower-pressure

An improved paradigm of Fluid Physiology and Therapy

259

equilibrium. Disequilibrium (from Latin, translated as ‘no equal balance’) occurs in other acute events. Rapid increase of capillary hydrostatic pressure leads to hyperfiltration, as evidenced by changing haematocrit after a large intravenous bolus of fluid and is an avoidable iatrogenic disequilibrium event. A rapid deterioration of the permeability barrier contributes to the pathophysiology of allergic and inflammatory disequilibrium events. The discovery of non-osmotic sodium binding by interstitial glycosaminoglycans raises the possibility of sudden change in interstitial osmolality as a contributor to disequilibrium events such as blood pressure dipping, flash pulmonary oedema, rapid blood loss, burns, and sepsis [38]. The mathematician René Thom suggested that a number of biological processes which displayed discontinuity could be explained by catastrophe theory. Describing the equilibrium between several drivers as a surface with a folded cusp, there is a zone of instability within which a sudden flip from one equilibrium condition to another can occur. As a practical example, a disequilibrium event affecting pulmonary permeability leads to increased Jv incompletely counterbalanced by increased lymph flow. Interstitial volume and pressure rise slowly, but eventually begin to restrict lymph flow by compressing lymph vessels. As the equilibrium of lower interstitial fluid volume enters the zone of instability it can at any moment flip to an equilibrium of higher volume or oedema. Salvador Dalí created a series of works based on the catastrophe theory of René Thom. The Swallow's Tail — Series of Catastrophes (La queue d'aronde — Série des catastrophes) was his last painting, completed in May 1983.

An alternative look at systemic inflammatory haemodynamics What follows is an example of how a sequence of disequilibrium events could be used to explain clinical observations. It is not to be interpreted as a proven hypothesis! Our starting point is the presence of an inflammatory source within the interstitium somewhere. 1. Integrins on the tissue side of the affected tissue are stimulated to cause water uptake and collagen conformational changes resulting in lowered interstitial pressure, raised οP and increased Jv. 2. Local lymphatic pumping is stimulated, also contributing to the

260

Chapter 11

reduction in interstitial pressure. 3. Inflammatory mediators generated in the interstitium enter the blood stream from increased lymph flow from the inflamed tissue and bring about systemic vasodilation and increased circulatory blood flow. 4. Capillary pressure rises, initiating generalised hyperfiltration (raised Jv). 5. Lymph flow more generally is increased, and raised lymph flow carrying inflammatory mediators from mesenteric tissues is a particular motor of systemic inflammation. 6. Generalised microvascular endothelial permeability increase accelerates the shift of intravascular albumin to the interstitium; the colloid osmotic pressure difference restraining Jv is reduced. 7. A new steady state eventually establishes as plasma volume falls (Jv exceeding the afferent lymph flow Qlymph), bringing the capillary pressure down and stabilising the initial hyperfiltration.

Concluding remarks It is still true to say that fluid resuscitation studies require us to reappraise the basics. The glycocalyx model paradigm has its foundations in the steady-state Starling principle which is gaining acceptance by clinical physiologists. Nonetheless, colloids are still being recommended by a few experts and prescribed for resuscitation from hypovolaemia, despite evidence-based protocols and guidelines favouring crystalloids. An important feature of the glycocalyx model paradigm is that it explains why albumin or plasma substitutes have no advantage over isotonic salt solutions when the transendothelial pressure difference οP is low. The important laboratory finding that general interstitial colloid osmotic pressure has little effect on Jv focuses our attention on the subglycocalyx space, the protected region. The endothelial surface layer is a fragile structure and its porous outer layer, along with the membrane-bound endothelial glycocalyx layer, is disrupted by acute hyperglycaemia, surgery, and sepsis. To this list we can add rapid intravenous infusion of fluids, and see an urgent need to reconsider the place of fluid bolus therapy in resuscitation protocols. The glycocalyx model describes how the subglycocalyx colloid osmotic pressure, the transendothelial hydrostatic pressure difference, and Jv balance one another, and raises concerns about disease processes or plasma substitute therapies that might disturb the protected low colloid

An improved paradigm of Fluid Physiology and Therapy

261

osmotic pressure of the subglycocalyx space. In the absence of absorption by capillaries and venules, filtered fluid returns to the circulation mostly by lymphatics, and the importance of preserving lymphatic flow is highlighted in this book. The new paradigm provides an explanation of context sensitivity of colloid and crystalloid volume kinetics in awake, anaesthetized, or hypotensive patients, and the rational prescriber will consider the desired effect on capillary pressure and the associated transendothelial pressure difference. Endothelial dysfunction associated with increased capillary porosity increases Jv at any capillary pressure, and lowers the capillary pressure at which Jv approaches zero. This J point can be taken into account when faced with a patient with systemic inflammation or sepsis. In this book I have raised the importance of mean systemic pressure, or venular filling pressure, or Rothe’s pivot point to the understanding of haemodynamics. We would expect the J point to be very close to the circulatory pivot point at steady-state. The J curve helps us to appreciate how fluid bolus therapy creates a disequilibrium and to anticipate the pathway to a new steady-state. It is likely that the revised Starling equation and glycocalyx model paradigm as presented here will be modified and refined in the light of physiology and clinical trial evidence. In its current form, it strengthens the arguments for preferring isotonic salt solutions over plasma or plasma substitutes for resuscitation, but accepts a rational use of colloids for euvolaemic or hypervolaemic haemodilution. The use of plasma or plasma substitutes to achieve a sustained supranormal plasma volume or to reduce tissue oedema is not rational.

References 1. Moore FD, Dagher FJ, Boyden CM, Lee CJ, Lyons JH. Hemorrhage in normal man. I. distribution and dispersal of saline infusions following acute blood loss: clinical kinetics of blood volume support. Ann Surg. 1966;163:485-504. 2. Lloyd SJ, Boulanger BR, Johnston MG. The lymphatic circulation plays a dynamic role in blood volume and plasma protein restitution after hemorrhage. Shock. 1996;5:416-423. 3. Boulanger BR, Lloyd SJ, Walker M, Johnston MG. Intrinsic pumping of mesenteric lymphatics is increased after hemorrhage in awake sheep. Circ Shock. 1994;43:95-101.

ʹ͸ʹ

Šƒ’–‡”ͳͳ

ͶǤ †ƒ•‘ ǡ ‹ŽŽƒ Ǥ Ž‘‘† ˜‘Ž—‡ ƒ† ’Žƒ•ƒ ’”‘–‡‹ ”‡’Žƒ…‡‡– ˆ‘ŽŽ‘™‹‰ ƒ…—–‡ „Ž‘‘† Ž‘•• ‹ ‘”ƒŽ ƒǤ Ǥ ͳͻ͸ͺǢʹͲͷǣ͸ͲͻǦ͸ͳʹǤ ͷǤ —…ƒ• ǡ ‡ƒ˜‡” ǡ ‹‰‰‹•  ǡ ‡†‰‡”™‘‘† ǡ ‘Š•‘ ǡ ‘—™ƒǤˆˆ‡…–•‘ˆƒŽ„—‹˜‡”•—•‘ǦƒŽ„—‹”‡•—•…‹–ƒ–‹‘ ‘ ’Žƒ•ƒ ˜‘Ž—‡ ƒ† ”‡ƒŽ ‡š…”‡–‘”› ˆ—…–‹‘Ǥ  ”ƒ—ƒǤ ͳͻ͹ͺǢͳͺǣͷ͸ͶǦͷ͹ͲǤ ͸Ǥ ƒ™•‘ ǡ —…ƒ• ǡ ‡†‰‡”™‘‘† Ǥ Ž–‡”‡† ‹–‡”•–‹–‹ƒŽ ˆŽ—‹† •’ƒ…‡ †›ƒ‹…• ƒ† ’‘•–”‡•—•…‹–ƒ–‹‘ Š›’‡”–‡•‹‘Ǥ ”…Š —”‰Ǥ ͳͻͺͳǢͳͳ͸ǣ͸ͷ͹Ǧ͸͸ʹǤ ͹Ǥ ƒš‡ ǡ ‘„‹ ǡ —…ƒ•  ǡ ‡†‰‡”™‘‘† ǡ —…ƒ• Ǥ ‡‘†›ƒ‹…ǡ’Žƒ•ƒ˜‘Ž—‡ǡƒ†’”‡‘†ƒŽ•‹Ž›’Š”‡•’‘•‡• –‘ ˜ƒ”‹‡† ”‡•—•…‹–ƒ–‹‘ ”‡‰‹‡•Ǥ  ”ƒ—ƒǤ ͳͻͻ͸ǢͶͳǣʹͺ͵ǦͻǢ †‹•…—••‹‘ʹͺͻǤ ͺǤ –¤ŠŽ‡ǡ‹Ž••‘ǡƒŠ Ǥ‘†‡ŽŽ‹‰–Š‡˜‘Ž—‡‘ˆ‡š’ƒ†ƒ„Ž‡ „‘†› ˆŽ—‹† •’ƒ…‡• †—”‹‰ ‹Ǥ˜Ǥ ˆŽ—‹† –Š‡”ƒ’›Ǥ ”  ƒ‡•–ŠǤ ͳͻͻ͹Ǣ͹ͺǣͳ͵ͺǦͳͶ͵Ǥ ͻǤ ˜‡•±ǡƒŠ Ǥ‘Ž—‡‹‡–‹…•‘ˆ‹‰‡”•‘Ž—–‹‘ǡ†‡š–”ƒ͹Ͳǡ ƒ† Š›’‡”–‘‹… •ƒŽ‹‡ ‹ ƒŽ‡ ˜‘Ž—–‡‡”•Ǥ ‡•–Š‡•‹‘Ž‘‰›Ǥ ͳͻͻ͹Ǣͺ͹ǣʹͲͶǦʹͳʹǤ ͳͲǤ‹‰Šƒ‹ƒ ǡ ‡‹•‡” ǡ ƒŠ Ǥ  —’‡†Ǧƒ”ƒ‡–‡” —„Œ‡…–Ǧ ’‡…‹ˆ‹…‘†‡Ž‘ˆŽ‘‘†‘Ž—‡‡•’‘•‡–‘ Ž—‹† ˆ—•‹‘Ǥ ”‘– Š›•‹‘ŽǤʹͲͳ͸Ǣ͹ǣ͵ͻͲǤ ͳͳǤ”‘„‹ ǡ ƒŠ  Ǥ ‘Ž—‡ ‹‡–‹…• ‘ˆ ‹‰‡”ǯ• •‘Ž—–‹‘ ‹ Š›’‘˜‘Ž‡‹…˜‘Ž—–‡‡”•Ǥ‡•–Š‡•‹‘Ž‘‰›ǤͳͻͻͻǢͻͲǣͺͳǦͻͳǤ ͳʹǤƒŠ  Ǥ ‘Ž—‡ ‹‡–‹…• ˆ‘” ‹ˆ—•‹‘ ˆŽ—‹†•Ǥ ‡•–Š‡•‹‘Ž‘‰›Ǥ ʹͲͳͲǢͳͳ͵ǣͶ͹ͲǦͶͺͳǤ ͳ͵Ǥ‹ ǡ ‹ ǡ Š— ǡ ƒŠ  Ǥ ‘Ž—‡ ‹‡–‹…• ‘ˆ ‹‰‡”ǯ• Žƒ…–ƒ–‡ •‘Ž—–‹‘‹ƒ…—–‡‹ˆŽƒƒ–‘”›†‹•‡ƒ•‡Ǥ” ƒ‡•–ŠǤʹͲͳͺǢͳʹͳǣͷ͹ͶǦ ͷͺͲǤ ͳͶǤ™ƒŽ†••‘ ǡ ƒŠ  Ǥ ‹‡–‹…• ƒ† ‡š–”ƒ˜ƒ•…—Žƒ” ”‡–‡–‹‘ ‘ˆ ƒ…‡–ƒ–‡†”‹‰‡”ǯ••‘Ž—–‹‘†—”‹‰‹•‘ˆŽ—”ƒ‡‘”’”‘’‘ˆ‘Žƒ‡•–Š‡•‹ƒ ˆ‘”–Š›”‘‹†•—”‰‡”›Ǥ‡•–Š‡•‹‘Ž‘‰›ǤʹͲͲͷǢͳͲ͵ǣͶ͸ͲǦͶ͸ͻǤ ͳͷǤƒ””‘™ƒ ǡ ‡””› ǡ ˜‹‡–›• ǡ ”ƒ‰‡” Ǥ š…Ž—•‹‘ ’Š‡‘‡‘‹–Š‡Ž‹˜‡”‹–‡”•–‹–‹—Ǥ Š›•‹‘ŽǤͳͻͺʹǢʹͶ͵ǣ ͶͳͲǦ ͶǤ ͳ͸Ǥœ‹‡…‹ƒ–‘™•ƒǡǯŽ‡••ƒ†”‘ǡ‘‘”‡‡–ƒŽǤ›’Š‹•‘–ƒ ’Žƒ•ƒ—Ž–”ƒˆ‹Ž–”ƒ–‡ǣƒ’”‘–‡‘‹…ƒƒŽ›•‹•‘ˆ‹Œ—”‡†’ƒ–‹‡–•ǤŠ‘…Ǥ ʹͲͳͶǢͶʹǣͶͺͷǦͶͻͺǤ ͳ͹ǤƒŠ  Ǥ ”–‡”‹ƒŽ ”‡••—”‡ ƒ† –Š‡ ƒ–‡ ‘ˆ Ž‹‹ƒ–‹‘ ‘ˆ ”›•–ƒŽŽ‘‹† Ž—‹†Ǥ‡•–ŠƒŽ‰ǤʹͲͳ͹ǢͳʹͶǣͳͺʹͶǦͳͺ͵͵Ǥ 

‹’”‘˜‡†’ƒ”ƒ†‹‰‘ˆ Ž—‹†Š›•‹‘Ž‘‰›ƒ†Š‡”ƒ’›

ʹ͸͵

ͳͺǤ›ƒ ǡ –‡” ǡ Ž‡–…Š‡” ǡ ”‘—†• ǡ Š‘†‡• ǡ ‡……‘‹ Ǥ Šƒ”ƒ…‘†›ƒ‹… ƒŽ›•‹• ‘ˆ ƒ Ž—‹† ŠƒŽŽ‡‰‡Ǥ ”‹– ƒ”‡ ‡†Ǥ ʹͲͳ͸ǢͶͶǣͺͺͲǦͺͻͳǤ ͳͻǤ ƒ”œ›Ñ•ƒ ǡ ‹‡–”—•œƒ Ǥ ‡”‹˜ƒ–‹‘ ‘ˆ ”ƒ…–‹…ƒŽ ‡†‡ Ǧ ƒ–…ŠƒŽ•›“—ƒ–‹‘•ˆ‘”‡„”ƒ‡—„•–ƒ…‡”ƒ•’‘”–ǤŽ†ƒ† ‡™‘…‡’–•‘ˆŠ›•‹…•ǤʹͲͲͺǢͷǣͶͷͻǦͶ͹ͶǤ ʹͲǤ—””› Ǥ Š‡ ‘Ž‡…—Žƒ” –”—…–—”‡ ‘ˆ –Š‡ †‘–Š‡Ž‹ƒŽ Ž›…‘…ƒŽ›š ƒ›‡”ȋ Ȍƒ†—”ˆƒ…‡ƒ›‡”•ȋȌ‘†—Žƒ–‹‘‘ˆ”ƒ•˜ƒ•…—Žƒ” š…Šƒ‰‡Ǥ†˜š’‡†‹‘ŽǤʹͲͳͺǢͳͲͻ͹ǣʹͻǦͶͻǤ ʹͳǤ—””› ǡ‹…Š‡ŽǤŠ‡‡†‘–Š‡Ž‹ƒŽ‰Ž›…‘…ƒŽ›šǣƒ””‹‡”ˆ—…–‹‘• ˜‡”•—• ”‡† …‡ŽŽ Š‡‘†›ƒ‹…•ǣ  ‘†‡Ž ‘ˆ •–‡ƒ†› •–ƒ–‡ —Ž–”ƒˆ‹Ž–”ƒ–‹‘–Š”‘—‰Šƒ„‹ǦŽƒ›‡”ˆ‘”‡†„›ƒ’‘”‘—•‘—–‡”Žƒ›‡”ƒ† ‘”‡ •‡Ž‡…–‹˜‡ ‡„”ƒ‡Ǧƒ••‘…‹ƒ–‡† ‹‡” Žƒ›‡”Ǥ ‹‘”Š‡‘Ž‘‰›Ǥ ʹͲͳͻ ʹʹǤ ‹ƒ‰ǡ‡–‹‘•ǡ—‘Ǥ‹…”‘˜ƒ•…—Žƒ” ‘”ƒ•’‘”––Š”‘—‰Š †‘–Š‡Ž‹ƒŽ Ž›…‘…ƒŽ›š ƒ›‡”ǣ ‡™ ‡…Šƒ‹• ƒ† ’”‘˜‡† –ƒ”Ž‹‰”‹…‹’Ž‡Ǥ Š›•‹‘Ž‡ƒ”–‹”…Š›•‹‘ŽǤʹͲͳͻ ʹ͵Ǥ Ǥ Š‡ ‡˜‹•‡† –ƒ”Ž‹‰ ”‹…‹’Ž‡ ƒ† –• ‡Ž‡˜ƒ…‡ –‘ ‡”‹‘’‡”ƒ–‹˜‡ Ž—‹† ƒƒ‰‡‡–Ǥ ǣ ƒ”ƒ‰ ǡ —”œ ǡ ‡†‹–‘”•Ǥ ‡”‹‘’‡”ƒ–‹˜‡ Ž—‹†ƒƒ‰‡‡–Ǥ’”‹‰‡”ǢʹͲͳ͸Ǥ ʹͶǤ‘–Š‡  Ǥ ‡ƒ …‹”…—Žƒ–‘”› ˆ‹ŽŽ‹‰ ’”‡••—”‡ǣ ‹–• ‡ƒ‹‰ ƒ† ‡ƒ•—”‡‡–Ǥ ’’ŽŠ›•‹‘ŽȋͳͻͺͷȌǤͳͻͻ͵Ǣ͹ͶǣͶͻͻǦͷͲͻǤ ʹͷǤ‘‘†…‘…ǤŽƒ•ƒ˜‘Ž—‡ǡ–‹••—‡‘‡†‡ƒǡƒ†–Š‡•–‡ƒ†›Ǧ•–ƒ–‡ –ƒ”Ž‹‰’”‹…‹’Ž‡Ǥ †—…ƒ–‹‘ǤʹͲͳ͹Ǣͳ͹ǣ͹ͶǦ͹ͺǤ ʹ͸Ǥ‘—‰ǡƒ‹Ž‡›ǡ‡ƒ•Ž‡›‡–ƒŽǤˆˆ‡…–‘ˆƒ—ˆˆ‡”‡†”›•–ƒŽŽ‘‹† ‘Ž—–‹‘ ˜• ƒŽ‹‡ ‘ …—–‡ ‹†‡› Œ—”› ‘‰ ƒ–‹‡–• ‹ –Š‡ –‡•‹˜‡ ƒ”‡ ‹–ǣ Š‡   ƒ†‘‹œ‡† Ž‹‹…ƒŽ ”‹ƒŽǤ Ǥ ʹͲͳͷǢ͵ͳͶǣͳ͹ͲͳǦͳ͹ͳͲǤ ʹ͹Ǥ‡††› ǡ ƒ‹Ž‡›  ǡ ‡ƒ•Ž‡›  ‡– ƒŽǤ ˆˆ‡…– ‘ˆ ͲǤͻΨ ƒŽ‹‡ ‘” Žƒ•ƒǦ›–‡ͳͶͺƒ•”›•–ƒŽŽ‘‹† Ž—‹†Š‡”ƒ’›‹–Š‡ –‡•‹˜‡ƒ”‡ ‹–‘Ž‘‘†”‘†—…–•‡ƒ†‘•–‘’‡”ƒ–‹˜‡Ž‡‡†‹‰ˆ–‡”ƒ”†‹ƒ… —”‰‡”›Ǥ ƒ”†‹‘–Š‘”ƒ…ƒ•…‡•–ŠǤʹͲͳ͹Ǣ͵ͳǣͳ͸͵ͲǦͳ͸͵ͺǤ ʹͺǤ‡Ž‡” ǡ ‡Žˆ ǡ ƒ†‡”‡”  ‡– ƒŽǤ ƒŽƒ…‡† ”›•–ƒŽŽ‘‹†• ˜‡”•—• ƒŽ‹‡ ‹ ”‹–‹…ƒŽŽ› ŽŽ †—Ž–•Ǥ  ‰Ž  ‡†Ǥ ʹͲͳͺǢ͵͹ͺǣͺʹͻǦ ͺ͵ͻǤ ʹͻǤƒ‹–Žƒ† ǡ ‹‰—Ž‹ ǡ ’‘ƒ  ‡– ƒŽǤ‘”–ƒŽ‹–›ƒˆ–‡”ˆŽ—‹† „‘Ž—• ‹ ˆ”‹…ƒ …Š‹Ž†”‡ ™‹–Š •‡˜‡”‡ ‹ˆ‡…–‹‘Ǥ  ‰Ž  ‡†Ǥ ʹͲͳͳǢ͵͸ͶǣʹͶͺ͵ǦʹͶͻͷǤ ͵ͲǤ‘„‘ǡ–ƒ‰ƒǡŽ‘›•‹—•‡–ƒŽǤˆˆ‡…–‘ˆ˜‘Ž—‡Ž‘ƒ†‹‰™‹–Š ͳ Ž‹–‡” ‹–”ƒ˜‡‘—• ‹ˆ—•‹‘• ‘ˆ ͲǤͻΨ •ƒŽ‹‡ǡ ͶΨ •—……‹›Žƒ–‡† ‰‡Žƒ–‹‡ ȋ ‡Ž‘ˆ—•‹‡Ȍ ƒ† ͸Ψ Š›†”‘š›‡–Š›Ž •–ƒ”…Š ȋ‘Ž—˜‡Ȍ ‘ 

ʹ͸Ͷ

Šƒ’–‡”ͳͳ

„Ž‘‘† ˜‘Ž—‡ ƒ† ‡†‘…”‹‡ ”‡•’‘•‡•ǣ ƒ ”ƒ†‘‹œ‡†ǡ –Š”‡‡Ǧ™ƒ› …”‘••‘˜‡”•–—†›‹Š‡ƒŽ–Š›˜‘Ž—–‡‡”•Ǥ”‹–ƒ”‡‡†ǤʹͲͳͲǢ͵ͺǣͶ͸ͶǦ Ͷ͹ͲǤ ͵ͳǤ ƒ…‘„ ǡ Šƒ’’‡ŽŽ ǡ ‘ˆƒǦ‹‡ˆ‡”  ‡– ƒŽǤ Š‡ ‹–”ƒ˜ƒ•…—Žƒ” ˜‘Ž—‡‡ˆˆ‡…–‘ˆ‹‰‡”ǯ•Žƒ…–ƒ–‡‹•„‡Ž‘™ʹͲΨǣƒ’”‘•’‡…–‹˜‡•–—†› ‹Š—ƒ•Ǥ”‹–ƒ”‡ǤʹͲͳʹǢͳ͸ǣͺ͸Ǥ ͵ʹǤƒ–‡•ǡƒ˜‹‡• ǡ‹Ž‡”ǡ‹Ž•‘ Ǥ”›•–ƒŽŽ‘‹†‘”…‘ŽŽ‘‹†ˆ‘” ‰‘ƒŽǦ†‹”‡…–‡† ˆŽ—‹† –Š‡”ƒ’› ‹ …‘Ž‘”‡…–ƒŽ •—”‰‡”›Ǥ ”  ƒ‡•–ŠǤ ʹͲͳͶǢͳͳʹǣʹͺͳǦʹͺͻǤ ͵͵Ǥ‡‹‰ ǡ Žƒ–”‘—• ǡ ƒ„‡”  ‡– ƒŽǤ ƒ‡‘†›ƒ‹… ”‡•’‘•‡ –‘ …”›•–ƒŽŽ‘‹†•‘”…‘ŽŽ‘‹†•‹•Š‘…ǣƒ‡š’Ž‘”ƒ–‘”›•—„‰”‘—’ƒƒŽ›•‹• ‘ˆƒ”ƒ†‘‹•‡†…‘–”‘ŽŽ‡†–”‹ƒŽǤ ’‡ǤʹͲͳ͹Ǣ͹ǣ‡Ͳͳ͸͹͵͸Ǥ ͵ͶǤž•œŽ× ǡ‡‡–‡” ǡY˜‡‰‡•‡–ƒŽǤ‘Ž—‡Ǧ”‡’Žƒ…‡‡–”ƒ–‹‘ˆ‘” …”›•–ƒŽŽ‘‹†• ƒ† …‘ŽŽ‘‹†• †—”‹‰ „Ž‡‡†‹‰ ƒ† ”‡•—•…‹–ƒ–‹‘ǣ ƒ ƒ‹ƒŽ‡š’‡”‹‡–Ǥ –‡•‹˜‡ƒ”‡‡†š’ǤʹͲͳ͹ǢͷǣͷʹǤ ͵ͷǤ—””› ǡ†ƒ•‘Ǥ‘‹…”‡‰—Žƒ–‹‘‘ˆ˜ƒ•…—Žƒ”’‡”‡ƒ„‹Ž‹–›Ǥ …–ƒŠ›•‹‘ŽȋšˆȌǤʹͲͳ͵ǢʹͲ͹ǣ͸ʹͺǦ͸ͶͻǤ ͵͸Ǥ‡”‡”ƒ ǡ ƒ‹ŽŠ‘– ǡ ‹Ž‡› ǡ ƒ†ƒ˜‹ƒ Ǥ Š‡  ‡šƒǣ ƒ’‹† Ž–”ƒ•‘—†‹‘…‹–Š‡‡˜ƒŽ—ƒ–‹‘‘ˆ–Š‡…”‹–‹…ƒŽŽ›ŽŽŽǤ‡”‰‡† Ž‹‘”–ŠǤʹͲͳͲǢʹͺǣʹͻǦͷ͸ǡ˜‹‹Ǥ ͵͹Ǥ‡ŽŽ‹‰ Ǥ ‹”‡…– ‡ˆˆ‡…–• ‘ˆ ˜ƒ•‘ƒ…–‹˜‡ •—„•–ƒ…‡• ‘ •—’‡”ˆ‹…‹ƒŽ Š—ƒ˜‡‹•‹˜‹˜‘Ǥ –‰‹‘ŽǤͳͻͺͷǢͶǣʹ͵ͷǦʹͶʹǤ ͵ͺǤŠƒ˜‡ ǡ‡‹Ž•‘ Ǥ‘†›ˆŽ—‹††›ƒ‹…•ǣ„ƒ…–‘–Š‡ˆ—–—”‡Ǥ  ‘…‡’Š”‘ŽǤʹͲͳͳǢʹʹǣʹͳ͸͸ǦʹͳͺͳǤ