- Fluid, Electrolyte, and Acid-Base Disorders in Small Animal Practice

SAVE OFFLINE

Acknowledgments

I am grateful to many people for help in completing the fourth edition of this book. The concept of a book on disturbances of fluid, electrolyte, and acid-base balance for veterinarians began a long time ago in discussions with Dr. Dennis Chew, and this book represents the evolution of material taught to second-year veterinary students at the College of Veterinary Medicine and the approach to fluid therapy used in small animal patients at The Ohio State University Veterinary Teaching Hospital. I am indebted to Dr. Helio de Morais for encouraging me to undertake revisions of the book, and for his support throughout the process. I also thank Dr. Shane Bateman for helping me to more fully develop the critical care specialist’s perspective in the most recent two editions. Warm thanks and sincere appreciation go to all contributors who have shared their expertise in specific areas and provided comprehensive chapters on clinically relevant topics. I especially thank Dr. Felipe Galvao for stepping in to help complete the chapters on fluid, electrolyte, and acid base disturbances in diseases of the liver, gastrointestinal tract, and pancreas. Thanks again are due to Tim Vojt of our Biomedical Media Department for his original artwork. Thanks also go to editors and staff members at Elsevier, including acquisitions editor Heidi Pohlman and developmental editor Maureen Slaten. Special thanks go to associate developmental editor Brandi Graham for her polite persistence in seeing the project to completion. As always, I again thank my family for their love and support.

xi

Contributors Sarah K. Abood, DVM, PhD Department of Small Animal Clinical Sciences College of Veterinary Medicine Michigan State University East Lansing, Michigan Enteral Nutrition Shane Bateman, DVM, DVSc, DACVECC Director Hill’s Pet Nutrition Primary Healthcare Center Ontario Veterinary College University of Guelph Guelph, ON, Canada Disorders of Magnesium: Magnesium Deficit and Excess Introduction to Fluid Therapy Shock Syndromes Alexander W. Biondo, DVM, MSC, PhD Adjunct Associate Professor Department of Comparative Pathobiology School of Veterinary Medicine Purdue University West Lafayette, Indiana Departmento de Medicina Veternaria Universidade Federal do Parana Curitiba, Parana, Brazil Disorders of Chloride: Hyperchloremia and Hypochloremia Nichole Birnbaum, DVM, DACVIM (SAIM) Veterinary Internal Medicine Practice of Northern Virginia Bethesda, Maryland Fluid and Electrolyte Disturbances in Gastrointestinal and Pancreatic Disease Amanda Boag, MA, VetMB, DACVIM, DACVECC, FHEA, MRCVS Clinical Director Vets Now Dunfermline, Fife, United Kingdom Fluid Therapy with Macromolecular Plasma Volume Expanders

John D. Bonagura, DVM, MS, DACVIM (Cardiology, Internal Medicine) Professor Department of Veterinary Clinical Sciences College of Veterinary Medicine The Ohio State University Columbus, Ohio Cardiology Service Head Ohio State University Veterinary Hospital Member Davis Heart & Lung Research Institute Fluid and Diuretic Therapy in Heart Failure C. A. Tony Buffington, DVM, MS, PhD, DACVN Professor Department of Veterinary Clinical Sciences College of Veterinary Medicine The Ohio State University Columbus, Ohio Enteral Nutrition Sharon A. Center, DVM, DACVIM (SAIM) Professor of Medicine Department of Clinical Sciences College of Veterinary Medicine Cornell University Ithaca, New York Fluid, Electrolyte, and Acid Base Disturbances in Liver Disease Daniel L. Chan, DVM, DACVECC, DACVN, MRCVS Lecturer in Emergency and Critical Care Clinical Nutritionist Section of Emergency and Critical Care Department of Veterinary Clinical Sciences The Royal Veterinary College University of London North Mymms, Hertfordshire, United Kingdom Total Parenteral Nutrition

v

vi

CONTRIBUTORS

Dennis J. Chew, DVM, DACVIM (SAIM) Professor Department of Veterinary Clinical Sciences College of Veterinary Medicine The Ohio State University Columbus, Ohio Disorders of Calcium: Hypercalcemia and Hypocalcemia

Lisa M. Freeman, DVM, PhD, DACVN Professor Department of Clinical Sciences Cummings School of Veterinary Medicine Tufts University, North Grafton, Massachusetts Total Parenteral Nutrition

Peter D. Constable, BVSc (Hon), MS, PhD, DACVIM (LAIM) Professor and Head Department of Veterinary Clinical Sciences School of Veterinary Medicine Purdue University West Lafayette, Indiana Strong Ion Approach to Acid Base Disorders

Bernie Hansen, DVM, MS, DACVIM (SAIM), DACVECC Associate Professor Department of Clinical Sciences College of Veterinary Medicine North Carolina State University Raleigh, North Carolina Technical Aspects of Fluid Therapy

Larry D. Cowgill, DVM, PhD, DACVIM (SAIM) Professor Department of Medicine and Epidemiology School of Veterinary Medicine University of California Davis, California Director University of California Veterinary Medical Center San Diego, California Hemodialysis and Extracorporeal Blood Purification

Ann E. Hohenhaus, DVM, DACVIM (Oncology and SAIM) Staff Oncologist Head Jaqua Transfusion Medicine Service Donaldson-Atwood Cancer Clinic The Animal Medical Center New York, New York Blood Transfusions and Blood Substitutes

Joao Felipe de Brito Galvao, MV Resident, Small Animal Internal Medicine Department of Veterinary Clinical Sciences College of Veterinary Medicine The Ohio State University Columbus, Ohio Fluid and Electrolyte Disturbances in Gastrointestinal and Pancreatic Disease Fluid, Electrolyte, and Acid Base Disturbances in Liver Disease Helio Autran de Morais, DVM, PhD, DACVIM (SAIM & Cardiology) Associate Professor Department of Clinical Sciences College of Veterinary Medicine Oregon State University Corvallis, Oregon Disorders of Chloride: Hyperchloremia and Hypochloremia Respiratory Acid Base Disorders Mixed Acid Base Disorders Strong Ion Approach to Acid Base Disorders Fluid and Diuretic Therapy in Heart Failure Thierry Francey, Dr. med. vet, DACVIM (SAIM) Vetsuisse Faculty Small Animal Internal Medicine Department of Clinical Veterinary Medicine University of Berne, Switzerland Hemodialysis and Extracorporeal Blood Purification

Melissa L. Holahan, DVM Emergency and Critical Care Service VCA Shoreline Veterinary Referral & Emergency Center Shelton, Connecticut Enteral Nutrition Kate Hopper, BVSc, PhD, DACVECC Assistant Professor of Small Animal Emergency and Critical Care Department of Veterinary Surgical and Radiological Sciences University of California Davis, California Shock Syndromes Dez Hughes, BVSc, MRCVS, DACVECC Adjunct Senior Lecturer Institute of Veterinary, Animal and Biomedical Sciences Massey University Palmerston North New Zealand Fluid Therapy with Macromolecular Plasma Volume Expanders Rebecca A. Johnson, DVM, PhD, DACVA Clinical Assistant Professor Department of Surgical Sciences School of Veterinary Medicine University of Wisconsin Madison, Wisconsin Respiratory Acid Base Disorders

CONTRIBUTORS

vii

Catherine W. Kohn, VMD, DACVIM (LAIM) Professor Department of Veterinary Clinical Sciences College of Veterinary Medicine The Ohio State University Columbus, Ohio Applied Physiology of Body Fluids in Dogs and Cats

Larry A. Nagode, DVM, MS, PhD Emeritus Associate Professor Veterinary Biosciences College of Veterinary Medicine The Ohio State University Columbus, Ohio Disorders of Calcium: Hypercalcemia and Hypocalcemia

Mary Anna Labato, DVM, DACVIM (SAIM) Clinical Professor Department of Clinical Sciences Cummings School of Veterinary Medicine Tufts University North Grafton, Massachusetts Staff Veterinarian Tufts Foster Hospital for Small Animals North Grafton, Massachusetts Peritoneal Dialysis

David L. Panciera, DVM, MS, DACVIM (SAIM) Anne Hunter Professor of Small Animal Medicine Department of Small Animal Clinical Sciences Virginia-Maryland Regional College of Veterinary Medicine Virginia Tech Blacksburg, Virginia Fluid Therapy in Endocrine and Metabolic Disorders

Cathy Langston, DVM, DACVIM Department of Nephrology, Urology, and Hemodialysis Animal Medical Center New York, New York Fluid Therapy During Intrinsic Renal Failure Linda B. Lehmkuhl, DVM, MS, DACVIM (Cardiology) Staff Cardiologist MedVet Medical and Cancer Center for Pets Worthington, Ohio Fluid and Diuretic Therapy in Heart Failure Andrew L. Leisewitz, BVSc (Hons), MMedVet(Med), PhD Professor Companion Animal Clinical Studies Faculty of Veterinary Science University of Pretoria Pretoria, South Africa Mixed Acid Base Disorders Karol A. Mathews, DVM, DVSc, DACVECC University Professor Emeritus Department of Clinical Studies Ontario Veterinary College University of Guelph Guelph, Ontario, Canada Monitoring Fluid Therapy and Complications of Fluid Therapy Mary A. McLoughlin, DVM, MS, DACVS Associate Professor Department of Veterinary Clinical Sciences College of Veterinary Medicine The Ohio State University Columbus, Ohio Enteral Nutrition

Peter J. Pascoe, BVSc, DACVA, DVA, DECVAA Professor of Anesthesiology Department of Surgical and Radiological Sciences School of Veterinary Medicine University of California Davis, California Perioperative Management of Fluid Therapy Thomas J. Rosol, DVM, PhD, DACVP Professor Department of Veterinary Biosciences Special Assistant to the Vice President for Research for Technology and Commercialization The Ohio State University Columbus, Ohio Disorders of Calcium: Hypercalcemia and Hypocalcemia Linda A. Ross, DVM, MS, DACVIM (SAIM) Associate Professor Department of Clinical Sciences Tufts Cummings School of Veterinary Medicine North Grafton, Massachusetts Tufts Foster Hospital for Small Animals North Grafton, Massachusetts Peritoneal Dialysis Patricia A. Schenck, DVM, PhD Assistant Professor Diagnostic Center for Population and Animal Health, Endocrinology Section Michigan State University Lansing, Michigan Disorders of Calcium: Hypercalcemia and Hypocalcemia

viii

CONTRIBUTORS

Deborah Silverstein, DVM Assistant Professor of Critical Care Department of Veterinary Clinical Studies College of Veterinary Medicine University of Pennsylvania Philadelphia, Pennsylvania Shock Syndromes Kenneth W. Simpson, BVM&S, PhD, DACVIM (SAIM), DECVIM Professor Department of Veterinary Clinical Sciences College of Veterinary Medicine Cornell University Ithaca, New York Fluid and Electrolyte Disturbances in Gastrointestinal and Pancreatic Disease

Maxey L. Wellman, DVM, PhD, DACVP (Clinical Pathology) Professor of Pathology Department of Veterinary Biosciences College of Veterinary Medicine The Ohio State University Columbus, Ohio Applied Physiology of Body Fluids in Dogs and Cats

Michael D. Willard, DVM, MS, DACVIM Professor Department of Small Animal Clinical Sciences Texas A&M University College Station, Texas Disorders of Phosphorus: Hypophosphatemia and Hyperphosphatemia

3251 Riverport Lane St. Louis, Missouri 63043 FLUID, ELECTROLYTE, AND ACID-BASE DISORDERS IN SMALL ANIMAL PRACTICE

ISBN: 9781437706543

Copyright # 2012, 2006 by Saunders, an imprint of Elsevier Inc. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/ permissions. This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein). Notices Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods, professional practices, or medical treatment may become necessary. Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility. With respect to any drug or pharmaceutical products identified, readers are advised to check the most current information provided (i) on procedures featured or (ii) by the manufacturer of each product to be administered, to verify the recommended dose or formula, the method and duration of administration, and contraindications. It is the responsibility of practitioners, relying on their own experience and knowledge of their patients, to make diagnoses, to determine dosages and the best treatment for each individual patient, and to take all appropriate safety precautions. To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein.

Previous editions copyrighted 2000, 1992 ISBN: 978-1-4377-0654-3

Vice President and Publisher: Linda Duncan Acquisitions Editor: Heidi Pohlman Associate Developmental Editor: Brandi Graham Publishing Services Manager: Julie Eddie, Hemamalini Rajendrababu Project Manager: Marquita Parker, Deepthi Unni Designer: Margaret Reid

Printed in the United States of America Last digit is the print number: 9 8 7 6 5 4 3 2 1

To Maxey . . . John Donne was right about love – acquired serendipitously, but kept by commitment. After all these years, you’re still the one.

Preface The purpose of the fourth edition is the same as that of previous editions, namely “to bring together in one place information about fluid, electrolyte, and acid-base physiology and fluid therapy as they apply to small animal practice.” I still believe that a good foundation in physiology and pathophysiology is an essential part of veterinary education and enhances the clinician’s approach to the patient. Thoughtful evaluation of laboratory results provides valuable insight into the fluid, electrolyte, and acid-base status of the animal and can only improve the veterinary care provided. The in-depth approach of previous editions has been retained in the fourth edition. The book is divided into five sections: applied physiology, electrolyte disorders, acid-base disorders, fluid therapy, and special therapy. The first sections of the book on fluid, electrolyte, and acid-base physiology and disorders have been changed and updated. Chapter 3 (Disorders ofSodium and Water)has been revised to include information on vasopressin receptor antagonists (vaptans), a new class of drugs that likely will revolutionize management of patients with chronic hyponatremia, especially those with cardiac and liver disease. With the fourth edition, I welcome the contributions of Dr. Joao Felipe de Brito Galvao in the chapter on gastrointestinal and

x

pancreatic disease (Chapter 18) and the chapter on liver disease (Chapter 19). The book now has a strong Brazilian connection with contributions from Felipe Galvao and Helio De Morais! I also welcome Dr. Cathy Langston of the Animal Medical Center to edition four with her new chapter on fluid and electrolyte disorders in renal failure (Chapter 22). Dr. Shane Bateman has expanded Chapter 23 on shock to include two additional critical care specialists, Drs. Kate Hopper and Deborah Silverstein. Also new to the fourth edition is the collaboration of Dr. Amanda Boag with Dr. Dez Hughes in the chapter on macromolecular plasma volume expanders (Chapter 27). As in previous editions, I encourage those who read and use this book to write or e-mail me ([email protected]) about errors, controversial issues, and suggestions for improvement. The use of textbooks should be supplemented by reading the veterinary and human medical literature. Such an approach allows clinicians to maintain both the historical perspective and a contemporary view of medicine. In keeping with this belief, the use of extensive references has been retained in this edition. Considerable effort has been expended to ensure the accuracy of information provided here, but drug dosages always should be verified.

Stephen P. DiBartola, DVM, DACVIM Associate Dean for Administration and Curriculum Professor of Medicine Department of Veterinary Clinical Sciences College of Veterinary Medicine The Ohio State University Columbus, Ohio

CHAPTER • 1

Applied Physiology of Body Fluids in Dogs and Cats Maxey L. Wellman, Stephen P. DiBartola, and Catherine W. Kohn

Appropriate treatment of fluid and electrolyte abnormalities requires a basic understanding of the physiology of fluid balance. The purpose of this chapter is to provide an overview of the principles of body fluid homeostasis, beginning with a brief review of body fluid compartments. This is followed by a discussion of measurement of solutes in body fluids and the concepts of anion gap, osmolal gap, and zero balance.

DISTRIBUTION OF BODY FLUIDS In health, approximately 60% of an adult animal’s body weight is water. Estimates of total body water in adult dogs that are neither very thin nor obese are 534 to 660 mL/kg.26,59 Total body water of adult cats also was determined to be approximately 60%.56 There are some species and individual variations in total body water, likely related to age, sex, and body composition. In humans, total body water decreases with age and is lower in women than in men.13 Neonatal dogs and cats have higher total body water content (80% of body weight) than adults (60% of body weight),30 and an age-related decrease in total body water has been described in puppies and kittens during the first 6 months of life.35 Total body water was approximately 70% of total body weight in racing Greyhounds, likely due to low body fat content.21 Because fat has a lower water content than lean tissue, fluid needs should be estimated on the basis of lean body mass to avoid overhydration, especially in patients with cardiac or renal insufficiency or in those with hypoproteinemia. Formulas for estimating lean body mass are based on the assumptions that (1) in normal small animal patients, approximately 20% of body weight is due to fat, (2) morbid obesity increases body fat to approximately 30% of body weight, and (3) body weight is a reasonable estimate of lean body mass in thin patients:

2

Normal body weight
0:8 ¼ Lean body mass Obese body weight
0:7 ¼ Lean body mass Thin body weight
1:0 ¼ Lean body mass Water is the major component of all body fluids, which are distributed into several physically distinct compartments. Body fluids in each compartment equilibrate with fluids in other compartments by multiple mechanisms across a wide variety of membranes to maintain homeostasis. The volume of fluid in each of these compartments has been estimated using various isotope or dye dilution techniques and calculating their volumes of distribution. Results are expressed either as a percentage of body weight, which is easy to measure when calculating fluid therapy needs, or as a percentage of total body water, which is a useful conceptualization of body fluid compartments. Studies of body fluid compartments often are performed in experimental animals that have been anesthetized, splenectomized, or nephrectomized. Data from these kinds of studies vary with the protocol used and thus provide only approximations of fluid compartment sizes in healthy awake animals. The second edition of this book contains a more detailed discussion of the techniques involved in determination of total body water and the amount of fluid in the various compartments. As shown in Figure 1-1, the largest volume of fluid in the body is inside cells. The intracellular fluid (ICF) compartment comprises approximately 40% of body weight (approximately two thirds of total body water). The composition of ICF is very different from extracellular fluid (ECF) (Fig. 1-2). Intracellular homeostasis is maintained by shifts in water, solutes, and numerous other substances across the cell membrane. Any fluid not contained inside a cell is in the extracellular fluid compartment (approximately one third of total body water). Fluid shifts that occur during changes in hydration can have a marked effect on the ECF, and in most disease states, loss of fluids occurs initially from the ECF. For example, in diarrhea, a large volume of gastrointestinal

Applied Physiology of Body Fluids in Dogs and Cats

3

Dry matter 40% Intracellular fluid (ICF) 40% 4L

Interstitial fluid 15% 1500 ml

10 kg

Total body water = 60% of body weight

Total body water 60% 6L

Extracellular fluid (ECF) 20% 2L

Plasma 5% 500 ml

Dry matter 40% Intracellular fluid (ICF) 40% 2L

Interstitial fluid 15% 750 ml 5 kg Total body water 60% 3L

Extracellular fluid (ECF) 20% 1L

Plasma 5% 250 ml

Figure 1-1 Compartments of total body water expressed as percentage of body weight and total body water for a 10-kg dog and a 5-kg cat.

Extracellular fluid Na+

145

12

K+

4

140

Ca2+

2.5

4

Mg2+

1

34

–

110

4

–

24

12

H2PO4

–

2

40

Protein–

15 *

50

Cl HCO3 2–,

HPO4

Intracellular fluid

mEq/L *0 in interstitial fluid, 15 in plasma

Figure 1-2 Average values for electrolyte concentrations in extracellular and intracellular fluid. Note the marked concentration differences for many electrolytes.

fluid is lost; in renal failure, a large volume of ECF may be excreted. Fluid losses often are treated using parenteral fluids, which initially enter the ECF. Therefore, it is important to be able to estimate the volume of the ECF compartment and the volume of fluid lost to initiate appropriate fluid replacement and monitor fluid therapy.

Unfortunately, data from dye dilution studies of ECF volume are difficult to interpret because no indicator is truly confined to the ECF space. Estimates of ECF vary dramatically with the indicator used. ECF volumes reported for adult, healthy dogs and cats vary between 15% and 30% of body weight. The wide range in estimates of ECF volume likely results from the variety of techniques used to measure this space and the heterogeneity of ECFs, which include interstitial fluid (ISF), plasma, and transcellular fluids. Dense connective tissue, cartilage, and bone also contain a small amount of ECF. From a physiologic perspective and based on multiple studies using various indicators, the most accurate estimate of the ECF in adult small animals is 27% of lean body weight. However, an easier distribution of body fluids to remember is the 60:40:20 rule: 60% of body weight is water, 40% of body weight is ICF, and 20% of body weight is ECF (see Fig. 1-1). Many clinicians use 20% as an estimate for ECF when calculating fluid therapy needs for their patients. As mentioned above and as shown in Figure 1-1, ECF is distributed among several different subcompartments. Most ECF (about three fourths) is in spaces surrounding cells and is called interstitial fluid. Although accurate studies of the size of the ISF compartment in dogs and cats have not been reported, estimates derived from measurement of fluids in other compartments indicate that the ISF comprises approximately 15% of body weight

4

APPLIED PHYSIOLOGY

(approximately 24% of total body water). About one fourth of the ECF is within blood vessels and is called the intravascular compartment (plasma). Intravascular fluids are approximately 5% of body weight (approximately 8% to 10% of total body water). Most of the intravascular fluid is plasma. Plasma volume estimates range from 42 to 58 mL/kg in adult dogs that are neither very thin nor obese.26 Estimates for plasma volume in cats are 37 to 49 mL/kg.26 Blood volume, which includes erythrocytes, is a function of lean body mass, and estimated blood volume in dogs is 77 to 78 mL/kg (8% to 9% of body weight) and in cats is 62 to 66 mL/kg (6% to 7% of body weight).24 Racing Greyhounds may have higher blood volumes (110 to 114 mL/kg) than other breeds, possibly related to higher lean body mass.21 Fluids produced by specialized cells to form cerebrospinal fluid, gastrointestinal fluid, bile, glandular secretions, respiratory secretions, and synovial fluid are in the transcellular fluid compartment, which is estimated as approximately 1% of body weight (approximately 2% of total body water). Dense connective tissues, bone, and cartilage contain approximately 15% of total body water. However, these tissues exchange fluids slowly with other compartments. Because this fluid usually is not taken into account for routine fluid therapy, this compartment is not shown in Figure 1-1. Thus, a more simplified distribution of total body water often used for fluid therapy is: ICF is approximately 2/3 of total body water ECF is approximately 1/3 of total body water ISF is approximately 3/4 of ECF Intravascular fluid is approximately ¼ of ECF Although body fluids traditionally are conceptualized anatomically within these various compartments, water and solutes in these spaces are in dynamic equilibrium across the cell membrane, capillary endothelium, and specialized lining cells. Fluids and electrolytes shift among compartments to maintain homeostasis within each compartment. In health, the concentration of a particular substance may be similar or very different among the various fluid compartments. During disease, fluid volumes and solute concentrations may change dramatically. Loss or gain of fluid or electrolytes from one compartment likely will alter the volume and solute concentrations of other compartments.

DISTRIBUTION OF BODY SOLUTES In addition to water, body fluids contain various concentrations of solutes. Total body content of solutes may be measured by cadaver analysis (desiccation) or by isotope dilution studies. Every solute has a space or apparent volume of distribution. Dilution studies of body solute content yield variable results, depending on the volume of distribution of the particular tracer used to

estimate the solute space. There are limited data in the literature from cadaver and isotope dilution studies of body solute content in small animals, and most of the following discussion is based on data from studies in humans.13,48 Solutes are not distributed homogeneously throughout body fluids. Vascular endothelium and cell membranes have different permeabilities for various solutes. Healthy vascular endothelium is relatively impermeable to the cellular components of blood and to plasma proteins. Consequently, the volume of distribution of cells and proteins is the plasma space itself. However, the vascular endothelium is freely permeable to ionic solutes, and the concentration of these ions is almost the same in ISF as in plasma. Cell membranes maintain intracellular solutes at very different concentrations from those of the ECF. The compositions of solutes in the ECF and ICF are compared in Figure 1-2, and concentrations of solutes in plasma and in ISF and ICF are listed in Table 1-1. The slightly increased concentration of cations and anions in ISF compared with plasma water occurs primarily because of the presence of negatively charged proteins in plasma. The equilibrium concentrations of permeable anions and cations across the vascular endothelium are determined by the Gibbs-Donnan equilibrium, which occurs because negatively charged, nondiffusible proteins affect the distribution of other small charged solutes. In clinical practice, the difference in concentrations of anions and cations across the vascular endothelium is negligible, and the effects of the Gibbs-Donnan equilibrium are usually ignored. Thus, in clinical practice, plasma concentrations of solutes are considered to reflect solute concentrations throughout the ECF. Average values for plasma concentrations of important solutes in dogs and cats are given in Table 1-2. Table 1-1 shows that, although the solute compositions of ECF and ICF are quite different, the total numbers of cations and anions in all body fluids are equal to maintain electroneutrality. The most abundant cation in the ECF is sodium (Naþ). Most of the body Naþ is in the extracellular space. Approximately 70% of body Naþ in humans is exchangeable, and 30% is fixed as insoluble salts in bone.48 The percentage of exchangeable sodium is important because only exchangeable solutes are osmotically active. Cell membranes are permeable to Naþ, which tends to diffuse into cells. In health, however, cell membrane sodium, potassium-adenosinetriphosphatase (Naþ, Kþ-ATPase) actively removes Naþ from cells, thus maintaining a steep extracellular-to-intracellular concentration gradient for Naþ. The ECF also contains a small but physiologically important concentration of Kþ. For example, alterations in ECF Kþ concentrations may result in muscle weakness (hypokalemia) or cardiotoxicity (hyperkalemia). The most abundant anions in ECF are chloride (Cl) and bicarbonate (HCO3). The volume of distribution of Cl is primarily

Applied Physiology of Body Fluids in Dogs and Cats

TABLE 1-1 Ion

Approximate Ionic Composition of the Body Water Compartments

Plasma (mEq/L)

Plasma Water* (mEq/L)

Interstitial Fluid{ (mEq/L)

Intracellular Fluid—Skeletal Muscle Cell (mEq/L)

142 4.3 2.5

152.7 4.6 2.7

145.1 4.4 2.4

12.0 140 4.0

1.1

1.2

1.1

149.9

161.2

153

104 24 2

111.9 25.8 2.2

117.4 27.1 2.3

4 12 40

14 5.9 149.9

15 6.3 161.2

0 6.2 153

50 84{ 190

Cations Naþ Kþ Ca2þ (ionized) Mg2þ (ionized) Total Anions Cl HCO3 HPO42, H2PO4 Proteins Other Total

5

34 190

Adapted from Woodbury DM. In: Ruch TC, Patton HD, editors. Physiology and biophysics, 20th ed. Philadelphia: WB Saunders, 1974; Rose BD. Clinical physiology of acid-base and electrolytes, 3rd ed. New York: McGraw-Hill, 1989, with permission of the McGraw-Hill Companies. *Plasma water content is assumed to be 93% of plasma volume. { Gibbs-Donnan factors used as multipliers are 0.95 for univalent cations, 0.90 for divalent cations, 1.05 for univalent anions, and 1.10 for divalent anions. { This largely represents organic phosphates, such as ATP.

TABLE 1-2

Substance Sodium Potassium Ionized calcium Total calcium Total magnesium Chloride Bicarbonate Phosphate Proteins Lactate

Average Plasma Concentrations of Electrolytes in Dogs and Cats Units mEq/L mEq/L mg/dL mg/dL mg/dL mEq/L mEq/L mg/dL g/dL mg/dL

Dog

Cat

145 4 5.4 10 3 110 21 4 7 15

155 4 5.1 9 2.5 120 20 4 7 15

the ECF volume. Bicarbonate is present in all body fluids and can be generated from CO2 and H2O in the presence of carbonic anhydrase. In contrast to ECF, the primary cations in ICF are Kþ and magnesium (Mg2þ). Most of the body Kþ is in the ICF, where Kþ is the most abundant cation. Cell membranes are permeable to Kþ. The Kþ concentration gradient between ICF and ECF is maintained by cell membrane Naþ, Kþ-ATPase, which moves Kþ into cells against a concentration gradient. The ratio of intracellular to extracellular Kþ concentration is important in

generating and maintaining the cell membrane potential at approximately 70 mV (see Appendix). Almost 100% of body Kþ in humans is exchangeable.48 Unfortunately, a reliable, practical method for measuring the intracellular Kþ concentration is not available, and changes in serum Kþ concentration may not reflect changes in total body Kþ stores (see Chapter 5). The predominant anions in the ICF are organic phosphates and proteins. ICFs are not homogeneous. Concentrations of solutes vary in different cell types and in different subcellular compartments. From a clinical perspective, these differences usually are ignored. The heterogeneity of solute distribution between ICF and ECF may, however, play an important role in some disease processes. Transcellular fluids include cerebrospinal fluid, gastrointestinal fluid, bile, glandular secretions, and joint fluid. Transcellular fluids usually are not simply transudates of plasma. Transcellular fluid composition varies according to the cells that form the fluid. Concentrations of solutes in transcellular fluids will be discussed in later chapters, related to alterations in fluid balance involving specific transcellular fluids, such as loss of enteric fluids in diarrhea.

UNITS OF MEASURE Definitions can be tedious, but familiarity with a few may help with understanding the subsequent sections in this chapter. The definitions are presented in sequence of discussion, not alphabetically.

APPLIED PHYSIOLOGY

6

ATOMIC MASS (ALSO REFERRED TO AS RELATIVE ATOMIC MASS OR ATOMIC WEIGHT) Most naturally occurring elements consist of one or more isotopes of that element, each of which has a different mass. For example, carbon in the environment consists of approximately 99% 12C and 1% 13C. The atomic mass of an element is an average mass based on the distribution of stable isotopes for that element, and is determined by the weight of that element relative to the weight of the 12 C isotope of carbon, which is defined as 12.000. Atomic mass usually is reported with no units or as atomic mass units. The atomic mass is shown in most periodic tables of the elements. The atomic weights of some biologically important elements in body fluids are listed in Table 1-3.

MOLECULAR MASS (MOLECULAR WEIGHT) Many elements combine to form physiologically important compounds. The molecular mass of a compound is the sum of the atomic masses of the atoms that form the compound. For example, the molecular mass of water (H2O) is 18 and represents two times the atomic mass of hydrogen (2
1) plus the atomic mass of oxygen (16).

TABLE 1-3 Substance Calcium ion Carbon Chloride ion Hydrogen ion Magnesium ion Nitrogen Oxygen Phosphorus Potassium ion Sodium ion Sulfur Ammonia Ammonium ion Bicarbonate ion Carbon dioxide Glucose Lactate ion Phosphate ion

Sulfate ion Urea Water

The molecular weights of important compounds in body fluids are shown in Table 1-3.

FORMULA WEIGHT Ionic compounds do not really form molecules, and a more appropriate term for the mass of these substances is formula weight. For example, the formula weight of CaCl2 is the atomic mass of Ca2þ (40) plus two times the atomic mass of Cl (2
35.5) ¼ 111.

MOLE A mole is defined as 6.023
1023 particles. Some physiology texts define a mole as the molecular (or atomic) weight of a substance in grams, but a mole really just describes 6.023
1023 (Avogadro’s number) particles. It is defined as the number of atoms in exactly 12 g of 12 C. One mole of a substance weighs its molecular weight in grams (see section on Molecular Mass).

MOLAR MASS The molar mass is the mass in grams of 1 mol of a substance. By definition, 1 mol of carbon has a mass of 12 g. Molar masses are numerically equivalent to atomic or molecular weights but are reported in grams. For example, 1 mol Naþ weighs 23 g. Molar mass and gram molecular weight often are used interchangeably.

Atomic or Molecular Weights of Physiologically Important Substances Symbol or Formula

Atomic or Molecular Weight

Valence

Ca C Cl H Mg N O P K Na S NH3 NH4 HCO3 CO2 C6H12O6 C3H5O3 PO4 HPO4 H2PO4 SO4 NH2CONH2 H2O

40.1 12.0 35.5 1.0 24.3 14.0 16.0 31.0 39.1 23.0 32.1 17.0 18.0 61.0 44.0 180.0 89.0 95.0 96.0 97.0 96.1 60.0 18.0

þ2 0 1 þ1 þ2 0 0 0 þ1 þ1 0 0 þ1 1 0 0 1 3 2 1 2 0 0

Adapted from Rose BD. Clinical physiology of acid-base and electrolyte disorders, 3rd ed. New York: McGraw-Hill, 1989, with permission of the McGraw-Hill Companies.

Applied Physiology of Body Fluids in Dogs and Cats

MOLALITY AND MOLARITY Molality refers to the number of moles of solute per kilogram of solvent, whereas molarity refers to the number of moles of solute per liter of solution. The molarity and molality of most biologic solutions are approximately equal because the density of water is 1 kg/L. The slight difference between molarity and molality of a substance in plasma is because of nonaqueous proteins and lipids, which make up about 6% of the total volume. In body fluids, this difference is relatively unimportant, and the terms molality and molarity often are used interchangeably.

MILLIMOLE AND MILLIGRAM The prefix “milli” refers to 1 one-thousandth. A millimole is 1
103 mol; a milligram is 1
103 g. Many biologic substances in body fluids are measured in millimoles or milligrams.

CONCENTRATION Concentration refers to the amount of a substance that is present in a specified volume. The amount of a substance can be expressed as mass (grams or milligrams), moles (or millimoles), or equivalents (or milliequivalents). Volume usually is expressed as liters (L), deciliters (dL), or milliliters (mL). A deciliter is one tenth of a liter (i.e., 100 mL). Many solutions used for fluid therapy are percent solutions. Percent concentration refers to a number of parts in 100 parts of solution. This may be used to express concentration in terms of weight per unit weight, weight per unit volume, or volume per unit volume. For example, a 0.9% solution of NaCl contains 0.9 g of NaCl per 100 mL of solution, because 100 mL of H2O is equal to 100 g of H2O (0.9 g NaCl/100 g H2O). Because a gram is equal to 1000 mg and a deciliter is equal to 100 mL of solution, a 0.9% solution of NaCl contains 900 mg of NaCl per deciliter (9 g NaCl/L). Similarly, a 10% solution of CaCl2 contains 10 g of CaCl2 per 100 mL of solution, or 10 g of CaCl2 per deciliter (100 g/L), and 5% dextrose contains 5 g of dextrose per deciliter (50 g/L).

7

negative one (e.g., Cl); a divalent cation has a charge of positive two (Ca2þ). One atom of Ca2þ combines with two atoms of Cl to form CaCl2. It is useful to express concentrations of solutes in body fluids in equivalents per liter (Eq/L) or milliequivalents per liter (mEq/L) to reflect the charge or valence of the solute. The equivalent weight of a substance is the atomic, molecular, or formula weight of a substance divided by the valance.

ELECTROCHEMICAL EQUIVALENCE Rose49 defines electrochemical equivalence as follows: One equivalent is defined as the weight in grams of an element that combines with or replaces 1 g of hydrogen ion (Hþ). Because 1 g of Hþ is equal to 1 mol of Hþ (containing approximately 6.023
1023 particles), 1 mol of any univalent anion (charge equals 1) will combine with this Hþ and is equal to 1 equivalent (Eq). For example, 1 mol (1 equivalent) of Cl combines with 1 mol of Hþ; 1 mol (1 equivalent) of Naþ could replace 1 mol of Hþ; 1 mol (2 equivalents) of Ca2þ combines with 2 mol (2 equivalents) of Cl to form 1 mol of CaCl2. Therefore, it is useful to express concentrations of solutes in body fluids in equivalents per liter (Eq/L), thus reflecting the charge or valence of the solute.

EQUIVALENT WEIGHT The equivalent weight of a substance is the atomic, molecular, or formula weight divided by the valence. The milliequivalent (mEq) weight is 103 times the equivalent weight. For an element such as sodium, which has a valence of þ1, the milliequivalent weight is equal to its atomic weight. Therefore, each millimole of Naþ provides 1 mEq. In contrast, the milliequivalent weight of Ca2þ is one half its atomic weight because its valence is þ2. Each millimole of Ca2þ provides 2 mEq (0.5 mmol provides 1 mEq). These relationships may be summarized as: Millimolecular weight=valence ¼ milliequivalent weight Millimoles
valence ¼ milliequivalents ðmEqÞ

CATION A cation is an atom or molecule with a positive charge. A monovalent cation has one positive charge (e.g., Naþ), and a divalent cation has two positive charges (e.g., Ca2þ).

ANION An anion is an atom or molecule with a negative charge. A monovalent anion has one negative charge (e.g., Cl), and a divalent anion has two negative charges (e.g., SO42).

VALENCE Ions in body fluids combine according to ionic charge (valence) rather than weight. The number of cations (positively charged ions) in a solution always equals the number of anions (negatively charged ions) to maintain electroneutrality. A univalent anion has a charge of

To convert concentrations: mEq=L ¼ mmol=L
valence mg=dL
10 mEqL ¼
valence molecular weight Note: Multiplication by 10 in the numerator converts mg/dL to mg/L. Dividing by the molecular weight converts milligrams to millimoles. Multiplying by the valence converts to milliequivalents. Phosphate can exist in body fluids in three different ionic forms: H2PO4, HPO42, and PO43 (see Chapter 7). At the normal pH of ECF, approximately 80% of phosphate is in the HPO42 form and 20% is in the H2PO4 form. Therefore, the average valence of

8

APPLIED PHYSIOLOGY

phosphate in ECF is 0.8
(2) þ 0.2
(1) ¼ 1.8. At a normal plasma phosphate concentration of 4 mg/dL, the phosphate concentration expressed in mEq/L would be: 4
10
1:8 ¼ 2:3 mEq=L 31

OSMOLALITY AND OSMOLARITY Regardless of its weight, 1 mol of any substance contains the same number of particles (6.023
1023; Avogadro’s law). Solutes exert an osmotic effect in solution that is dependent only on the number of particles in solution, not their chemical formula, weight, size, or valence. One osmole (Osm) is defined as 1 g molecular weight of any nondissociable substance; therefore, each osmole also contains 6.023
1023 molecules. If a substance does not dissociate in solution (e.g., glucose), 1 mol equals 1 Osm. If a substance dissociates in solution, the number of osmoles equals the number of dissociated particles. For example, assuming that NaCl completely dissociates into Naþ and Cl in solution, each millimole of NaCl provides 2 milliosmoles (mOsm): 1 mOsm of Naþ and 1 mOsm of Cl. If a compound in solution dissociates into three particles, the number of osmoles in solution is increased three times (e.g., CaCl2). The milliosmolar concentration of a solution may be expressed as the solution’s milliosmolarity or milliosmolality. Osmolality refers to the number of osmoles per kilogram of solvent. An aqueous solution with an osmolality of 1.0 results when 1 Osm of a solute is added to 1 kg of water. The volume of the resulting solution exceeds 1 L by the relatively small volume of the solute. In clinical veterinary medicine, osmolality is expressed as milliosmoles per kilogram. Osmolarity refers to the number of osmoles per liter of solution. If 1 Osm of a solute is placed in a beaker and enough water is added to make the total volume 1 L, the osmolarity of the resulting solution is 1. In clinical medicine, osmolarity is expressed as milliosmoles per liter. In biologic fluids, there is a negligible difference between osmolality and osmolarity, and the term osmolality is used in this discussion In clinical medicine, osmolality is measured in serum, because the addition of anticoagulants for plasma samples would increase solute in the sample. Serum osmolality usually is measured by freezing-point depression, which is more precise and accurate than vapor pressure determinations. One osmole of a solute in 1 kg of water depresses the freezing point of the water by 1.86 C.55 Average values for measured serum osmolality in the dog and cat are 300 and 310 mOsm/kg, respectively.8,17 Measured osmolality may not be the same as calculated osmolality (see later discussion).

Effective and Ineffective Osmoles In any fluid compartment, the osmotic effect of a solute is in part dependent on the permeability of the solute across the membranes separating the compartment. Consider the two fluid compartments in a rigid box in Figure 1-3. Assume that the membrane dividing the two compartments is freely permeable to urea and water but is impermeable to glucose. When urea is added to the left compartment (top of figure), it moves down its concentration gradient from left to right, and water moves down its concentration gradient from right to left until there are equal concentrations of urea and water on both sides of the membrane. No fluid rises in the column attached to the left fluid compartment because urea is an ineffective osmole and does not generate osmotic pressure. In biologic fluids, urea is a small molecule that freely diffuses across most cell membranes and therefore does not contribute to effective osmolality. When glucose is added to the left compartment (bottom of figure), water moves down its concentration gradient from right to left, but glucose cannot move across the membrane. This movement of water from a solution of lesser solute concentration across a semipermeable

Urea H2O

= Urea

Ineffective osmole

Semipermeable membrane Osmotic pressure

H2O

= Glucose

Effective osmole

Figure 1-3 Effective and ineffective osmoles. Top, Effect of adding a permeable solute such as urea (small closed circles) to the fluid on one side of a membrane. In this setting, equilibrium is reached by urea equilibration across the membrane rather than water movement into the urea compartment. Consequently, no osmotic pressure is generated. Bottom, Effect of adding an impermeable solute such as glucose (large open circles) to the fluid on one side of a membrane. As water moves into the glucose compartment, hydraulic pressure is generated (measured by the height of the column of water above the glucose compartment), which at equilibrium equals the osmotic pressure of the solution.

Applied Physiology of Body Fluids in Dogs and Cats

9

membrane to a solution of greater solute concentration is called osmosis. The influx of water into the left compartment resulting from the osmotic effect of glucose causes the solution to rise in the column. The height of fluid in the column is proportional to the osmotic pressure generated by glucose. In this example, glucose is an effective osmole because it generates osmotic pressure by causing a shift of water across the boundary membrane. Glucose is an effective osmole in this setting because the boundary membrane is impermeable to glucose but permeable to water. In biologic fluids, glucose can contribute to osmolality because it is not freely diffusible.

tonicity of a solution may be less than the measured osmolality if both effective and ineffective osmoles are present. Thus, the tonicity and osmolality of a solution are not necessarily equal—a circumstance that often is true in biologic solutions.

Tonicity

The calculated osmolality is an estimate of serum osmolality using various formulas. The formulas include solutes that have a major contribution to total osmolality. Calculated osmolality often is less than measured osmolality because the formulas either exclude some osmotically active particles or estimate their contribution.

The effective osmolality of a solution is referred to as the tonicity of the solution. A freezing-point depression osmometer measures all osmotically active particles in the solution. Thus, the measured osmolality of a solution includes both effective and ineffective osmoles. The Example 1

Measured Osmolality The osmolality determined with an osmometer is the measured osmolality, which typically is not the same as the calculated osmolality estimated using various formulas.

Calculated Osmolality

Determine how many millimoles, milliequivalents, and milliosmoles of sodium and chloride there are in 1 L of a 0.9% solution of NaCl.

Concentration of 0.9% NaCl:

0.9 g NaCl/100 mL of solution ¼ 900 mg NaCl/dL

Convert milligrams to grams and deciliters to liters

900 mg NaCl/100 dL
1 g/1000 mg
10 dL/L ¼ 9 g NaCl/L

Formula weight of NaCl: (use atomic weight from Table 1-3 or periodic table)

Atomic mass of Na þ atomic mass of Cl ¼ 23 þ 35.5 ¼ 58.5

Molar mass of NaCl:

58.5 g

Convert grams to moles:

9 g NaCl
(1 mol/58.5g) ¼ 0.154 mol NaCl 0.154 mol
(1000 mmol/mol) ¼ 154 mmol NaCl

Convert moles to millimoles: þ

Determine millimoles of Na and Cl



NaCl in solution dissociates into Naþ and Cl, yielding 154 mmol/L of Naþ and 154 mmol/L of Cl

Determine milliequivalents of Naþ and Cl

millimoles
valence ¼ milliequivalents Naþ and Cl each have a valence of 1 154 mmol
1 ¼ 154 mEq of Naþ 154 mmol
1 ¼ 154 mEq of Cl

Determine milliosmoles of Naþ and Cl

NaCl in solution dissociates into Naþ and Cl, so the mOsm/L in 0.9% NaCl is the sum of the mOsm for each component: 154 mEq/L Naþ þ 154 mEq/L Cl 154 mOsm/L Naþ þ 154 mOsm/L Cl ¼ 308 mOsm/L

Example 2

Determine how many millimoles, milliequivalents, and milliosmoles of calcium and chloride there are in 1 L of a 10% solution of CaCl2.

Concentration of 10% CaCl2:

10 g CaCl2/100 mL of solution ¼ 10 g CaCl2/dL

To convert deciliters to liters:

10 g CaCl2/dL
10 dL/L ¼ 100 g CaCl2/L

Formula weight of CaCl2: (use atomic weight from Table 1-3 or periodic table)

Atomic mass of Ca þ 2
(atomic mass of Cl) ¼ 40.1 þ (2
35.5) ¼ 111.1 Continued

10

APPLIED PHYSIOLOGY

Molar mass of CaCl2:

111.1 g

Convert grams to moles:

100 g CaCl2
(1 mol/111.1 g) ¼ 0.9 mol CaCl2

Convert moles to millimoles

0.9 mol
(1000 mmol/mol) ¼ 900 mmol CaCl2

Determine millimoles of Caþ2 and Cl:

CaCl2 in solution dissociates into Caþ2 and 2Cl, yielding 900 mmol/L of Caþ2 and 1800 mmol/L of Cl

Determine milliequivalents of Caþ2 and Cl

millimoles
valence ¼ milliequivalents Caþ2 has a valence of 2; Cl has a valence of 1 900 mmol Caþ2
2 ¼ 1800 mEq of Caþ2 1800 mmol Cl
1 ¼ 1800 mEq of Cl

Determine milliosmoles of Caþ2 and Cl:

CaCl2 in solution dissociates into Caþ2 þ 2Cl mOsm/L in 10% CaCl2 is the sum of the milliosmoles for each component: 1800 mEq/L Caþ2 þ 1800 mEq/L Cl 900 mOsm/L Caþ2 þ 1800 mOsm/L Cl ¼ 2700 mOsm/L

Osmolal Gap Osmolal gap is the difference between measured osmolality and calculated osmolality.

Colloid Osmotic Pressure (Oncotic Pressure) Colloids are large molecular weight (MW ¼ 30,000) particles present in a solution. The component of the total osmotic pressure in plasma contributed by colloids is called the colloid osmotic pressure (oncotic pressure). Plasma proteins are the major colloids present in normal plasma. Although colloid osmotic pressure is only about 0.5% of the total osmotic pressure, oncotic pressure is extremely important in transcapillary fluid dynamics. Oncotic pressure can be measured using a colloid osmometer (oncometer). Several examples related to fluid therapy are included here to illustrate how these definitions may be used in clinical veterinary medicine.

EXCHANGE OF WATER BETWEEN EXTRACELLULAR AND INTRACELLULAR FLUID SPACES The number of osmotically active particles in each space determines the volume of fluid in the ECF and ICF compartments. The osmolality of physiologic fluids is dominated by small solutes that are present in high concentrations. In serum, sodium, potassium, chloride, bicarbonate, urea, and glucose are present in high enough concentrations to individually affect osmolality. Together these make up more than 95% of the total osmolality of

serum. Larger molecules like albumin contribute little to the osmolality. As mentioned above, osmotic activity depends on the solute and its permeability across the membrane. Sodium is the most abundant cation in the ECF. Although there is variation among different types of cells, many cell membranes are impermeable to sodium. Sodium movement across most cell membranes occurs by active transport. Consequently, Naþ and its associated anions account for most of the osmotically active particles in the ECF and as such are considered effective osmoles. Glucose and urea are two other substances with potential osmotic activity. Many cell membranes are not freely permeable to glucose, in which case glucose would be osmotically active. In contrast, urea does not make a major contribution to effective osmolality in the ECF because it is a small molecule that is freely diffusible across most cell membranes. However, urea may have an impact on serum osmolality if its concentration is increased. Osmolality of the ECF may be estimated using various formulas. This is called the calculated osmolality because it is based on estimating the contribution of osmotically active substances. Calculated osmolality by itself is not very useful because it is simply an estimate based on the concentration of commonly measured solutes. Calculated osmolality, which is an estimate, may not be the same as measured osmolality, which is determined by an osmometer. The formulas for calculated osmolality include various combinations of the most osmotically active solutes, but none includes all osmotically active solutes because not all solutes are measured on routine biochemical profiles. The formulas also assume complete dissociation of some solutes, which may not be true in serum.

Applied Physiology of Body Fluids in Dogs and Cats One of the most commonly used formulas for calculating the osmolality of serum is:48 ECF osmolalityðmOsm=kgÞ ¼ 2ð½Naþ  þ ½K þ Þ þ

Effective ECF osmolality ¼ 2
Naþ þ

½glucose 18

In healthy dogs and cats, the contribution of glucose to the effective osmolality of the ECF is small (about 4 to 6 mOsm/kg) based on blood glucose concentrations of 70 to 110 mg/dL. Therefore, 2
[Naþ] is a good approximation of the ECF effective osmolality. All body fluid spaces are isotonic with one another. Thus, the effective osmolality of the ICF also may be estimated by doubling the ECF Naþ concentration, [Naþ], even though the Naþ concentration in ICF is small. Because all body fluid spaces are isotonic, the tonicity of total body water also may be approximated by doubling the plasma [Naþ]. The tonicity of total body water also may be expressed as the ratio of the sum of all exchangeable cations and all exchangeable anions to the volume of total body water. Exchangeable ions (denoted by the subscript letter “e”) are able to move throughout the fluid compartment. The total number of milliosmoles of exchangeable cations and anions may be estimated from the expression: 2½Naþ e þ 2½K þ e

Therefore:

½glucose ½BUN þ 18 2:8

In all formulas, Naþ and Kþ are measured in millimoles per liter or milliequivalents per liter. In this formula, the contribution of Cl and HCO3 is estimated by multiplying the major cations by 2, assuming the serum must remain electrically neutral. The concentrations of glucose and blood urea nitrogen (BUN) are divided by 18 and 2.8, respectively, to convert milligram per deciliter to millimoles per liter (The molecular weight of glucose is 180 and the molecular weight of urea is 28, and there are 10 dL/L). Several other formulas have been suggested for estimation (calculated osmolality) of the true serum osmolality (measured osmolality). These formulas vary based on which major solutes are included and whether constants are added to estimate the effects of other solutes. Including Kþ is a more accurate estimate of measured osmolality. Remember, in all formulas, Naþ and Kþ are measured in millimoles per liter or milliequivalents per liter. If glucose and BUN are measured in milligrams per deciliter, the conversion factor is included in the formula. If glucose and BUN are measured in millimoles per liter, delete the conversion factor (see later discussion). Alternate formulas are listed in the second edition of this book. Not all potentially osmotic substances are osmotically active in body fluids. Cell membranes are permeable to urea and Kþ; therefore, these solutes are ineffective osmoles. Effective osmolality (tonicity) is calculated as:48

11

2
plasma½Naþ  

2½Na þ e þ 2½K þ e TBW

and plasma½Naþ  

2½Na þ e þ 2½K þ e TBW

This relationship is represented graphically in Figure 1-4.14,49 Examination of Figure 1-4 shows that when total exchangeable Naþ increases, serum sodium concentration also increases,49 and these changes are usually associated with body fluid hypertonicity. A decrease in total exchangeable Naþ or Kþ is associated with hyponatremia, a decrease in plasma osmolality, and hypotonicity. The effect of a decrease in total exchangeable Kþ on serum [Naþ] is not intuitively obvious but is clinically important.49 A decrease in serum [Kþ] results in a shift of Kþ out of cells. To maintain electroneutrality, Naþ shifts into cells, thus causing hyponatremia. Serum (and therefore ECF) osmolality in dogs is approximately 300 mOsm/kg, and fluids with effective osmolalities greater than 300 mOsm/kg are hypertonic to plasma, whereas those with effective osmolalities less than 300 mOsm/kg are hypotonic to plasma. Those with effective osmolalities of 300 mOsm/kg are isotonic to plasma. In health, addition or loss of fluid or solute to or from the body results in alterations in body fluid space volumes and tonicity. These alterations elicit homeostatic shifts of fluid between compartments so that fluid spaces return to isotonicity (see Chapter 3). In most disease states, fluid and solutes initially are lost from the ECF. Three basic types of fluid and solute loss may occur: solute in excess of water (loss of hypertonic fluids), isotonic loss (loss of isotonic fluids), or water in excess of solute (loss of hypotonic fluids) (Table 1-4).28 Solute and water losses theoretically may occur in any proportion along the continuum between solute loss with no water loss (e.g., peritoneal dialysis with a salt-poor solution) and water loss with no solute loss (e.g., water deprivation). When solute is lost in excess of water (hypertonic fluid loss), the osmolality of the ECF decreases relative to that of the ICF. This could be seen in oozing of serum from the skin of burn patients, which occurs much more commonly in human medicine than in veterinary medicine. Water passes from the ECF through the cell membrane to the ICF, thus diluting the ICF solute until the effective osmolalities of ECF and ICF are again equal. The osmolalities of both ICF and ECF decrease. This homeostatic fluid shift decreases ECF volume. When hypertonic fluid is lost from the ECF and volume depletion occurs, homeostatic water shifts further compromise the ECF volume and effective circulating blood volume, thus compounding fluid losses.

12

APPLIED PHYSIOLOGY

"Corrected" serum sodium concentration (mEq / L serum water) (Na's)

165 160

Na's = 1.11 (Nae + Ke)/ TBW –25.6 SY.X = 5.6 r = 0.83

155 150 145 140 135 130 125 120 115

No edema Edema

110 110 115 120 125 130 135 140 145 150 155 160 165 170 175 180 (Nae+Ke) / TBW (mEq/L)

Figure 1-4 Relationship of plasma [Naþ] to ([Naþ]e  [K]e)/TBW. [Naþ]e, total exchangeable Naþ; [Kþ]e, total exchangeable Kþ; TBW, total body water. (From Edelman IS, Leibman J, O'Meara MP, et al. Interrelations between serum sodium concentration, serum osmolarity and total exchangeable sodium, total exchangeable potassium, and total body water. J Clin Invest 1958;37:1236–1256.)

TABLE 1-4

Loss Hypotonic Isotonic Hypertonic

Effect of Water and Solute Losses from Body Fluids

ECF Hypertonic Isotonic Hypotonic

Theoretical Replacement Fluid Hypotonic Isotonic Isotonic/hypertonic

During water deprivation, the tonicity of ECF increases relative to that of the ICF. Water shifts out of cells and into ECF until the osmolalities of the two compartments are equal. The osmolalities of both ICF and ECF are greater than those during the state of normal hydration. This water shift augments the ECF volume, thus helping to preserve the effective circulating blood volume and protecting against the development of shock. Loss or gain of isotonic fluid from the ECF results in no change in ECF osmolality, and no osmotically mediated water shifts between the ICF and ECF occur. Loss of isotonic fluid results in a decrease in ECF volume, whereas gain

of isotonic fluid increases the ECF volume. Isotonic fluid loss, if of sufficient magnitude, results in hypovolemia and shock. These concepts are discussed further in Chapter 3.

EXCHANGE OF WATER BETWEEN PLASMA AND INTERSTITIAL SPACES Most of the ECF is in either the interstitial compartment (approximately three quarters of the ECF) or the intravascular compartment, most of which is plasma (approximately one quarter of the ECF). The partitioning of fluid between plasma and ISF spaces is critically important for maintenance of the effective circulating blood volume. The effective blood volume has been defined as “the component of blood volume to which the volume-regulatory system responds by causing renal sodium and water retention in the setting of cardiac and hepatic failure even though measured total blood and plasma volume may be increased.”42,51 Exchange of solutes and fluid between plasma and interstitial spaces occurs at the capillary level. The volume of the vascular space is controlled by a balance between forces that favor filtration of fluid through the vascular endothelium into the interstitial space (capillary

Applied Physiology of Body Fluids in Dogs and Cats

πp

Pcap

Arteriole

Anions

Cations

Vessel

13

Pif

πif

Venule Cl– 110 mEq/L

Interstitium

Figure 1-5 Factors affecting fluid movement at the level of the capillary. Pcap, Capillary hydrostatic pressure; Pif, interstitial hydrostatic pressure; pp, capillary oncotic pressure; pif, interstitial oncotic pressure.

Na+ 145 mEq/L

HCO3– 21 mEq/L

hydrostatic pressure and tissue oncotic pressure) and forces that tend to retain fluid within the vascular space (plasma oncotic pressure and tissue hydrostatic pressure). Oncotic pressure is the osmotic pressure generated by plasma proteins in the vascular space. Starling’s law describes these relationships (Fig. 1-5): Net filtration ¼ K f ½ðPcap  Pif Þ  ðpp pif Þ where Kf represents the net permeability of the capillary wall, P represents the hydrostatic pressure generated by the heart (Pcap) or tissues (Pif), and p represents the oncotic pressure generated by plasma proteins (pp) or filtered proteins and mucopolysaccharides in the interstitium (pif). The net filtration pressure in healthy capillaries is about 0.3 to 0.5 mm Hg at the proximal (arteriolar) end of the capillary.49 Near the venule, the forces favoring filtration are less than the forces favoring reabsorption of fluid into the vascular space, because capillary hydrostatic pressure decreases along the length of the capillary, but capillary oncotic pressure remains approximately the same.49 Some of the fluid that is filtered into the interstitium at the proximal end of the capillary is reabsorbed distally; the remainder of the filtered fluid is transported by lymphatics in the interstitium. The hydrostatic pressure transferred from arterioles to the capillaries is controlled by autoregulation of the precapillary sphincter. Autoregulation protects the capillary from increases in hydrostatic pressure caused by systemic hypertension, which otherwise could cause a dangerous loss of vascular fluid into the ISF by filtration. During water depletion, capillary oncotic pressure increases and hydrostatic pressure may decrease if depletion is severe enough to cause hypovolemia. These alterations in Starling’s forces favor a decrease in net filtration of fluid into the interstitium at the level of the capillary. Increased reabsorption of ISF augments effective circulating blood volume, thus decreasing plasma protein concentration and increasing hydrostatic pressure. Conversely, loss of plasma protein decreases plasma oncotic pressure and increases the net force favoring filtration of fluid out of the capillary. Loss of intravascular fluid increases plasma oncotic pressure, but filtration of fluid into the interstitium produces the edema observed in hypoproteinemic states. Thus, in the healthy animal, maintenance of plasma volume depends on a fine balance

HPO42–, H2PO4– 2 mEq/L SO42– 2 mEq/L UA Lactate– 2 mEq/L Other– 3 mEq/L Proteins–n 16 mEq/L

K+ 4 mEq/L UC

Ca2+ 5 mEq/L Mg2+ 2 mEq/L

UA - UC = Anion gap

Figure 1-6 Relative concentrations of unmeasured anions (UAs) and cations (UCs) in extracellular fluid (ECF).

between the forces favoring filtration and those favoring reabsorption in the capillary.

ELECTRONEUTRALITY AND THE ANION GAP In body fluids, the sum of all cations must equal the sum of all anions to fulfill the law of electroneutrality. In the clinical setting, however, all anions and cations in body fluids are not routinely measured. Figure 1-6 compares the concentrations of the commonly measured anions and the commonly measured cations in a gamblegram. The commonly measured cations are Naþ and Kþ, and the commonly measured anions are Cl and HCO3. The sum of the concentrations of commonly measured anions is less than the sum of the concentrations of commonly measured cations. In other words, there are more unmeasured anions (UAs) than unmeasured cations (UCs). From this observation, the concept of the anion gap was developed. It is important to remember that there is no real difference between the total number of anions and the total number of cations in the body. In the clinical setting, the anion gap is used to predict changes in the UAs or less commonly in the UCs. The anion gap is defined as the difference between the UAs and the UCs. According to the law of electroneutrality, Naþ þ K þ þ UC ¼ Cl þ HCO 3 þ UA Rearranging this equation, ðNaþ K þ Þ  ðCl þ HCO3 Þ ¼ UA  UC ¼ anion gap The range for the normal anion gap varies by species and is approximately 12 to 24 mEq/L in dogs and 13

APPLIED PHYSIOLOGY

14

to 27 mEq/L in cats (see Chapter 9). The average anion gap in dogs is 18 to 19 mEq/L11 and in cats is approximately 24 mEq/L.32 The higher average anion gap in cats suggests a higher net charge on proteins in this species. Younger animals may have a lower anion gap. Little information on variations in anion gap in pediatric small animal patients is available. In 3-day-old puppies, however, anion gap values were reported to be approximately 16 mEq/L in one study,37 suggesting that anion gaps in neonatal puppies are within the reference ranges for adults. The primary usefulness of the anion gap is to detect an increase in UAs as an aid in the diagnosis of metabolic acidosis. Clinically relevant changes in the anion gap usually are from changes in UAs, and most of these changes are caused by increases in UAs associated with organic acids. For example, the ketoacidosis that occurs in some diabetic patients causes an increase in UAs, resulting in an increase in the anion gap. Similarly, the increased UAs that occur with ethylene glycol intoxication result in an increased anion gap. The derivation and clinical application of the principle of the anion gap are discussed further in Chapters 9 and 10.

THE OSMOLAL GAP The osmolal gap is defined as the difference between the measured and the calculated serum osmolalities: Osmolal gap ¼ Osmm  Osmc Reference values for osmolal gaps in dogs are given in Table 1-5. Data for osmolal gaps in cats are not reported in the literature. Attempts to derive osmolal gaps from published data on measured serum osmolalities and electrolyte concentrations in cats have yielded confusing results (see footnote to Table 1-5). Values for the osmolal

TABLE 1-5

Reference Ranges for Osmolal Gap

Species

Osmolal Gap (mOsm/kg)

Reference

Dog Dog Dog Dog

10  6 10.1  5.9 0-10 56

Grauer15 Hauptman20 Shull53 Burkitt7

Serum osmolality values in normal cats were reported to be approximately 308  5 mOsm/kg (Chew et al8). When mean values for serum Na (155 mEq/L), K (4 mEq/L), glucose (120 mg/dL), and blood urea nitrogen (BUN; 24 mg/dL) are substituted into the equation 2(Na þ K) þ glucose/18 þ BUN/2.8, a value of 333 mOsm/kg is obtained for cats. Calculated plasma osmolality values greater than measured values have generally been attributed to laboratory error. Why calculated plasma osmolality exceeds measured plasma osmolality using mean values from normal cats is unclear.

gap vary with the formula used to calculate osmolality. Numerous formulas have been derived to calculate serum osmolality (see earlier section on exchange of water between ICF and ECF spaces). One of the most commonly used formulas to estimate osmolality is: 2ðNaþ þ K þ Þ þ

glucose BUN þ 18 2:8

Some laboratories report a calculated osmolality based on these various formulas because it is easy to program the analyzer to perform the calculation. These are estimates of the actual osmolality, which must be measured using an osmometer. Serum osmolality most frequently is measured by freezing-point depression. Measured osmolality is higher than calculated osmolality because Osmm measures all osmotically active solutes, whereas the formulas used for Osmc do not account for all osmotically active solutes in serum. The difference (gap) between the measured (actual) and calculated (estimated) osmolality is called the osmolal gap. Calculation of the osmolal gap is most helpful when unsuspected osmoles are present in ECF, thus increasing the osmolal gap as a result of an increase in the measured but not the calculated osmolality (e.g., ethylene glycol poisoning), and when assessing the significance of the serum Naþ concentration (see Chapter 3). During the acute stage (6 to 12 hours after exposure) of ethylene glycol toxicity, the osmolal gap is increased. This increased osmolal gap could be helpful in the diagnosis of ethylene glycol toxicity if a measured osmolality is requested. Hyponatremia with a normal osmolal gap suggests dilutional hyponatremia (e.g., overhydration). This rules out the presence of abnormal osmotically active particles that could cause a shift of water from ICF to ECF, thus decreasing the serum sodium concentration. The osmolal gap is discussed further in Chapter 3.

HOMEOSTASIS: ZERO BALANCE In the healthy adult animal at rest in a thermoneutral environment, the daily intake of water, nutrients, and minerals is balanced by daily excretion of these substances or their metabolic by-products. Thus, in this homeostatic state, the animal does not experience a net gain or loss of water, nutrients, or minerals and is said to be in zero balance. In a sedentary dog or cat in a thermoneutral environment, obligatory daily losses of water occur (Fig. 1-7). Input is equal to output in zero balance, and the volume of water added to body fluids by food and water consumption and by metabolism is equal to the volume of water lost in urine, feces, and saliva (i.e., sensible water loss) and evaporation from cutaneous and respiratory epithelia (i.e., insensible water loss).31

Applied Physiology of Body Fluids in Dogs and Cats Drinking

Water in food

Metabolic water

Saliva GUT Fecal water Evaporation Skin Sweat

ECF

ICF

Evaporation Respiratory Panting

Free water loss: ADH Obligatory urinary water loss: Renal solute load

Figure 1-7 Total body water: daily input and obligatory losses. (Adapted from Chew RW. Water metabolism of mammals. In: Mayer WW, Van Gelder RG, editors. Physiologic mammalogy, Vol II: mammalian reaction to stressful environments. New York: Academic Press, 1965: 43–177.)

15

function of both the amount of solute in the diet and the osmolality of the urine. Diets with higher solute contents require greater total water intake than do diets of relatively lower solute content. Most small animals have free access to water and therefore ingest sufficient water to support urinary excretion of dietary solutes. Sick animals often are inactive and have a poor appetite or are anorexic. Water requirements to replace insensible losses related to activity and to support renal solute excretion thus are decreased, and maintenance water requirements presumably are lower than those in healthy individuals. Increased insensible water losses caused by fever or increased metabolic rate during disease may offset this decrease in water requirement. Basal needs must be defined accurately if water requirements during disease are to be estimated using increments of basal requirements. To address this issue, the following discussion focuses on the relationship between basal water requirements and dietary solute in sedentary small animals in a thermoneutral environment.

WATER LOSSES Although the classic definition of insensible water loss in healthy animals is water lost via the skin or lungs, in clinical veterinary medicine water lost in the feces and saliva also is included in insensible losses. This approach is used because it usually is impractical to measure fecal and salivary water losses, which are small under normal conditions. This chapter uses the clinical definition of insensible water loss. Although evaporative losses may be great in heat-stressed, exercising, or active animals, the most important and predictable obligatory daily loss of water in healthy, sedentary dogs and cats in a thermoneutral environment occurs via urine. Estimates for water input by drinking and water loss via urine, feces, or total insensible avenues in healthy dogs and cats are variable (Table 1-6). Daily maintenance fluid requirement may be defined as the volume of fluid needed each day to maintain the animal in zero fluid balance. Maintenance needs thus are determined by daily sensible and insensible losses, by ambient temperature and humidity, by the animal’s voluntary or forced activity, and by disease. A high ambient temperature, especially with low humidity, results in increased insensible evaporative losses and, therefore, in increased maintenance fluid requirements. Similarly, fever and increased metabolic rate associated with disease may increase fluid requirements. Estimates of maintenance fluid requirements during thermal stress or disease usually are based on empirical adjustments of the estimated basal fluid requirements. Maintenance fluid requirements also are determined partially by composition of the diet. In dogs and cats, most absorbed dietary nitrogen and minerals not required to maintain zero balance or to provide for growth or tissue repair are excreted daily in urine. The volume of urine required for solute excretion thus is a

URINARY AND FECAL WATER LOSS Daily urinary water losses may be divided into obligatory water loss (i.e., water needed to excrete the daily renal solute load) and free water loss (i.e., water excreted unaccompanied by solute under the control of antidiuretic hormone [ADH]). Clearance of free water increases during relative water excess, thus protecting the animal from the overhydration and hypotonicity that would result from retention of water in excess of solutes. Obligatory renal water loss must occur even in states of relative water deficit so that solute may be eliminated from the body. Similarly, a small daily, obligatory fecal water loss is required for fecal excretion of solute. Obligatory fecal water loss may increase if fecal solute increases (e.g., addition of CaCl2 or MgCl2 to the diet). These ions increase fecal solute, because Ca2þ and Mg2þ are poorly absorbed from the gastrointestinal tract. Maintenance water requirements must include at least enough water to allow renal and fecal solute excretion.

OBLIGATORY URINARY AND FECAL WATER LOSSES The amount of water required for elimination of the urinary solute load in theory depends on the maximal urine osmolality that can be achieved by the animal (Table 1-7). However, solute usually is not excreted at maximal urine osmolality, especially when water is readily available for voluntary consumption. Urinary osmolalities from experimental dogs at rest and in water balance ranged from 1000 to 2000 mOsm/kg.17 In a study of client-owned dogs, urine osmolality ranged from 161 to 2830 mOsm/kg, and urine osmolality was greater in the morning (mean, 1541  527 mOsm/kg; range,

APPLIED PHYSIOLOGY

16

TABLE 1-6

Measurements of Daily Water Intake and Output in Sedentary Dogs and Cats

Measurement Input Water drunk

Output Urine volume Fecal water Insensible loss

Species

mL/kg/day*

Condition or Diet

Reference Chew9 Thrall and Miller57 Chew9 O’Connor40

Feline Feline Canine Canine

71.3 50.6 56.1-70.8 38.9 (19.5-84)

Canine Feline

13.3 (10.5-17.9) 25-29 g/day

Caged Caged

O’Connor40 Jackson and Tovey23

Feline Canine Feline Feline Canine

56 g/day 20.5 12.42 29.0 26.2 (8.1-70.7)

Caged 69% H2O diet 70% H2O diet Dry ration Beef and biscuit

Thrall57 Smith et al54 Hamlin and Tashjian16 Thrall and Miller57 O’Connor40

*Except as noted in table.

Species Dog Dog Cat Cat Cat

Maximal Urine Osmolalities (mOsm/kg)

mOsm/kg 2425 2791 3200 3420-4980 2984

Reference Chew9 Hardy and Osborne17* Chew9 Thrall and Miller57 Ross and Finco50*

1000

600 Urine volume (ml/day)

TABLE 1-7

1200 1600 2000

400

200

*Values obtained after dehydration resulting in 5% body weight loss. 200

273 to 2620 mOsm/kg) than in the evening (mean, 1400  586 mOsm/kg; range, 161 to 2830 mOsm/ kg).58 There was no effect of sex on urine osmolality, but urine osmolality decreased significantly with age. Figure 1-8 depicts urine volume and urine osmolality plotted as a function of urine solute in a dog fed varying quantities of food.40 Increased intake produced increased renal solute and increased urine volume; however, urine osmolality remained approximately 1600 mOsm/kg (1200 to 2000 mOsm/kg).41 Urine osmolalities did not, as might be expected, increase toward the maximum attainable (2400 to 2800 mOsm/kg) in water-deprived dogs.9,17 Thus, urine osmolality is conserved in the presence of increased urine solute load by an increase in urine volume. The physiologic mechanisms that conserve urine osmolality as the renal solute load varies are not well defined. The renal solute load is derived from dietary sources of protein and minerals and comprises urea, Naþ, Kþ, Ca2þ, Mg2þ, NH4þ, and other cations and PO43, Cl, SO42, and other anions. When estimating solute load

400 600 2Na + 2K + Urea (mmol/day)

800

Figure 1-8 Urine volume of a dog plotted against urinary excretion of solute (2Na þ 2K þ urea) during consumption of 320 (1), 385 (l), and 770 (s) grams of food. Each symbol represents data from 1 day. The lines labeled 1000, 1200, 1600, and 2000 indicate urine osmolality (mOsm/kg). (From O'Connor WJ, Potts DJ. Kidneys and drinking in dogs. In: Michell AR, editor. Renal disease in dogs and cats: comparative and clinical aspects. Oxford, UK: Blackwell Scientific, 1988: 35.)

from the diet of an animal in zero balance, all nitrogen is assumed to form urea. Urea constitutes two thirds of the urinary solute load in dogs.40 The amount of solute in the diet is determined by the composition and the quantity of food and minerals ingested. Increasing dietary protein results in increased urea production. Metabolism of carbohydrates and fats yields only CO2 and H2O and does not produce urea or other solutes that must be excreted in the urine. Diets high in minerals that are well absorbed from the gut (usually NaCl) provide more solute for excretion.

Applied Physiology of Body Fluids in Dogs and Cats Not all solute produced by metabolism of ingested and absorbed food is necessarily excreted in the urine. Fecal excretion of solutes does occur. In most healthy dogs, however, daily fecal Naþ, Kþ, and Cl excretion is substantially lower than urinary mineral excretion. The daily renal solute load is thus a function of the quantity of food ingested and of diet composition. Assuming a range of urine osmolalities in healthy dogs between 1000 and 2000 mOsm/kg and a urine solute load of approximately 400 mOsm in a 10-kg dog, the range of urine output would be 200 to 400 mL or 20 to 40 mL/kg/day. Urine volume is thus a function of renal solute load. Another important factor that determines urine volume is the total quantity of water ingested per day. Total water consumption depends both on water in the diet and on water voluntarily consumed by drinking.

URINARY FREE WATER Excretion of urinary free water is controlled by stimulation or inhibition of secretion of ADH and by thirst. Urinary free water increases when enough water has been ingested to dilute body solute and result in hypotonicity. A 1% to 2% decrease in serum osmolality inhibits secretion of ADH and abolishes thirst in humans.45,46 During water depletion, body water osmolality increases and ADH secretion is stimulated. An increase in serum osmolality of 1% to 2% is sufficient to provoke maximal ADH secretion in humans.45,46 In dogs, increases in osmolality of 1% to 3% stimulate thirst.39,41 A water loss of 5 mL/kg of body weight provoked drinking in experimental dogs.47 Therefore, daily urinary free water losses are very small during water deficiency in otherwise healthy dogs and cats.

RESPIRATORY AND CUTANEOUS EVAPORATIVE LOSSES Cutaneous evaporative water losses usually are small in dogs and cats. Cats in hot environments are reported to lick themselves with saliva to promote evaporative cooling.9 This phenomenon is rarely observed in clinical practice, but if it occurs, salivary water losses could significantly increase water need. Evaporative water loss from the skin is minimal in dogs and cats because eccrine sweat glands (which are limited in distribution to the foot pads) do not participate in thermoregulation in these species. Evaporative water losses usually are less in healthy, sedentary cats in a thermoneutral environment compared with dogs (see Table 1-6), probably because cats rarely pant. Evaporative losses in caged, sedentary laboratory dogs are quite variable from dog to dog, and some individuals experience significant daily losses via this route (Table 1-8). Dogs in the study summarized in Table 1-8 fell into two categories: those that remained quiet in their cages all of the time and those that ran in circles, barking for several hours each day. The mean evaporative loss for all dogs was 27 mL/kg/day. This value overestimates evaporative loss in quiet dogs. For dogs at rest,

17

evaporative losses usually were less than 1 mL/kg/hr. During periods of activity, evaporative losses were estimated at almost 7 mL/kg/hr.40 The total water intake per day of the dogs in Table 1-8 was quite variable from dog to dog, ranging from approximately 20 to 91 mL/kg/day. If insensible loss (primarily composed of respiratory evaporative loss) for each dog is subtracted from total daily water intake, the range of water intake unrelated to insensible losses is narrower (11 to 20 mL/kg/day) than the range for total water intake. This emphasizes the profound effect that insensible losses may have on daily water balance in dogs. Increases in ambient temperature, especially in association with low relative humidity, may result in marked increases in respiratory water evaporation in dogs. The panting response to heat is more efficient in dogs than in cats. At an ambient temperature of 40 C, cats increase their respiratory rate 4.5 times, whereas dogs can increase their respiratory rate 12 to 20 times.9 The estimated respiratory water loss in a panting dog at 41 C was 469 mL/day, whereas that for a cat under the same conditions was 41 mL/day (Table 1-9).

WATER INTAKE WATER IN FOOD The percentage of water in pet foods is variable. In general, canned foods are more than 70% water, semimoist foods 20% to 40% water, and dry foods less than 10% water.29 Two representative cat diets are described in Table 1-10. Therefore water in food makes up a variable proportion of total daily water consumption, depending on what type of diet is fed (Figs. 1-9 and 1-10). Cats can exist without drinking water if fed a diet of cod, salmon, or beefsteak.43 If the beefsteak or salmon was partially desiccated, cats became hydropenic (increased serum osmolality and serum sodium concentration), anorexic, and cachectic. Thus, cats may meet their water needs solely from the water in some foods.

DRINKING The volume of water voluntarily ingested each day by healthy, sedentary dogs and cats in a thermoneutral environment depends on the composition and the quantity of the diet ingested. Water intake decreases in experimental dogs if food intake is limited (Fig. 1-11).9,10,33,34 After a 1-day fast, drinking decreased to 25% to 50% of the normal volume in dogs. After a 14- to 18-day fast, drinking was 45% of the normal volume in dogs.9 Conversely, if water intake is limited, food intake decreases in dogs and cats.43 As mentioned earlier, cats continue to eat and survive on some diets without drinking water. In dogs that are chronically deprived of food, a basal level of drinking is maintained.2 In sick, anorexic small animal patients, drinking may decrease, and because such

18

Water Intake and Urinary Losses of Solute and Water in Six Sedentary Experimental Dogs Receiving the Same Diet Water Intake

Water Loss

Average Total Water

Dog Titch Lassie Kim Gina Blackie Sandy Mean

Evaporation

Food and Average Metabolic Drunk (mL/ (mL/ Wt. (kg) (mL/day) (mL/day) day) kg/day) (mL/day) 12.7 15.2 20.5 10.7 20.7 19.8

300 300 363 187 250 250 275

854  465 291  102 482  188 135  53 153  80 151  62 344.3

1154 691 842 322 403 401 633.5

90.9 45.5 41.1 30.1 19.5 20.3 41.2

961  358 386  115 545  176 209  49 167  54 165  50 405.5

Urine

Total H2O Average Urine Average Evaporated (mL/kg/ (mL/kg/ (mOsm/ (mL/kg/ day) day) (mL/day) kg) day) 75.7 25.4 26.6 19.5 8.1 8.3 27.3

243  41 204  71 292  50 123  16 217  29 258  31 22

Adapted from O’Connor WJ, Potts DJ. The external water exchanges of normal laboratory dogs. Q J Exp Physiol 1969;54:244–265.

19.1 13.4 14.2 11.5 10.5 13 13.6

15.2 20.1 14.5 10.6 11.4 12 14

1706  441 1836  139 1519  211 2016  216 1652  305 1079  124 1634.7

Urine Solute/ Day (mOsm) 415 375 444 248 359 278 353

APPLIED PHYSIOLOGY

TABLE 1-8

Applied Physiology of Body Fluids in Dogs and Cats

TABLE 1-9

Respiratory Water Losses of Panting Mammals* Respiratory Water Loss

Species

19

Weight (kg)

(g/min)

(g/day)

Percentage Heat Production Lost

16 3.5

0.326 0.029

469 41.2

57 9.4

Dog Cat

Data from Chew RM. Water metabolism of mammals. In: Mayer WW, Van Gelder RG, editors. Physiologic mammalogy, Vol II: mammalian reaction to stressful environments. New York: Academic Press, 1965: 43–177. *Temperature, 41˚ C; relative humidity, 32%.

TABLE 1-10

Effect of Diet on Water Intake in Cats

Food

Dry Matter Intake (g/day)

Food Water* (g/day)

Water Drunk{ (mL/day)

Total Water Intake (mL/day)

Ratio of Total Water to Dry Matter

Dry Canned

76.9  17.4 35.2  7.2

7.4  1.7 (8.8) 116  23.6 (76.8)

167.2  40.1 (>90) 22.8  12.8 (14)

174.6  41.6 139.0  31.4

2.3  0.2 3.9  0.3

Data from Seefeldt SL, Chapman TE. Body water content and turnover in cats fed dry and canned rations. Am J Vet Res 1979;40:183–185. *Figures in parentheses represent approximate percentage of diet that was water. { Figures in parentheses represent approximate percentage of total water intake that was drunk.

Water intake (ml/day)

3000

2000

1000

CD

CBD

SMD

SDD

DD

Food type

Figure 1-9 Effect of food type on water intake in dogs. Each column represents the total daily water intake (mean  SD) for four dogs fed different diets. The solid area shows the amount of endogenous food water; the clear area shows water drunk. CD, canned; CBD, canned meat and biscuit mixture; SMD, SDD, intermediate moisture foods; DD, dry. (From Burger IH, Anderson RS, Holme DW. Nutritional factors affecting water balance in the dog and cat. In: Anderson RS, editor. Nutrition of the dog and cat. Oxford, UK: Pergamon Press, 1980: 149.)

animals do not have access to water from food, total water intake may decrease drastically. However, the water requirement of such animals is probably quite low. In

quiet, sick animals, the major obligatory water loss occurs via urine (assuming no other major contemporary fluid loss, such as in diarrhea or vomitus). The renal solute load and obligatory renal water loss decrease because the animal is not eating. However, animals in a catabolic state obviously do produce urea and ions for excretion as a result of catabolism of lean body mass. Figures for renal solute loads generated from endogenous sources are not readily available in the literature. Water requirements of a sick animal may be increased if the animal is febrile, having seizures, or experiencing abnormal losses, such as in vomitus or diarrhea. These contemporary water needs are in addition to the maintenance water required to maintain zero balance during inanition and inactivity in the presence of diminished but still present obligatory urinary water losses. The volume of water drunk increases as the water in the diet decreases (see Table 1-10). Dogs maintain a uniform total water intake when food water is decreased by commensurately increasing drinking (see Fig. 1-9). However, cats may not increase drinking enough to maintain total water intake when consuming a diet low in water (see Fig. 1-10). Cats receiving dry food diets may ingest insufficient water. This issue has been investigated extensively as a contributing factor in the development of lower urinary tract disease in cats. Some investigators believe that a low ratio of total water intake to dry matter in the diet predisposes a cat to lower urinary tract disease. Diets with a ratio of total water to dry matter greater than 3 have been suggested as an aid in the

APPLIED PHYSIOLOGY

20 340

TABLE 1-11

Water intake (ml/day)

300

Investigator

Food type Salt content (% D.M.)

Thrall and Miller58 Jackson and Tovey24 Holme23 Seefeldt and Chapman53 Jenkins and Coulter24a

CCb 1.6

a

DCac 3.6

DC 1.3

DCbc 4.6

SMCabc 3.7

Figure 1-10 Effects of food type and salt content on water intake in cats. Each column represents the total daily water intake (mean  SD) for cats on various diets. The same group of six cats was used for all foods except a DC diet 4.6% salt, data for which were obtained from a different experiment using another group of 12 cats. The solid area shows food water, and the clear area shows water drunk. Total water intake for foods bearing different superscript letters is significantly different (P 1 to 2 mL/kg/hr in dogs and cats). The term solute, or osmotic, diuresis refers to increased urine flow caused by excessive amounts of nonreabsorbed solute within the renal tubules (e.g., polyuria associated with diabetes mellitus, administration of mannitol). During osmotic diuresis, urine osmolality approaches plasma osmolality. The term water diuresis refers to increased urine flow caused by decreased reabsorption of solutefree water in the collecting ducts (e.g., polyuria associated with psychogenic polydipsia or diabetes insipidus). During water diuresis, urine osmolality is less than plasma osmolality. The term isosthenuria refers to urine with an osmolality equal to that of plasma, and hyposthenuria refers to urine with an osmolality less than that of plasma. The term hypersthenuria, or baruria, refers to urine with an osmolality greater than that of plasma, but this term is rarely used and only to describe urine that is very concentrated.

TYPES OF DEHYDRATION Dehydration occurs when fluid loss from the body exceeds fluid intake. Dehydration may be classified according to the type of fluid lost from the body and the tonicity of the remaining body fluids. Pure water loss and loss of hypotonic fluid result in hypertonic dehydration because the tonicity of the remaining body fluids is increased. Loss of fluid with the same osmolality as that of ECF results in isotonic dehydration, because there is no osmotic stimulus for water movement and the

Disorders of Sodium and Water: Hypernatremia and Hyponatremia remaining body fluids are unchanged in tonicity. Loss of hypertonic fluid or loss of isotonic fluid with water replacement results in hypotonic dehydration because the remaining body fluids become hypotonic. The types of dehydration and their relative effects on the volume and tonicity of the intracellular and extracellular compartments are shown in Figure 3-1.

SERUM SODIUM CONCENTRATION The serum sodium concentration is an indication of the amount of sodium relative to the amount of water in the ECF and provides no direct information about total body sodium content. Patients with hyponatremia or hypernatremia may have decreased, normal, or increased total body sodium content. An increased serum sodium concentration (hypernatremia; >155 mEq/L in dogs or >162 mEq/L in cats) implies hyperosmolality, whereas a decreased serum sodium concentration (hyponatremia; 155 mEq/L in dogs; >162 mEq/L in cats)

Evaluation of volume status

Hypervolemia

Normovolemia

Hypovolemia

Impermeant solute gain

Pure water deficit

Hypotonic loss

Salt poisoning Hyperaldosteronism Hypertonic fluid administration

Primary hypodipsia Central diabetes insipidus Nephrogenic diabetes insipidus High environmental temperature Fever Inadequate access to water

Extrarenal (FCNa 1%)

Gastrointestinal • Vomiting • Diarrhea • Small intestinal obstruction

Appropriate • Osmotic diuresis (mannitol, hyperglycemia) • Chemical diuresis (furosemide, ethacrynic acid)

Third space loss • Pancreatitis • Peritonitis

Inappropriate • Chronic renal failure • Nonoliguric acute renal failure • Post-obstructive diuresis

Cutaneous • Burns The following clinical information suggests hypovolemia: a history of hypotonic losses (e.g., vomiting, diarrhea), decreased skin turgor, dry mucous membranes, delayed capillary refill time, tachycardia, flat jugular veins, systemic hypotension, low central venous pressure. Hypervolemic patients may develop pulmonary edema, especially those with underlying cardiac disease.

Figure 3-7 Clinical approach to the patient with hypernatremia. FCNa, Fractional clearance of sodium.

Rarely, chronic hypernatremia may occur in fully conscious animals that have access to water. In these cases, abnormal osmoregulation of ADH release caused by underlying hypothalamic lesions results in hypodipsia. Animals that are unable to obtain water because central nervous system disease has resulted in an altered sensorium may also be hypernatremic; but in these instances, the hypernatremia is simply a result of water deprivation. Hypodipsic hypernatremia related to defective osmoregulation of ADH has been reported in a dog with hydrocephalus and normal pituitary function.31 In normal individuals, administration of hypertonic saline increases plasma osmolality and simultaneously causes volume expansion. Osmoreceptors are stimulated by hyperosmolality but inhibited by volume expansion. Normally, the response to hyperosmolality takes precedence, and ADH secretion increases, resulting in decreased urine volume and increased urine osmolality. The affected dog experienced increased urine volume and decreased urine osmolality in response to an infusion of hypertonic saline, indicating defective osmoreceptor function as observed in human patients with hypodipsic hypernatremia. Similarly, destruction of osmoreceptors in the hypothalamus was thought to be responsible for adipsia and hypernatremia in a dog with focal granulomatous meningoencephalitis.104 Weakness and polymyopathy have been reported in a young cat with hypodipsia, hypernatremia, and hypertonicity associated with hydrocephalus and hypopituitarism, and hypernatremia,

adipsia, and diabetes insipidus have been observed in a young dalmatian dog with dysplasia of the rostral diencephalon.5,34 Hypernatremia also has been reported in a dog63 and cat115 with central nervous system lymphoma. Hypodipsia, hypernatremia, and hypertonicity caused by an abnormal thirst mechanism have been reported in young female miniature schnauzers and in a young Great Dane.27,70,76,112,159 One miniature schnauzer with hypodipsic hypernatremia had severe behavioral disturbances, and holoprosencephaly was found at necropsy.153 Another had dysgenesis of the corpus callosum and other forebrain structures.112 Grossly visible neuroanatomic abnormalities were not identified in a previous report.27 Whether a spectrum of neuroanatomic abnormalities exists in these dogs (which appear to have a form of congenital adipsic hypernatremia) is not known. Infusion of hypertonic saline has been shown to lead to an increase in urine volume and a decrease in urine osmolality compatible with defective osmoregulation of ADH.27 Clinical signs in affected dogs are associated with hypertonicity and include anorexia, lethargy, weakness, disorientation, ataxia, and seizures. Affected dogs can be managed clinically by addition of water to their food, but hypernatremia and neurologic dysfunction recur whenever water supplementation is discontinued. In a Norwegian elkhound with adipsic hypernatremia, the adipsia resolved spontaneously at 2 years of age,62 and an adipsic Labrador retriever with hypothyroidism responded to treatment with levothyroxine.81

ELECTROLYTE DISORDERS

56

2L

1L

ECF

ICF

600 mOsm

1200 mOsm

300 mOsm/Kg

300 mOsm/Kg

ECF

ICF

4L

H2O 1200 mOsm 1L

1.67 L

4L

600 mOsm

ECF

ICF

600 mOsm

1200 mOsm

360 mOsm/Kg

3.33 L

360 mOsm/Kg

Figure 3-8 Effect of loss of 1 L of water on volume and tonicity of extracellular fluid (ECF) and intracellular fluid (ICF). (Drawing by Tim Vojt.)

Central or pituitary diabetes insipidus (CDI) is caused by a partial or complete lack of vasopressin production and release from the neurohypophysis.65 It may result from trauma or neoplasia or may be idiopathic in dogs and cats.* Visceral larva migrans also has been reported to cause CDI in a dog.99 In one dog with hypernatremia, hypertonicity, and gastric dilation-volvulus, CDI was present and caused by neurohypophyseal atrophy secondary to a cystic craniopharyngeal duct.36 Congenital CDI is rare58,86,163 but has been reported in two sibling Afghan pups.131 Traumatic CDI may be transient in nature. Hypophysectomy for treatment of hyperadrenocorticism results in transient CDI that may take several weeks to resolve.102 Marked hypernatremia occurs in dogs in the first 24 hours after hypophysectomy and can be prevented by prophylactic treatment with desmopressin (DDAVP).64 In the month after surgery, serum sodium concentrations in control dogs were not markedly different from those observed in the DDAVP*References 4, 18, 29, 45, 57, 58, 72, 106, 119, 120, 130, 133, 140.

treated group, suggesting that the dogs with untreated CDI drank sufficient water to maintain relatively normal plasma osmolality. The transient nature of CDI after hypophysectomy may result from the fact that some of the vasopressin-producing neurons from the hypothalamus terminate in the median eminence. Animals with CDI have severe polydipsia and polyuria. Their urine typically is hyposthenuric (urine osmolality, 60 to 200 mOsm/kg), but urine osmolality may approach 400 to 500 mOsm/kg in the presence of dehydration. Variability in USG and urine osmolality values at the time of presentation in dogs and cats with diabetes insipidus presumably is related to hydration status and severity of vasopressin deficiency. In one study, dogs were classified as having complete or partial CDI based on the magnitude of increase in their USG and urine osmolality after induction of 5% dehydration.65 Dogs with complete CDI had USG values of 1.001 to 1.007 that did not change substantially after induction of 5% dehydration, whereas dogs with partial CDI had USG values of 1.002 to 1.016 that increased to 1.010 to 1.018 after induction of 5% dehydration. In both groups, there was a substantial (>50%) increase in USG 2 hours after administration of 1 to 5 U of aqueous arginine vasopressin. Affected dogs responded well to administration of DDAVP acetate (1 to 2 drops in both eyes every 12 to 24 hours), but the prognosis was dependent on the underlying cause of CDI. Many older dogs with CDI had tumors in the region of the pituitary gland and developed neurologic signs. Increased plasma osmolality and hypernatremia may occur in dogs and cats with CDI. These results suggest that some affected dogs and cats do not obtain enough water to maintain water balance and are presented in a hypertonic state. Severe hypernatremia and neurologic dysfunction may occur if the animal cannot maintain adequate water intake.36,133 In contrast, with psychogenic polydipsia, plasma osmolality and serum sodium concentration may be lower than normal at presentation.91 Administration of vasopressin leads to an increase in urine osmolality or specific gravity in dogs and cats with CDI, but the initial response may be less than expected because of renal medullary washout of solute. In one study, USG values increased to 1.018 to 1.022 after vasopressin administration in dogs with complete CDI and to 1.018 to 1.036 in dogs with partial CDI.65 Treatment with vasopressin restores medullary hypertonicity and normal urinary concentrating ability. Historically, vasopressin tannate in oil (pitressin tannate) has been used to treat CDI in small animal practices. The dosage is 3 to 5 U for dogs or 1 to 2 U for cats given intramuscularly or subcutaneously every 24 to 72 hours as needed to control polyuria and polydipsia. To avoid the possibility of water intoxication, it is recommended that the treatment interval be determined by recurrence of polyuria. This product is no longer commercially available.

Disorders of Sodium and Water: Hypernatremia and Hyponatremia DDAVP is a structural analogue of vasopressin (see Fig. 3-4) that has a more potent antidiuretic effect than vasopressin but a minimal vasopressive effect and is relatively resistant to metabolic degradation. DDAVP is available as a nasal spray (0.1 mg/mL), injectable solution (4 mg/mL), or tablet for oral administration (0.1 and 0.2 mg). The injectable solution is much more expensive than the nasal spray, and the nasal spray has been used subcutaneously in dogs and in a cat with CDI at a dosage of 1 mg/kg without adverse effects.86,87 Polyuria and polydipsia in a cat with CDI were controlled with 1 mg/kg administered subcutaneously every 12 hours or 1.5 mg/kg administered conjunctivally every 8 hours. One drop of the nasal spray contains 1.5 to 4 mg of DDAVP, and the duration of effect varies from 8 to 24 hours.43 In humans, the bioavailability of DDAVP after oral administration was 0.1% as compared with 3% to 5% after intranasal administration, and gastrointestinal absorption was improved when it was given in a fasted state.46,136 In dogs, an antidiuretic effect was observed even after orally administered doses as low as 50 mg.160 Chlorpropamide is a sulfonylurea hypoglycemic agent that potentiates the renal tubular effects of small amounts of vasopressin and may be useful in management of animals with partial CDI. Its effect may occur by up-regulation of ADH receptors in the kidneys.35 The recommended dosage of chlorpropamide is 10 to 40 mg/kg/day orally, and hypoglycemia is a potential adverse effect. It has been useful in the management of CDI (up to 50% reduction in urine output) in some reports but not in others, possibly because some animals have partial and some have complete CDI.86,140 In the broadest sense, the term nephrogenic diabetes insipidus (NDI) may be used to describe a diverse group of disorders in which structural or functional abnormalities interfere with the ability of the kidneys to concentrate urine (Box 3-2).13,90 Congenital NDI is a rare disorder in small animal medicine.13,80,90 Affected animals are presented at a very young age for severe polyuria and polydipsia. In reported cases, urine osmolality and specific gravity have been in the hyposthenuric range. Affected animals show no response to water deprivation testing, exogenous vasopressin administration, or hypertonic saline infusion. In one case report, the plasma vasopressin concentration was markedly increased.80 Congential NDI in human patients can arise from mutations in the V2 receptor (X-linked recessive inheritance) or from mutations in the AQP2 channel (autosomal recessive inheritance). Low affinity V2 receptors were thought to be responsible for congenital NDI in a family of Siberian huskies.103 Thiazide diuretics (chlorothiazide 20 to 40 mg/kg every 12 hours or hydrochlorothiazide 2.5 to 5.0 mg/kg twice a day) have been used to treat animals with CDI and NDI. Diuretic administration results in mild dehydration, enhanced proximal renal tubular reabsorption of

BOX 3-2

57

Causes of Nephrogenic Diabetes Insipidus

Congenital (primary) Acquired (secondary) Functional Drugs Glucocorticoids Lithium Demeclocycline Methoxyflurane Escherichia coli endotoxin (e.g., pyelonephritis, pyometra) Diuretics Electrolyte disturbances Hypokalemia Hypercalcemia Altered medullary hypertonicity Hypoadrenocorticism Multifactorial or unknown mechanism Hepatic insufficiency Hyperthyroidism Hyperadrenocorticism Postobstructive diuresis Acromegaly Structural Medullary interstitial amyloidosis (e.g., in cats, SharPei dog) Polycystic kidney disease Chronic pyelonephritis Chronic interstitial nephritis

sodium, decreased delivery of tubular fluid to the distal nephron, and reduced urine output. Thiazides have been reported to result in a 20% to 50% reduction in urine output in dogs with NDI and in cats with CDI.13,18,86,90,154 In other reports, thiazides were reported to be ineffective in reducing urine output in a dog and a cat with CDI.57,72 Restriction of dietary sodium and protein reduces the amount of solute that must be excreted in the urine each day and thus further reduces obligatory water loss and polyuria. A low-salt diet and hydrochlorothiazide (2 mg/kg orally twice a day) were used successfully to manage a dog with congenital NDI for 2 years.154 The dog’s water consumption decreased from an average of approximately 900 mL/kg/day to 200 mL/kg/day with treatment.

HYPOTONIC FLUID LOSS When hypotonic fluid is lost from the extracellular compartment, the osmotic stimulus for water to move from the intracellular to the extracellular compartment is less than the stimulus for water movement created by pure water loss. Thus, hypotonic losses cause a greater reduction in the ECF volume, and the animal is more likely to show clinical signs of volume depletion (e.g., tachycardia,

ELECTROLYTE DISORDERS

58

weak pulses, and delayed capillary refill time). As the tonicity of the fluid lost increases toward the normal tonicity of ECF, the volume deficit of the extracellular compartment becomes progressively more severe (Fig. 3-9). In the case of isotonic losses, no osmotic stimulus for water movement is present. The entire loss is borne by the extracellular compartment, and hypovolemic shock may occur if the loss has been of sufficient magnitude (e.g., severe hemorrhage). Consider what would occur in the previous example if our 10-kg dog had suffered a loss from the extracellular compartment of 1 L of fluid with an osmolality of 150 mOsm/kg. Such a loss would leave 450 mOsm of solute and 1 L of water in the extracellular compartment. Once again, water moves from the intracellular to the extracellular compartment until the osmolality has been equalized. Thus: New ECF osmolality ¼ new ICF osmolality 450 mOsm=ð1 þ xÞ L ¼ 1200 mOsm=ð4  xÞL

tonicity, the greater the volume loss from the ECF compartment. For simplicity, these examples are based on many assumptions that in reality may not be true. For example, TBW is not 60% of body weight in all individuals, the number of osmoles in the ECF may have been altered by electrolyte losses not detected clinically, the effects of hydrostatic forces resulting from extracellular volume depletion have not been considered, some solutes may not be strictly impermeant, and compensatory physiologic responses have not been considered. Nonetheless, such calculations are helpful in understanding the pathophysiology of hypertonic states, and they provide useful clinical approximations. Hypotonic fluid losses are the most common type encountered in small animal medicine. They may be classified as extrarenal (e.g., gastrointestinal, third-space loss, and cutaneous) or renal. Causes of gastrointestinal losses include vomiting, diarrhea, and small intestinal obstruction; causes of third-space losses include pancreatitis and peritonitis. Cutaneous losses are usually not clinically

where x is the volume of water moving between compartments:

ECF

ICF

600 mOsm

1200 mOsm

300 mOsm/Kg

300 mOsm/Kg

ECF

ICF

450ð4  xÞ ¼ 1200ð1 þ xÞ x ¼ 0:36 L 2L

The new volumes and osmolalities are: ECF : ICF :

450 mOsm=1:36 L 1200 mOsm=3:64 L

¼ ¼

330 mOsm=kg 330 mOsm=KG

Note that the extracellular volume deficit is more severe than in the previous example of pure water loss (0.64 L vs. 0.33 L). These changes are depicted in Figure 3-10. The more closely the fluid lost approximates ECF in

1L

150 mOsm 1200 mOsm

1.0 1L

ECF volume deficit (L)

0.9

4L

4L

450 mOsm

0.8 0.7

ECF

ICF

0.6 0.5 0.4 1.36 L

0.3 0.2

450 mOsm

1200 mOsm

3.64 L

0.1 0

25 50 75 100 125 150 175 200 225 250 275 300 Tonicity of fluid lost (mOsm/kg)

Figure 3-9 Magnitude of extracellular fluid (ECF) volume deficit caused by loss of 1 L of fluid of varying tonicity.

330 mOsm/Kg

330 mOsm/Kg

Figure 3-10 Effect of loss of 1 L of hypotonic fluid (150 mOsm/ kg) on volume and tonicity of extracellular fluid (ECF) and intracellular fluid (ICF). (Drawing by Tim Vojt.)

Disorders of Sodium and Water: Hypernatremia and Hyponatremia important in dogs and cats. Eccrine sweat glands are limited to the foot pads and serve no thermoregulatory function, and burns are encountered uncommonly in small animal practice. Renal losses may result from osmotically (e.g., diabetes mellitus, mannitol) or chemically (e.g., furosemide, corticosteroids) induced diuresis or from defective urinary concentrating ability related to intrinsic renal disease (e.g., chronic renal failure, nonoliguric acute renal failure, postobstructive diuresis).

200 mOsm (5.85 g NaCl)

2L

GAIN OF IMPERMEANT SOLUTE Gain of impermeant solute is uncommon in small animal medicine. The addition of a sodium salt to ECF causes hypernatremia, whereas gain of an impermeant solute that does not contain sodium (e.g., glucose and mannitol) initially causes hyponatremia because water is drawn into ECF. However, hypernatremia occurs as osmotic diuresis develops because urine osmolality approaches plasma osmolality and the sodium-free solute replaces sodium in urine. The sodium displaced from urine remains in the ECF and contributes to hypernatremia. The development of hypertonicity as a result of excessive salt ingestion is unlikely if the animal in question has an intact thirst mechanism and access to water. The addition of impermeant solute without water expands the extracellular compartment at the expense of the intracellular compartment as water moves from ICF to ECF to equalize osmolality. This volume overload may lead to pulmonary edema if the patient has underlying cardiac disease. Consider again our example of the 10-kg dog. The addition of 200 mOsm of solute to the ECF without any water would be equivalent to ingestion of 5.85 g of sodium chloride (5.85 g NaCl ¼ 100 mmol Na and 100 mmol Cl). The addition of this impermeant solute to ECF causes movement of water from the intracellular to extracellular compartments until osmolality has been equalized. Thus: New ECF osmolality ¼ new ICF osmolality 800 mOsm=ð2 þ xÞ L ¼ 1200 mOsm=ð4  xÞL where x is the volume of water moving between compartments: 800ð4  xÞ ¼ 1200ðz þ xÞ x ¼ 0:4 L The new volumes and osmolalities are: ECF : 800 mOsm=2:4 L ICF : 1200 mOsm=3:6 L

¼ ¼

333 mOsm=kg 333 mOsm=kg

Note that ECF volume has been expanded by 0.4 L and that this volume has been derived from ICF. In the normal animal, this expansion of the extracellular compartment

ECF

ICF

600 mOsm

1200 mOsm

300 mOsm/Kg

300 mOsm/Kg

59

4L

ECF ICF

800 mOsm 1200 mOsm

2.4 L

333 mOsm/Kg

3.6 L

333 mOsm/Kg

Figure 3-11 Effect of addition of 200 mOsm solute (5.85 g NaCl) on volume and tonicity of extracellular fluid (ECF) and intracellular fluid (ICF). (Drawing by Tim Vojt.)

leads to natriuresis, and the volume deficit is repaired by ingestion of water in response to plasma hyperosmolality. These changes are depicted in Figure 3-11. In one report of salt poisoning in dogs, a defective water softener resulted in delivery of drinking water containing 10% sodium chloride as compared with normal tap water containing less than 0.1%.78 The affected dogs developed progressive ataxia, seizures, prostration, and death. Their serum sodium concentrations ranged from 185 to 190 mEq/L. Histopathology showed focal areas of perivascular hemorrhage and edema in the midbrain. In another case report, presumptive salt poisoning resulted from ingestion of seawater and subsequent restriction of fresh drinking water.23 Another dog developed fatal hypernatremia after it ingested a large amount of a salt-flour mix.84 After ingestion of a salt-flour figurine, the dog began vomiting and developed polyuria and polydipsia. The owner removed the dog’s water source, and it ingested more of the salt-flour mix. Seizures, pyrexia, and sinus tachycardia developed, and the serum sodium concentration reached 211 mEq/L. Approximately 22% of dogs that ingest paintballs (which may contain polyethylene glycol, glycerol, and sorbitol) develop hypernatremia.32 Hyperchloremia and hypokalemia also are reported. Clinical signs include

60

ELECTROLYTE DISORDERS

vomiting, ataxia, diarrhea, and tremors. These ingredients act as osmotic laxatives, causing a shift in water from the tissues into the lumen of the bowel and resulting in hypernatremia. Warm water enemas may facilitate removal of paintball ingredients from the bowel, but activated charcoal products generally should not be used because they may contain sorbitol. Depending on the duration of onset, 5% dextrose in water (acute onset) or 0.45% NaCl (unknown onset) can be administered to gradually correct hypernatremia. Parenteral fluids can be supplemented with potassium chloride if serum potassium concentration decreases below 2.5 mEq/L. Therapeutic administration of hyperosmolar solutions containing large amounts of sodium during cardiac resuscitation can cause hypernatremia and hypertonicity (e.g., hypertonic saline, sodium bicarbonate). For example, serum sodium concentration reached 174 mEq/L within 15 minutes after beginning infusion of 7.2% NaCl at a rate of 15 mL/kg in normal beagles.2 Sodium phosphate enemas may also result in mild hypernatremia.3 Primary hyperaldosteronism also may be associated with hypernatremia. It is rare in dogs, but several cases have been reported in cats (see Chapter 5 ). Mild hypernatremia also may occur in dogs with hyperadrenocorticism.101,128

patients with underlying cardiac disease. Patients with CDI or NDI typically are presented for evaluation of severe polydipsia and polyuria.

CLINICAL SIGNS OF HYPERNATREMIA

If we assume that body water (TBW) is 60% of body weight measured in kilograms (Wt) and that 2.1  PNa is an estimate of Posm:

The clinical signs of hypernatremia primarily are neurologic and related to osmotic movement of water out of brain cells. A rapid decrease in brain volume may cause rupture of cerebral vessels and focal hemorrhage. The severity of clinical signs is related more to the rapidity of onset of hypernatremia than to the magnitude of hypernatremia. In dogs and cats, clinical signs of hypernatremia are observed when the serum sodium concentration exceeds 170 mEq/L.66,78,84,133 If hypernatremia develops slowly, the brain has time to adapt to the hypertonic state by production of intracellular solutes (e.g., inositol and amino acids) called osmolytes or idiogenic osmoles. These substances prevent dehydration of the brain and allow patients with chronic hypernatremia to be relatively asymptomatic. Where described in dogs and cats, clinical signs of hypernatremia and hypertonicity have included anorexia, lethargy, vomiting, muscular weakness, behavioral change, disorientation, ataxia, seizures, coma, and death.* If hypotonic losses are the cause of hypernatremia, clinical signs of volume depletion (e.g., tachycardia, weak pulses, and delayed capillary refill time) may be observed on physical examination. If hypernatremia has developed as a result of a gain of sodium, signs of volume overload (e.g., pulmonary edema) may be observed, especially in *References 5, 23, 27, 31, 34, 78, 84, 133, 159.

TREATMENT OF HYPERNATREMIA The main goals in treating patients with hypernatremia are to replace the water and electrolytes that have been lost and, if necessary, to facilitate renal excretion of excess sodium. The first priority in treatment should be to restore the ECF volume to normal. The next priority is to diagnose and treat the underlying disease responsible for the water and electrolyte deficits.

PURE WATER LOSS Total body solute (TBS) is the product of TBW and plasma osmolality (Posm). If a patient’s fluid loss has been limited to pure water, the following relationship is true: TBSðpresentÞ ¼ TBSðpreviousÞ TBW ðpresentÞ  Posm ðpresentÞ ¼ TBW ðpreviousÞ Posm ðpreviousÞ

2:1  PNa ðpresentÞ  0:6  WtðpresentÞ ¼ 2:1  PNa ðpreviousÞ  0:6 WtðpreviousÞ This equation reduces to: PNa ðpresentÞ  WtðpresentÞ ¼ PNa ðpreviousÞ  WtðpreviousÞ WtðpreviousÞ ¼

PNa ðpresentÞ  WtðpresentÞ PNa ðpreviousÞ

The water deficit is the difference between the previous and present body weights: WtðpreviousÞ  WtðpresentÞ ¼ PNa ðpresentÞ  WtðpresentÞ  WtðpresentÞ PNa ðpreviousÞ or


PNa ðpresentÞ 1 WtðpresentÞ  PNa ðpreviousÞ



Consider a previously normal dog that has been deprived of water for several days. The dog weighs 10 kg at

Disorders of Sodium and Water: Hypernatremia and Hyponatremia presentation, and its serum sodium concentration is 170 mEq/L. Assuming a previously normal serum sodium concentration of 145 mEq/L, the dog’s water deficit can be calculated: 0

1

PNa ðpresentÞ  1A PNa ðpreviousÞ 1

Water deficit ¼ WtðpresentÞ  @ 0 Water deficit ¼ 10  @

170  1A ¼ 1:72 L 145

The original estimates of TBW and serum sodium concentration may be modified based on information available to the clinician at presentation. For example, if the dog’s normal serum sodium concentration is known from a previous admission, this value can be substituted in place of 145 mEq/L. If the dog’s previous normal body weight is known, the water deficit may simply be estimated as the difference between the previous and present body weights. The assumption inherent in the latter calculation is that the patient has not gained or lost tissue mass. For a short period, this is a reasonable assumption because loss of 1 kg of tissue mass requires an expenditure of approximately 1600 kcal. This caloric expenditure would require fasting for 2 to 3 days in a normal 10-kg dog with a basal energy requirement of approximately 700 kcal. A pure water deficit can be replaced by giving 5% dextrose in water intravenously. This solution technically is only slightly hypotonic to plasma (278 mOsm/kg), but the glucose ultimately enters cells and is metabolized so that administration of 5% dextrose is equivalent to administration of water. The water deficit must be replaced and hypernatremia corrected slowly over 48 hours. The brain adapts to hypertonicity by the production of osmolytes or idiogenic osmoles that prevent cellular dehydration. Excessively rapid lowering of the serum sodium concentration may result in movement of water into brain cells and development of cerebral edema. In human patients with hypernatremia of chronic or unknown duration, correction of the serum sodium concentration at a rate of less than 10 to 12 mEq/L per 24 hours minimizes the risk of neurologic complications related to water intoxication.6,152 The animal’s serum sodium concentration should be monitored serially during replacement of the water deficit.

HYPOTONIC LOSS As described earlier, hypotonic losses cause more severe extracellular volume contraction than do losses of pure water. As the tonicity of the fluid lost approaches the tonicity of ECF, the extracellular volume deficit becomes greater (see Fig. 3-9). As a result, signs of volume depletion are more likely with hypotonic losses, and the

61

original replacement fluid should be isotonic so that extracellular volume repletion can proceed rapidly. In the presence of hemorrhagic shock, whole blood, plasma, or a colloid solution is the ideal fluid to administer. The hemoglobin in whole blood improves oxygencarrying capacity. The plasma proteins in whole blood and plasma or the dextrans in a colloid solution increase and maintain intravascular volume by increasing oncotic pressure. In many animals that have experienced severe hypotonic losses over an extended time period, replacement of the ECF volume with an isotonic crystalloid solution (e.g., 0.9% NaCl and lactated Ringer’s solution) is adequate. A volume up to four times the suspected intravascular deficit may be required because the isotonic crystalloid solution distributes rapidly throughout the ECF compartment (ECF volume is four times intravascular volume). After the extracellular volume has been expanded, hypotonic fluids (e.g., 0.45% NaCl and halfstrength lactated Ringer’s solution) can be administered to provide fluids for maintenance needs and ongoing losses (see Chapter 14).

GAIN OF IMPERMEANT SOLUTE The patient with an excess of sodium-containing impermeant solute in the ECF can be treated by administration of 5% dextrose intravenously. The main disadvantage of this approach is that it causes further expansion of the extracellular compartment in a patient already suffering from ECF volume expansion. In an animal with normal cardiac and renal function, this volume expansion leads to diuresis and natriuresis, and ECF volume returns to normal. In an animal with underlying cardiac disease or oliguria related to primary renal disease, this approach may lead to development of pulmonary edema. Administration of a loop diuretic (e.g., furosemide and ethacrynic acid) promotes excretion of the existing sodium load and hastens return of ECF volume to normal. As in the case of pure water deficit, it is essential that fluid administration proceeds slowly and that serum sodium concentration be lowered gradually over 48 hours to avoid neurologic complications.

CLINICAL APPROACH TO THE PATIENT WITH HYPONATREMIA The presence of hyponatremia usually, but not always, implies hypoosmolality. Thus, the first step in the approach to the patient with hyponatremia is to determine whether hypoosmolality of the ECF is present. This can be determined by measurement of plasma osmolality. The evaluation of hyponatremia then may be approached using the patient’s plasma osmolality as a guide. This approach is outlined in Fig. 3-12, and the causes of hyponatremia are listed in Box 3-3.

62

ELECTROLYTE DISORDERS Decreased serum sodium concentration ( 300 mOsm/kg in human patients with SIADH. A urine osmolality >100 mOsm/kg should be considered abnormal in a patient with hyponatremia and plasma hypoosmolality. A urine osmolality 20 mEq/L in human patients). 5. No evidence of hypovolemia, which could result in nonosmotic stimulation of vasopressin release. 6. No evidence of ascites or edema, which could result in hypervolemic hyponatremia (i.e., no evidence of severe liver disease, congestive heart failure, or nephrotic syndrome). 7. Correction of hyponatremia by fluid restriction. Impaired osmoregulation of vasopressin release was observed in 11 dogs with liver disease (7 of which had large congenital portosystemic shunts).143 Either the threshold for vasopressin release was increased or the magnitude of response decreased in these dogs, but plasma vasopressin and sodium concentrations were within the normal reference range. Affected dogs had

67

evidence of excessive glucocorticoid secretion, and their response was similar to that previously described for dogs with spontaneous hyperadrenocorticism.8 Cerebral salt wasting occurs in critically ill human patients with intracranial injury, often subarachnoid hemorrhage, but also may be observed after neurosurgical procedures.7,26 Hyponatremia resulting from cerebral salt wasting must be differentiated from that caused by SIADH because patients with the former disorder are volume depleted and require NaCl and water replacement, whereas those with SIADH require water restriction. Atrial and brain natriureticfactors likelyare responsiblefor the urinary loss of sodium in affected patients.110 Recognition of volume depletion depends on clinical findings such as changes in skin turgor, systemic blood pressure, central venous pressure, heart rate, and character of peripheral pulses. Hyponatremia must be corrected slowly (see Treatment of Hyponatremia section), and SIADH should be suspected if hyponatremia worsens after saline infusion. Fludrocortisone also has been used in human patients with cerebral salt wasting to facilitate sodium retention. Severe hypothyroidism with myxedema in humans can result in hyponatremia, possibly because of decreased distal delivery of tubular fluid and nonosmotic stimulation of vasopressin release. Hyponatremia in this setting is corrected by thyroid hormone replacement. In four reported cases of myxedema coma in dogs, hyponatremia was found in two of three dogs in which the serum sodium concentration was measured.20,83 Exercise-associated hyponatremia has been reported in human athletes during prolonged exertion. It is thought to be caused by excessive consumption of water or hypotonic fluids as well as nonosmotic stimulation of vasopressin release associated with volume depletion that impairs water excretion.141 Sodium loss in sweat also may play a contributory role. Women, people of low body weight, and those taking nonsteroidal anti-inflammatory drugs are at increased risk.138 Total body exchangeable cation content (sodium and potassium) decreases during longdistance exercise in Alaskan sled dogs despite no apparent change in TBW, and consequently hyponatremia develops.74 The observed hyponatremia was mild (i.e., 139.7 mEq/L after the race as compared with 148.6 mEq/L before the race).75 Dogs do not sweat appreciably, and the mild hyponatremia was attributed to increased urinary losses of sodium in association with increased protein catabolism and excretion of large amounts of urea in urine. Drugs that stimulate the release of vasopressin or potentiate its renal effects may lead to normovolemic hyponatremia. Nitrous oxide, barbiturates, isoproterenol, and narcotics are drugs used during anesthesia and surgery that stimulate vasopressin release from the neurohypophysis and may contribute to impaired water excretion in the postoperative period. Anxiety, stress, and pain associated with surgical procedures also may contribute

68

ELECTROLYTE DISORDERS

to increased plasma vasopressin concentrations and result in decreased renal excretion of water. These events may result in postoperative hyponatremia, especially if the patient receives hypotonic fluids.6,38 In fact, routine use of hypotonic fluids in patients in whom water excretion is impaired by nonosmotic stimulation of vasopressin release is thought to be the main contributing factor to hospital-acquired hyponatremia in human patients and can be prevented by administration of isotonic fluids such as 0.9% NaCl.114 Chlorpropamide potentiates the action of vasopressin, possibly by inhibiting vasopressin-stimulated production of prostaglandin E2 or by up-regulating vasopressin receptors in the kidneys.35 Nonsteroidal anti-inflammatory drugs havea similar effect because oftheir inhibitionof prostaglandin production. The antineoplastic drugs vincristine and cyclophosphamide also impair water excretion. Figure 3-6 shows the effects of various drugs on the release and action of vasopressin.

CLINICAL SIGNS OF HYPONATREMIA The clinical signs of hyponatremia are related more to the rapidity of onset than to the severity of the associated plasma hypoosmolality. In human patients, deaths and severe complications of hyponatremia were most common when the serum sodium concentration acutely decreased to less than 120 mEq/L or at a rate greater than 0.5 mEq/L/hr.25 Cerebral edema and water intoxication occur if hyponatremia develops faster than the brain’s defense mechanisms can be called into play. Reduction in plasma osmolality and influx of water into the central nervous system cause the clinical signs observed in acute hyponatremia. A 30- to 35-mOsm/kg gradient can result in translocation of water between plasma and the brain in dogs.61 Clinical signs are often absent in chronic disorders characterized by slower decreases in serum sodium concentration and plasma osmolality. During hyponatremia of chronic onset, brain volume is adjusted toward normal by loss of potassium and organic osmolytes from cells.6,134 Acute water intoxication is likely only if the patient has some underlying cause of impaired water excretion at the time a water load occurs. For example, water-loaded dogs given repositol vasopressin developed signs of acute water intoxication.69 Early signs were mild lethargy, nausea, and slight weight gain; more severe signs included vomiting, coma, and a marked increase in body weight. One dog in this study died from pulmonary and cerebral edema. Weakness, incoordination, and seizures may also result from acute water intoxication. In one clinical report, a Labrador retriever developed acute hyponatremia (125 mEq/L) and severe neurologic signs (i.e., coma) after swimming for many hours in a lake.157 The dog spontaneously underwent marked diuresis and recovered with supportive care,

suggesting that it was capable of suppressing vasopressin release in response to the water load.

TREATMENT OF HYPONATREMIA The two main goals of treatment in hyponatremia are to diagnose and manage the underlying disease and, if necessary, increase serum sodium concentration and plasma osmolality. Severe, symptomatic hyponatremia of acute onset (48 hours’ duration) hyponatremia is corrected too rapidly (i.e., when the serum sodium concentration is increased by >10 to 12 mEq/L in 24 hours).95,151 When hyponatremia and hypoosmolality are corrected, potassium and organic osmolytes lost during adaptation must be restored to the cells of the brain. If replacement of these solutes does not keep pace with the increase in serum sodium concentration that occurs as a result of treatment, brain dehydration and injury— called osmotic demyelination or myelinolysis—may result.95,151 Experimental studies have confirmed that this syndrome is a result of a rapid and large increase in serum sodium concentration and is not a consequence of hyponatremia and hypoosmolality. Human patients with hyponatremia of more than 72 hours’ duration are more susceptible than those with hyponatremia of less than 24 hours’ duration.151 The neural lesions of myelinolysis develop several days after correction of hyponatremia and consist of myelin loss and injury to oligodendroglial cells in the pons and other sites in the brain (e.g., thalamus, subcortical white matter, and cerebellum). Lesions may take several days to develop, but on magnetic resonance imaging they are hyperintense on T2-weighted images, hypointense on T1-weighted images, and are not enhanced after gadolinium injection.95 The ability to reaccumulate organic osmolytes may vary among different regions of the brain and thus account for why some regions (e.g., midbrain) are more susceptible to osmotic demyelination.

Disorders of Sodium and Water: Hypernatremia and Hyponatremia Similar lesions have been reported in experimental dogs with hyponatremia with correction rates of 15 mEq/L/ day even without overcorrection to hypernatremia.94 In veterinary medicine, myelinolysis first was reported in two dogs after correction of hyponatremia associated with trichuriasis.122 In one dog, a serum sodium concentration of 101 mEq/L had been corrected to 136 mEq/L in less than 38 hours (correction rate, >22 mEq/L/day), and in the other, a serum sodium concentration of 108 mEq/L had been corrected to 134 mEq/L in less than 38 hours (correction rate, >16 mEq/L/day). Clinical signs developed 3 to 4 days after correction of hyponatremia and consisted of lethargy, weakness, and ataxia progressing to hypermetria and quadriparesis. Lesions were detected by magnetic resonance imaging and were located in the thalamus as compared with the more typical pontine location in affected human patients. From this experience, it was recommended that dogs with asymptomatic chronic hyponatremia be treated by mild water restriction and monitoring of serum sodium concentration. Symptomatic dogs with chronic hyponatremia should be treated conservatively at correction rates of less than 10 to 12 mEq/L/ day (0.5 mEq/L/hr). Serial monitoring of serum sodium concentration is necessary because the actual rate of correction may not correspond to the calculated rate of correction. Correction should be carried out with conventional crystalloid solutions (e.g., lactated Ringer’s solution and 0.9% NaCl) in a volume calculated specifically to replace the patient’s volume deficit. The clinician must remember that volume repletion in hypovolemic patients abolishes the nonosmotic stimulus for vasopressin release and allows the animal to excrete solute-free water via the kidneys. This in itself tends to correct the hyponatremia. Thus, caution should be exercised even when using conventional crystalloid fluid therapy. Three additional cases of suspected myelinolysis in dogs with chronic hyponatremia caused by hypoadrenocorticism or trichuriasis have been reported.11,24,105 The rates of correction of hyponatremia in these dogs were 22 mEq/L on day 1 and 17 mEq/L on day 2, 32 mEq/L over 2 days, and 17 mEq/L in 9 hours.11,24,105 The neurologic signs that developed (e.g., spastic tetraparesis, loss of postural and proprioceptive responses, dysphagia, trismus, and decreased menace response) were similar to those originally described by O’Brien.122 The dogs of these reports gradually recovered over several weeks. Water intake should be carefully restricted to a volume less than urine output in normovolemic patients with hyponatremia (e.g., psychogenic polydipsia), or drugs causing an antidiuretic effect should be discontinued if possible. Demeclocycline and lithium inhibit vasopressin release and have been used to treat SIADH in humans, but water restriction is probably the safest approach.48 In edematous patients, dietary sodium restriction and diuretic therapy should be considered. A 0.9% NaCl solution can be administered concurrently with loop diuretics

69

(e.g., furosemide) to effect more rapid correction of hyponatremia in overhydrated symptomatic patients. The occurrence of chronic hyponatremia in patients with congestive heart failure is often a sign of advanced disease and responds poorly to treatment. Administration of furosemide and an angiotensin-converting enzyme inhibitor (e.g., enalapril) may improve stroke volume and cardiac output by reducing preload and afterload and may decrease vasopressin secretion and enhance water excretion, which in turn may facilitate resolution of hyponatremia. Arginine vasopressin (AVP) receptor antagonists (vaptans) block either V2 receptors (lixivaptan, tolvaptan, satavaptan) or both V2 and V1A receptors (conivaptan).* Based on their mechanism of action, these drugs increase free water excretion by the kidneys and effectively normalize serum sodium concentration in patients with non-osmotic release of AVP causing euvolemic (e.g., SIADH) or hypervolemic (e.g., congestive heart failure, liver failure) hyponatremia.111,117 Patients with hypovolemic hyponatremia should be treated with an infusion of 0.9% NaCl or other isotonic fluid to replace their volume deficits. The AVP receptor antagonists increase water but not solute excretion via the kidneys, and likely will have major impact on the clinical management of euvolemic and hypervolemic hyponatremia in the near future.60 Conivaptan is administered as an intravenous bolus in 5% dextrose followed by a constant rate infusion, whereas tolvaptan, lixivaptan, and satavaptan are administered orally. In humans, adverse effects of the vaptans generally are limited to thirst and dry mouth. To minimize the risk of osmotic demyelination, correction of serum sodium concentration should be limited to 10 miles) duration.95 Corrected hyperchloremia also was observed in dogs during and after agility competition.85 Treatment of corrected hyperchloremia should be directed at correction of the underlying disease process. The effects of fluid therapy on chloride concentration should be anticipated, especially in patients with diabetes mellitus or abnormal renal function. Special attention should be given to plasma pH because patients with corrected hyperchloremia tend to be acidotic. Bicarbonate therapy can be instituted whenever plasma pH is less than 7.2 or bicarbonate concentration is less than 12 mEq/L in patients with hyperchloremic metabolic acidosis.

CONCLUSION Although it is the major anion in ECF, chloride has not received much attention in the clinical setting. It should be remembered that the chloride ion also is important in the metabolic regulation of acid-base balance. The kidneys regulate acid-base balance by changing the amount of chloride that is reabsorbed with sodium. Chloride is important in determining the patient’s SID, and therefore changes in chloride concentration will reflect the patient’s acid-base status. Corrected hypochloremia is associated with increased SID and metabolic alkalosis. Chloride is the only anion in ECF that can contribute to a substantial increase in SID. Administration of chloride is necessary for correction of hypochloremic metabolic alkalosis. Corrected hyperchloremia is associated with decreased SID and metabolic acidosis. Treatment with sodium bicarbonate should be carried out in hyperchloremic patients with a pH of less than 7.2.

REFERENCES 1. Abboud HE, Luke RG, Galla JH, et al. Stimulation of renin by acute selective chloride depletion in the rat. Circ Res 1979;44:815–21. 2. Adrogue´ HJ, Eknoyan G, Suki WK. Diabetic ketoacidosis: role of the kidneys in the acid-base homeostasis reevaluated. Kidney Int 1984;25:591–8. 3. Angle CT, Wakshlag JJ, Gillete RL, et al. Hematologic, serum biochemical, and cortisol changes associated with anticipation of exercise and short duration high-intensity exercise in sled dogs. Vet Clin Pathol 2009;38:370–4. 4. Atkins EL, Schwartz WB. Factors governing correction of the alkalosis associated with potassium deficiency: the critical role of chloride in the recovery process. J Clin Invest 1962;41:218–29. 5. Bernardini D, Gerardi G, Contiero B, et al. Interference of haemolysis and hyperproteinemia on sodium, potassium, and chloride measurements in canine serum samples. Vet Res Commun 2009;33(Suppl. 1):S173–26.

Disorders of Chloride: Hyperchloremia and Hypochloremia 6. Bia M, Thier SO. Mixed acid-base disturbances: a clinical approach. Med Clin North Am 1981;65:347–61. 7. Brown SA, Spyridakis LK, Crowell WA. Distal renal tubular acidosis and hepatic lipidosis in a cat. J Am Vet Med Assoc 1986;189:1350–2. 8. Burnell JM, Dawbron JK. Acid base parameters in potassium depletion in the dog. Am J Physiol 1970;218:1583–9. 9. Burnell JM, Teubner EJ, Simpson DP. Metabolic acidosis accompanying potassium deprivation. Am J Physiol 1974;227:329–33. 10. Calia CM, Hohenhaus AE, Fox PR, et al. Acute tumor lysis syndrome in a cat with lymphoma. J Vet Intern Med 1996;10:409–11. 11. Chew DJ, Leonard M, Muir III WW. Effect of sodium bicarbonate infusion on serum osmolality, electrolyte concentration, and blood gas tensions in cats. Am J Vet Res 1991;52:12–7. 12. Ching SV, Fettman MJ, Hamar DW, et al. The effect of chronic dietary acidification using ammonium chloride on acid-base and mineral metabolism in the adult cat. J Nutr 1989;111:902–15. 13. Christopher MM, Broussard JD, Peterson ME. Heinz body formation associated with ketoacidosis in cats. J Vet Intern Med 1995;9:24–31. 14. Constable PD. Hyperchloremic acidosis: the classic example of strong ion acidosis. Anesth Analg 2003;96:919–22. 15. Corrigan AM, Behrend EN, Martin LG, et al. Effect of glucocorticoid administration on serum aldosterone concentration in clinically normal dogs. Am J Vet Res 2010;71:649–54. 16. de Morais HSA. A nontraditional approach to acid-base disorders. In: DiBartola SP, editor. Fluid therapy in small animal practice. Philadelphia: WB Saunders; 1992. p. 297–320. 17. de Morais HSA. Chloride ion in small animal practice: the forgotten ion. J Vet Emerg Critic Care 1992;2:11–24. 18. de Morais HSA, Muir III WW. Strong ions and acid-base disorders. In: Bonagura JD, Kirk RW, editors. Current veterinary therapy XII. 12th ed. Philadelphia: WB Saunders; 1995. p. 121–7. 19. de Rouffignac C, Elalouf JM. Hormonal regulation of chloride transport in the proximal and distal nephron. Annu Rev Physiol 1988;50:123–40. 20. DiBartola SP, de Morais HSA. Case examples. In: DiBartola SP, editor. Fluid therapy in small animal practice. Philadelphia: WB Saunders; 1992. p. 599–688. 21. DiBartola SP, Green RA, de Morais HSA. Electrolyte and acid-base abnormalities. In: Willard MD, Tvedten H, Turnwald GH, editors. Small animal clinical diagnosis by laboratory methods. 2nd ed. Philadelphia: WB Saunders; 1994. p. 97–114. 22. DiBartola SP, Johnson SE, Davenport DJ, et al. Clinicopathologic findings resembling hypoadrenocorticism in dogs with primary gastrointestinal disease. J Am Vet Med Assoc 1985;187:60. 23. DiBartola SP, Leonard PO. Renal tubular acidosis in a dog. J Am Vet Med Assoc 1982;180:70–3. 24. Dobbins J. Gastrointestinal disorders. In: Arieff AI, DeFronzo RA, editors. Fluid, electrolyte, and acid-base disorders. New York: Churchill Livingstone; 1985. p. 827–49. 25. Dow SW, Fettman MJ, Smith KR, et al. Effects of dietary acidification and potassium depletion on acid-base balance, mineral metabolism and renal function in adult cats. J Nutr 1990;120:569–78.

89

26. Drazner FH. Distal renal tubular acidosis associated with chronic pyelonephritis in a cat. Calif Vet 1980;34:15–21. 27. Driscoll JL, Martin HF. Detection of bromism by an automated chloride method. Clin Chem 1966;12:314–8. 28. Elin RJ, Robertson EA. Bromide interference with determination of chloride by each of four methods (letter). Clin Chem 1981;27:778–9. 29. Emancipator K, Kroll MH. Bromide interference: is less really better? Clin Chem 1990;8:1470–3. 30. Fencl V, Rossing TH. Acid-base disorders in critical care medicine. Annu Rev Med 1989;40:17–29. 31. Finco DR, Barsanti JA, Brown SA. Ammonium chloride as an urinary acidifier in cats: efficacy, safety and rationale for its use. Mod Vet Pract 1986;67:537–41. 32. Fine A. Effects of carbonic anhydrase inhibition on renal ammoniagenesis in the dog. Pharmacology 1986;33: 217–20. 33. Gabow PA. Disorders associated with altered anion gap. Kidney Int 1985;27:472–83. 34. Galla JH. Metabolic alkalosis. J Am Soc Nephrol 2000;11:369–73. 35. Galla JH, Bonduris DN, Luke RG. Correction of chloride depletion metabolic alkalosis (CDA) without volume expansion (abstract). Clin Res 1982;30:540A. 36. Galla JH, Bonduris DN, Luke RG. Effect of hypochloremia in glomerular filtration rate (GFR) on euvolemic rats (abstract). Clin Res 1982;30:785A. 37. Galla JH, Bonduris DN, Luke RG. The correction of acute chloride-depletion alkalosis in the rat without volume expansion. Am J Physiol 1983;244:F217–F221. 38. Galla JH, Bonduris DN, Luke RG, et al. Effects of chloride and extracellular fluid volume on bicarbonate reabsorption along the nephron in metabolic alkalosis in the rat: reassessment of the classical hypothesis of the pathogenesis of metabolic alkalosis. J Clin Invest 1987;80:41–50. 39. Galla JH, Gifford JD, Luke RG, et al. Adaptations to chloride-depletion alkalosis. Am J Physiol 1991;261: R771–R781. 40. Galla JH, Luke RG. Chloride transport and disorders of acid-base balance. Annu Rev Physiol 1988;50:141–58. 41. Garella S, Cohen JJ, Northrup TE. Chloride-depletion metabolic alkalosis induces ECF volume depletion via internal fluid shifts in nephrectomized dogs. Eur J Clin Invest 1991;21:273–9. 42. Gennari FJ, Goldstein MB, Schwartz W. The nature of the renal adaptation to chronic hypocapnia. J Clin Invest 1972;51:1722–30. 43. Goodkin DA, Krishna GG, Narins RG. The role of the anion gap in detecting and managing mixed metabolic acid-base disorders. Clin Endocrinol Metab 1984;13:333–49. 44. Gougoux A, Vinay P, Zizian L, et al. Effect of acetazolamide on renal metabolism and ammoniagenesis in the dog. Kidney Int 1987;31:1279–90. 45. Graber ML, Quigg RJ, Slempsey WE, et al. Spurious hyperchloremia and decreased anion gap in hyperlipidemia. Ann Intern Med 1983;98:607–9. 46. Greger R. Chloride transport in thick ascending loop, distal convolution, and collecting duct. Annu Rev Physiol 1988;50:111–22. 47. Gunnerson KJ. Clinical review: the meaning of acid-base abnormalities in the intensive care unit part I — epidemiology. Crit Care 2005;9(5):508–16. 48. Halperin ML,Bun-ChenC.Influence ofacutehyponatremia on renal ammoniagenesis in dogs with chronic metabolic acidosis. Am J Physiol 1990;258:F328–F332.

90

ELECTROLYTE DISORDERS

49. Halperin ML, Vinay P, Gougoux A, et al. Regulation of the maximum rate of renal ammoniagenesis in the acidotic dog. Am J Physiol 1985;248:F607–F615. 50. Hara Y, Nezu Y, Harada Y, et al. Secondary chronic respiratory acidosis in a dog following the cervical cord compression by an intradural glioma. J Vet Med Sci 2002;64:863–6. 51. Haskins SC, Munger RJ, Helphrey MG, et al. Effects of acetazolamide on blood acid-base and electrolyte values in dogs. J Am Vet Med Assoc 1981;179:792–6. 52. Heird WC, Dell B, Driscoll JM, et al. Metabolic acidosis resulting from intravenous alimentation with synthetic amino acids. N Engl J Med 1972;287:943–5. 53. Hughes DE, Sokolowski J. Sodium chloride poisoning in the dog. Canine Pract 1978;5:28–31. 54. Hulter HN, Licht JH, Sebastian A. Kþ deprivation potentiates the renal acid excretory effect of mineralocorticoid: obliteration by amiloride. Am J Physiol 1979;236: F48–F57. 55. Jones NL. Blood gases and acid base physiology. 2nd ed. New York: Thieme Medical Publishers; 1987. 56. Kassirer JP, Berkman PM, Lawrenz DR, et al. The critical role of chloride in the correction of hypokalemic alkalosis in man. Am J Med 1965;38:172–89. 57. Koch SM, Taylor RW. Chloride ion in intensive care medicine. Crit Care Med 1992;20:227–40. 58. Kreisberg RA, Wood BC. Drugs and chemical-induced metabolic acidosis. Clin Endocrinol Metab 1983;12: 391–411. 59. Lemieux G, Gervais M. Acute chloride depletion alkalosis: effect of anions on its maintenance and correction. Am J Physiol 1964;207:1279–86. 60. Lemieux G, Lemieux C, Duplessis S, et al. Metabolic characteristics of cat kidney: failure to adapt to metabolic acidosis. Am J Physiol 1990;259:R277–R281. 61. Ling G, Stabenfeldt GH, Comer KM, et al. Canine hyperadrenocorticism: pretreatment clinical and laboratory evaluation of 117 cases. J Am Vet Med Assoc 1979;174:1211–5. 62. Luke RG, Galla JH. Chloride-depletion alkalosis with a normal extracellular fluid volume. Am J Physiol 1983;254:F419–F424. 63. MacCallum WG, Lintz J, Vermilye HN, et al. The effect of pyloric obstruction relation to gastric tetany. Bull John Hopkins Hosp 1920;31:1–7. 64. Madias NE, Bossed WH, Adrogue´ HJ. Ventilatory response to chronic metabolic acidosis and alkalosis in the dog. J Appl Physiol 1984;56:1640–6. 65. Madias NE, Wolf CJ, Cohen JJ. Regulation of acid-base equilibrium in chronic hypercapnia. Kidney Int 1985;27:538–43. 66. McKenzie EC, Jose-Cullineras E, Hinchcliff KW, et al. Serum chemistry alterations in Alaskan sled dogs during five successive days of prolonged endurance exercise. J Am Vet Med Assoc 2007;230:1486–92. 67. McLoughlin MA, Walshaw R, Thomas MW, et al. Gastric conduit urinary diversion in normal dogs. Part II. Hypochloremic metabolic alkalosis. Vet Surg 1992;21:33–9. 68. Moon PF, Kramer GC. Hypertonic saline-dextran resuscitation from hemorrhagic shock induces transient mixed acidosis. Crit Care Med 1995;23:323–31. 69. Morimatsu H, Rocktaschel J, Bellomo R, et al. Comparison of point-of-care versus central laboratory measurement of electrolyte concentrations on calculations of the anion gap and the strong ion difference. Anesthesiology 2003;98(5):1077–84.

70. Narins RG, Emmett M. Simple and mixed acid-base disorders: a practical approach. Medicine 1980;59:161–87. 71. Needle MA, Kaloyanides GJ, Schwartz WB. The effects of selective depletion of hydrochloric acid on acid-base and electrolyte equilibrium. J Clin Invest 1964;43:1836–46. 72. Norris SH, Kurtzman NA. Does chloride play an independent role in the pathogenesis of metabolic alkalosis? Semin Nephrol 1988;7:101–8. 73. Oh MS, Carrol HJ, Goldstein DA, et al. Hyperchloremic acidosis during the recovery phase of diabetic ketosis. Ann Intern Med 1978;89:925–7. 74. Penman RW, Luke RF, Jarboe TM. Respiratory effects of hypochloremic alkalosis and potassium depletion in the dog. J Appl Physiol 1972;33:170–4. 75. Peterson ME, Greco DS, Orth DN. Primary hypoadrenocorticism in ten cats. J Vet Intern Med 1989;3:55–8. 76. Phillips SF. Small and large intestinal disorders: associated fluid and electrolyte complications. In: Maxwell MH, Kleeman CR, Narins RG, editors. Clinical disorders of fluid and electrolyte metabolism. New York: McGrawHill; 1987. p. 865–77. 77. Ployngam T, Tobias AH, Smith SA, et al. Hemodynamic effects of methylprednisolone acetate administration in cats. Am J Vet Res 2006;67:583–7. 78. Polak A, Haynie GD, Hays RM, et al. Effects of chronic hypercapnia on electrolyte and acid-base equilibrium. I. Adaptation. J Clin Invest 1961;40:1223–37. 79. Polzin DJ, Stevens JB, Osborne CA. Clinical application of the anion gap in evaluation of acid-base disorders in dogs. Comp Cont Educ Pract Vet 1982;4:1021–33. 80. Reynolds BS, Concordet D, Germain CA, et al. Breed dependency of reference intervals for plasma biochemical values in cats. J Vet Intern Med 2010;24:809–18. 81. Rose BD. Clinical physiology of acid-base and electrolyte disorders. 3rd ed. New York: McGraw-Hill; 1989. 82. Rose RJ. Some physiological and biochemical effects of the intravenous administration of five different electrolyte solutions in the dog. J Vet Pharmacol Ther 1979;2:279–89. 83. Rose RJ, Caner J. Some physiological and biochemical effects of acetazolamide in the dog. J Vet Pharmacol Ther 1979;2:215–21. 84. Rosen RA, Bruce JA, Dubovsky EV, et al. On the mechanism by which chloride corrects metabolic alkalosis in man. Am J Med 1988;84:449–58. ˜ oz A, Benito M. Fluid and electrolyte shifts 85. Rovira S, Mun during and after agility competitions in dogs. J Vet Med Sci 2007;69:31–5. 86. Schild L, Giebisch G, Green R. Chloride transport in the proximal renal tubule. Annu Rev Physiol 1988;50:97–110. 87. Schwartz WB, Brackelt NC, Cohen JJ. The response of extracellular hydrogen ion concentration to graded degrees of chronic hypercapnia: the physiologic limits of defense of pH. J Clin Invest 1965;44:291–301. 88. Schwartz WB, Hays RM, Polak A, et al. Effects of chronic hypercapnia on electrolyte and acid-base equilibrium. II. Recovery, with special reference to the influence of chloride intake. J Clin Invest 1961;40:1238–49. 89. Senior DF, Sundstrom DA, Wolfson BB. Testing the effects of ammonium chloride and d-methionine on the urinary pH of cats. Vet Med 1986;81:88–93. 90. Sharkey L, Gjevre K, Hegstad-Davies R, et al. Breedassociated variability in serum biochemical analytes in four large-breed dogs. Vet Clin Pathol 2009;38:375–80. 91. Story DA, Morimatsu H, Egi M, et al. The effect of albumin concentration on plasma sodium and chloride measurements in critically ill patients. Anesth Analg 2007;104(4):893–7.

Disorders of Chloride: Hyperchloremia and Hypochloremia 92. Tietz NW, Pruden EL, Sigaard-Andersen O. Electrolytes, blood gases, and acid-base balance. Section one. Electrolytes. In: Tietz NW, editor. Textbook of clinical chemistry. Philadelphia: WB Saunders; 1986. p. 1172–91. 93. Toll PW, Gaehtgens P, Neuhaus D, et al. Fluid, electrolyte, and packed cell volume shifts in racing greyhounds. Am J Vet Res 1995;56:227–32. 94. van Ypersele de Strihou C, Gulyassy PF, Schwartz WB. Effects of chronic hypercapnia on electrolyte and acid-base equilibrium. III. Characteristics of the adaptive and recovery process as evaluated by provision of alkali. J Clin Invest 1962;41:2246–53. 95. Wakshlag J, Snedden K, Reynolds A. Biochemical and metabolic changes due to exercise in sprint-racing sled dogs: implications for postexercise carbohydrate supplements and hydration management. Vet Ther 2004;5:52–9. 96. Wall BM, Byrum GV, Galla JH, et al. Importance of chloride for the correction of chronic metabolic alkalosis in the rat. Am J Physiol 1987;253:F1031–F1039.

91

97. Warnock DG. Uremic acidosis. Kidney Int 1988;34:278–87. 98. Waters JH, Miller LR, Clack S, et al. Causes of metabolic acidosis in prolonged surgery. Crit Care Med 1999;27:2142–6. 99. Watson AD, Culvenor JA, Middleton DJ, et al. Distal renal tubular acidosis in a cat with pyelonephritis. Vet Rec 1986;119:65–8. 100. Whitehair KJ, Haskins SC, Whitehair JG, et al. Clinical applications of quantitative acid-base chemistry. J Vet Intern Med 1995;9:1–12. 101. Widmer B, Gerhardt RE, Harrington JT, et al. Serum electrolyte and acid base composition: the influence of graded degrees of chronic renal failure. Arch Intern Med 1979;139:1099–102. 102. Winter SD, Pearson JR, Gabow PA, et al. The fall of serum anion gap. Arch Intern Med 1990;150:311–13.

CHAPTER • 5

Disorders of Potassium: Hypokalemia and Hyperkalemia Stephen P. DiBartola and Helio Autran De Morais

92

and ECF, however, the serum potassium concentration can change without any change in the total body potassium content. One of the most important functions of potassium in the body is its role in generation of the normal resting cell membrane potential.

THE RESTING CELL MEMBRANE POTENTIAL The normal relationship between ECF and ICF potassium concentrations is maintained by sodium, potassiumadenosinetriphosphatase (Naþ, Kþ-ATPase) in cell membranes. This enzyme pumps sodium ions out of, and potassium ions into, the cell in a 3:2 Na/K ratio so that 50

Serum [K+] (mEq /L)

Potassium is the major intracellular cation in mammalian cells, whereas sodium is the major extracellular cation. Normally, the extracellular fluid (ECF) sodium concentration is approximately 140 mEq/L, and the ECF potassium concentration is approximately 4 mEq/L. This relationship is reversed in intracellular fluid (ICF), in which the sodium concentration is approximately 10 mEq/L and the potassium concentration is approximately 140 mEq/L. In experimental studies of dogs, control values for ICF sodium and potassium concentrations in skeletal muscle were 8.4 to 13.7 and 139 to 142 mEq/L, respectively.20,107 Total body potassium content in humans is approximately 50 to 55 mEq/kg body weight, and almost all of this potassium is readily exchangeable.6,70 In one study of potassium depletion in dogs, the control value for total exchangeable potassium as determined by 50K dilution was 47.1 mEq/kg body weight (range, 39.8 to 61.1 mEq/ kg).1 In cats, total body potassium is approximately 55 mEg/kg body weight.184a As much as 95% or more of total body potassium is located within cells, with muscle containing 60% to 75% of this potassium. Muscle potassium content in normal dogs and cats isapproximately400 mEq/ kg.20,107,147,191 As a solute, intracellular potassium is crucial for maintenance of normal cell volume. Intracellular potassium also is important for normal cell growth because it is required for the normal function of enzymes responsible for nucleic acid, glycogen, and protein synthesis. The remaining 5% of the body’s potassium is located in the ECF. Maintaining the ECF potassium concentration within narrow limits is critical to avoid the life-threatening effects of hyperkalemia on cardiac conduction. In humans, the serum potassium concentration is inversely correlated with the total body deficit of potassium (Fig. 5-1). Likewise, in dogs with potassium depletion induced by dietary restriction, the muscle potassium content was strongly correlated (r ¼ 0.87) with the serum potassium concentration.147 During translocation of potassium between ICF

40 r = .893

30

20 0

100

200 300 400 500 600 700 K+ deficit (mEq / 70 kg body weight)

800

Figure 5-1 Relationship of serum potassium concentration to bodily potassium deficit. The data are derived from seven metabolic balance studies carried out on 24 human subjects depleted of potassium. (From Sterns RH et al.: Medicine 60:339, 1981.)

Disorders of Potassium: Hypokalemia and Hyperkalemia

Em ¼ 61 log10

½K þ I ½K þ O

The Goldman-Hodgkin-Katz equation is a modification of the Nernst equation that allows prediction of Em based on the ionic permeability characteristics of the cell membrane to sodium and potassium and the concentrations of these ions inside and outside the cell: Em ¼ 61 log10

rPk½K þ I þ PNa ½Naþ I rPk½K þ O þ PNa ½Naþ O

where PNa and PK are the membrane permeabilities for sodium and potassium. The term r is included in the equation to account for the effect of the electrogenic Naþ, Kþ-ATPase pump under steady-state conditions. This term is assigned the Na/K transport ratio of 3:2 so that r ¼ 1.5. If the membrane permeability for potassium is assigned a value of 1.0 and the cell membrane is 100 times more permeable to potassium than sodium: Em ¼ 61 log10

1:5½K þ I þ 0:01½Naþ I 1:5½K þ O þ 0:01½Naþ O

For example, using the hypothetical ECF and ICF concentrations of sodium and potassium given at the beginning of this chapter: Em ¼ 61 log10

1:5½140 þ 0:01½10 1:5½4 þ 0:01½140

Em ¼ 61 log10 ð28:4Þ ¼ 89mV In one study of dogs with potassium deficiency, the predicted Em was 86.6 mV and the measured Em in

skeletal muscle of control animals was 90.1 mV.20 The resting cell membrane potential plays a vital role in the normal function of skeletal and cardiac muscle, nerves, and transporting epithelia.

THE THRESHOLD CELL MEMBRANE POTENTIAL The threshold cell membrane potential is reached when sodium permeability increases to the point that sodium entry exceeds potassium exit, depolarization becomes self-perpetuating, and an action potential develops. The ability of specialized cells to develop an action potential is crucial to normal cardiac conduction, muscle contraction, and nerve impulse transmission. The excitability of a tissue is determined by the difference between the resting and threshold potentials (the smaller the difference, the greater the excitability). Hypokalemia increases the resting potential (i.e., makes it more negative) and hyperpolarizes the cell, whereas hyperkalemia decreases the resting potential (i.e., makes it less negative) and initially makes the cell hyperexcitable (Fig. 5-2). If the resting potential decreases to less than the threshold potential, depolarization results, repolarization cannot occur, and the cell is no longer excitable. Translocation of potassium between body compartments results in a greater change in the ratio of intracellular to extracellular potassium concentrations ([Kþ]I/[Kþ]O) than does a change in total body potassium. In the former instance, the potassium concentrations of the two compartments change in opposite directions, whereas in the latter instance, they change in the same direction. Membrane excitability also is affected by ionized calcium concentration and acid-base balance. Calcium affects the threshold potential rather than the resting +30 Rights were not granted to include this content in electronic media. Action Please refer to the printed book. potential

0

Millivolts

the intracellular concentration of potassium is much higher than its extracellular concentration. As a result, Kþ ions diffuse out of the cell down their concentration gradient. However, the cell membrane is impermeable to most intracellular anions (e.g., proteins and organic phosphates). Therefore, a net negative charge develops within the cell as Kþ ions diffuse out, and a net positive charge accumulates outside the cell. Consequently, a potential difference is generated across the cell membrane. The principal extracellular cation is sodium, and it enters the cell relatively slowly down its concentration and electrical gradients, because the permeability of the cell membrane to potassium is 100-fold greater than its permeability to sodium. Diffusion of Kþ ions from the cell continues until the ECF acquires sufficient positive charge to prevent further diffusion of Kþ ions out of the cell. The ratio of the intracellular to extracellular concentrations of potassium ([Kþ]I/[Kþ]O) is the major determinant of the resting cell membrane potential as described by the Nernst equation:

93

–30

–60

–90

Normal threshold Resting

–120 Normal

Low K+ High K+

High Ca++ Low Ca++

Figure 5-2 Effects of serum calcium and potassium on membrane potentials of excitable tissues. The concentration of potassium in extracellular fluids affects the resting potential, whereas calcium concentrations alter the threshold potential. (From Leaf A, Cotran R. Renal pathophysiology. New York: Oxford University Press, 1976: 116.)

ELECTROLYTE DISORDERS

potential. Ionized hypocalcemia increases membrane excitability by allowing self-perpetuating sodium permeability to be reached with a lesser degree of depolarization, whereas ionized hypercalcemia requires greater than normal depolarization for this threshold to be reached (see Fig. 5-2). Thus, hypercalcemia counteracts hyperkalemia by normalizing the difference between the resting and threshold potentials, whereas hypocalcemia exacerbates the effect of hyperkalemia on membrane excitability. This principle is the basis for treating hyperkalemia with calcium salts (see the Treatment of Hyperkalemia section). Membrane excitability is increased by alkalemia and decreased by acidemia. As a result of these factors, clinical signs are not necessarily correlated with serum potassium concentrations. Electrocardiographic findings and muscle strength reflect the functional consequences of abnormalities in serum potassium concentration.

POTASSIUM BALANCE EXTERNAL POTASSIUM BALANCE External balance for potassium is maintained by matching output (primarily in urine) to input (from the diet). In the normal animal, potassium enters the body only through the gastrointestinal tract, and virtually all ingested potassium is absorbed in the stomach and small intestine. Transport of potassium in the small intestine is passive, whereas active transport (responsive to aldosterone) occurs in the colon. Colonic secretion of potassium may play an important role in extrarenal potassium homeostasis in some disease states (e.g., chronic renal failure) (Fig. 5-3). Potassium derived from the diet and endogenous cellular breakdown is removed from the body primarily by the kidneys and, to a much lesser extent, by the gastrointestinal tract. During zero balance, 90% to 95% of ingested potassium is excreted in urine, and the remaining 5% to 10% is excreted via the gastrointestinal tract. This pattern of output has been observed during control balance studies in normal dogs.12,139,152,169,182 In a study of renal handling of potassium in dogs, 90% to 98% of potassium intake was eliminated from the body by the kidneys.24 Adaptation occurs during chronic potassium loading so that the animal is protected from hyperkalemia that could occur as a result of an acute potassium load. This effect results from enhanced renal and colonic excretion of potassium, as well as from enhanced uptake of potassium by the liver and muscle, mediated by the effects of insulin and catecholamines. Potassium deprivation is associated with decreased aldosterone secretion, suppression of potassium secretion in the distal nephron, and increased reabsorption of potassium in the inner medullary collecting ducts. Skeletal muscle potassium

80

Percent of daily intake

94

Stool Potassium Excretion

60

40

20

0 0

5

10 15 20 25 40 60 Creatinine clearance (mL/ min)

80

100

Figure 5-3 Relationship between the degree of renal insufficiency and fecal potassium excretion. Data points are compiled from three studies comprising 98 balance periods in 40 human patients. Variation in dietary protein or sodium intake did not produce consistent changes in fecal potassium excretion; thus, data points from these balance periods were included without special designation. (From Alexander EA, Perrone RD. Regulation of extrarenal potassium metabolism. In: Maxwell MH, Kleeman CR, Narins RG, editors. Clinical disorders of fluid and electrolyte metabolism, 4th ed. New York: McGraw-Hill, 1987: 105–117, with permission of the McGraw-Hill Companies.)

concentration decreases, but brain and heart potassium concentrations are minimally affected during potassium depletion.20,107,178 The colon adapts to potassium deprivation by decreasing its secretion of potassium.

INTERNAL POTASSIUM BALANCE Internal balance for potassium is maintained by translocation of potassium between ECF and ICF. One half to two thirds of an acute potassium load appears in the urine within the first 4 to 6 hours, and effective translocation of potassium from ECF to ICF is crucial in preventing life-threatening hyperkalemia until the kidneys have sufficient time to excrete the remainder of the potassium load. Endogenous insulin secretion and stimulation of b2-adrenergic receptors by epinephrine promote cellular uptake of potassium in the liver and muscle by increasing the activity of cell membrane Naþ, Kþ-ATPase. The main effects of these hormones are to facilitate distribution of an acute potassium load and not to mediate minor adjustments in serum potassium concentration. The ECF concentration of potassium itself plays an important role in translocation because potassium movement into cells is facilitated by the change in chemical concentration gradient resulting from addition of potassium to ECF. The fraction of an acute potassium load taken up by the body is increased during chronic potassium depletion and decreased when total body potassium is excessive. In summary, any change in serum potassium concentration must arise from a change in intake, distribution, or excretion (Fig. 5-4).

Disorders of Potassium: Hypokalemia and Hyperkalemia

K+ intake (diet, parenteral fluids)

Translocation

ECF

[K+]

ICF [K+]

K+ excretion

Colon (feces) 10%

Kidneys (urine) 90%

Figure 5-4 Components of potassium homeostasis. ECF, Extracellular fluid; ICF, intracellular fluid. (Drawing by Tim Vojt.)

EFFECT OF ACID-BASE BALANCE ON POTASSIUM DISTRIBUTION The effect of acute pH changes on translocation of potassium between ICF and ECF is complex. In general, acidosis is associated with movement of potassium ions from ICF to ECF, and alkalosis is associated with movement of potassium ions from ECF to ICF. Early animal studies and observations in a small number of human patients led to the prediction that acute metabolic acidosis would be associated with a 0.6-mEq/L increment in serum potassium concentration for each 0.1-U decrement in pH. This rule of thumb has circulated widely among clinicians.31,181,188 However, a critical review of experimental studies in animals and humans demonstrated that changes in serum potassium concentration during acute acid-base disturbances were quite variable.4 The change in serum potassium concentration was greatest during acute mineral acidosis. In dogs, the increase in serum potassium concentration after administration of a mineral acid (e.g., HCl or NH4Cl) was very variable, ranging from a 0.17- to 1.67-mEq/L increment in serum potassium concentration per 0.1-U decrement in pH (mean, 0.75 mEq/L). The increment in serum potassium concentration during acute respiratory acidosis in dogs was much lower, averaging only 0.14 mEq/L per 0.1-U decrement in pH. The decrement in serum potassium concentration during metabolic alkalosis in dogs averaged 0.18 mEq/L per 0.1-U increment in pH, whereas it averaged 0.27 mEq/L per 0.1-U increment in pH during respiratory alkalosis. In another study, respiratory alkalosis induced by hyperventilation in anesthetized dogs caused a somewhat greater decrement in serum

95

potassium concentration (0.4 mEq/L) for each 0.1-U increment in pH.136 An increase in serum potassium concentration did not occur in acute metabolic acidosis caused by organic acids (e.g., lactic acid and ketoacids).* Acute infusion of b-hydroxybutyric acid in normal dogs caused an increase in insulin in portal venous blood and hypokalemia, presumably as a result of potassium uptake by cells. Conversely, acute infusion of HCl led to increased portal vein glucagon concentration and hyperkalemia, possibly caused by potassium release from cells.5 In summary, only mineral acidosis is expected to cause any clinically relevant change in serum potassium concentration during acute acid-base disturbances. Many factors probably contribute to the variable changes observed in serum potassium concentration during acute acid-base disturbances, including blood pH and HCO3 concentration, nature of the acid anion (mineral versus organic), osmolality, hormonal activity (e.g., catecholamines, insulin, glucagon, and aldosterone), and the metabolic and excretory roles of the liver and kidneys.4 Hyperosmolality and lack of insulin are more likely to be responsible for hyperkalemia observed in patients with diabetic ketoacidosis than is the acidosis itself. Hyperkalemia associated with acute metabolic acidosis induced by mineral acids is transient. In a study of acute and chronic metabolic acidosis induced in dogs by administration of HCl or NH4Cl, hyperkalemia was observed after acute infusion of HCl, but hypokalemia developed after 3 to 5 days of NH4Cl administration.122 The observed hypokalemia was associated with inappropriately high urinary excretion of potassium and increased plasma aldosterone concentration.122 Similar findings have been reported in rats with chronic metabolic acidosis induced by NH4Cl. Despite a total body deficit of potassium, rats with chronic metabolic acidosis did not conserve potassium appropriately.170 This effect may be caused by a decreased filtered load of HCO3, increased distal delivery of sodium, and increased distal tubular flow. Thus, metabolic acidosis of at least 2 to 3 days’ duration is associated with increased urinary potassium excretion and mild hypokalemia rather than hyperkalemia.79

RENAL HANDLING OF POTASSIUM The kidneys are the primary regulators of potassium balance. Potassium is filtered at the glomerulus, and approximately 70% of the filtered load is reabsorbed isosmotically with water and sodium in the proximal tubule. An additional 10% to 20% of filtered potassium is reabsorbed in the ascending limb of Henle’s loop. Finally, 10% to 20% of the filtered load is delivered to the distal nephron, where final adjustments in potassium

*References 4, 5, 98, 142, 143, 195.

96

ELECTROLYTE DISORDERS

reabsorption and secretion are made. Potassium experiences either net reabsorption or secretion in the connecting tubule, cortical collecting duct, and first portion of the outer medullary collecting duct, depending on the body’s needs. Net movement of potassium in these segments of the nephron determines urinary excretion of potassium. Potassium once again experiences reabsorption in the last portion of the outer medullary collecting duct and inner medullary collecting duct regardless of the body’s needs.

MECHANISMS OF RENAL TUBULAR TRANSPORT OF POTASSIUM The transepithelial electrical potential difference is lumen negative in the early proximal tubule, but no active transport mechanism for potassium has been discovered in this segment of the nephron. In the proximal tubule, potassium is reabsorbed along with water by solvent drag via the paracellular route. Apparently, water reabsorption increases the luminal concentration of potassium enough to overcome the unfavorable transepithelial potential difference. The transepithelial electrical potential difference becomes lumen positive in the late proximal tubule, and this facilitates reabsorption of potassium by the paracellular route. Transcellular transport of potassium in the proximal tubular cells occurs by means of potassium channels in both luminal and basolateral membranes and by a Kþ-Cl cotransporter in basolateral membranes (Fig. 5-5). In the thick ascending limb of the Henle loop, the transepithelial electrical potential difference is strongly lumen positive, and most potassium reabsorption occurs by the paracellular route. Potassium channels in the luminal membranes allow potassium to exit the cell down its concentration gradient and facilitate the electrochemical

Tubular fluid

Interstitial fluid

Cell

TEPD Lumen negative (early in proximal tubule)

gradient for potassium reabsorption via the paracellular route. Transcellular reabsorption of potassium is facilitated by the luminal Naþ-Kþ-2Cl cotransporter and by potassium channels and a Kþ-Cl cotransporter in the basolateral membranes (Fig. 5-6). The mechanisms of renal potassium handling in the distal convoluted tubule are shown in Figure 5-7. The thiazide-sensitive Naþ-Cl cotransporter and the KþCl cotransporter in the luminal membranes of these tubular cells result in secretion of potassium and reabsorption of sodium while chloride is recycled across the luminal membrane. The basolateral Naþ, Kþ-ATPase Tubular fluid

Impermeable to water

ATP

Inhibited by loop diuretics

+

K+

K

ATP

2K+

+

K

K+ Cl–

K+

Cl–

K+ TEPD Lumen positive

Figure 5-6 Renal tubular transport mechanisms for potassium in the thick ascending limb of Henle’s loop. TEPD, Transepithelial potential difference. (Drawing by Tim Vojt.)

Tubular fluid

Interstitial fluid

Cell

Na+ 2K+

3Na+

1Na+ 2Cl– 1K+

Inhibited by thiazide diuretics 3Na+

Interstitial fluid

Cell

3Na+

ATP

2K+

–

Cl

K+ K+ Cl– Cl–

Cl– K+

Cl–

K+

TEPD Lumen positive (majority of proximal tubule)

Figure 5-5 Renal tubular transport mechanisms for potassium in the proximal tubule. TEPD, Transepithelial potential difference. (Drawing by Tim Vojt.)

TEPD Lumen negative

Figure 5-7 Renal tubular transport mechanisms for potassium in the distal convoluted tubule (early distal tubule). TEPD, Transepithelial potential difference. (Drawing by Tim Vojt.)

Disorders of Potassium: Hypokalemia and Hyperkalemia maintains a low intracellular concentration of sodium and a high intracellular concentration of potassium that facilitate sodium reabsorption and potassium secretion across the luminal membranes. Principal cells are found in the connecting tubule and collecting duct and are responsible for potassium secretion. The basolateral membranes of principal cells are rich in Naþ, Kþ-ATPase, which maintains a high intracellular potassium concentration. The luminal membranes of the principal cells contain an electrogenic sodium channel (ENaC). This sodium channel is directly blocked by the diuretics amiloride and triamterene, whereas spironolactone antagonizes the effect of aldosterone on the channel. Sodium movement through this channel renders the tubular lumen negative, and the resultant increase in lumen electronegativity facilitates secretion of Kþ ions through luminal Kþ channels (Fig. 5-8). There are two types of intercalated cells in the distal nephron. Type A or a intercalated cells contain HþATPase and Hþ, Kþ-ATPase in their luminal membranes and Cl- HCO3 countertransporters and Cl and Kþ channels in their basolateral membranes. They also contain carbonic anhydrase. This arrangement allows the intercalated cell to secrete Hþ ions and reabsorb Kþ and HCO3 ions. Potassium is actively transported across the luminal membranes of type a intercalated cells by Hþ, Kþ-ATPase and then diffuses down its concentration gradient through potassium channels in the basolateral membranes (Fig. 5-9). Type A and a intercalated cells are found in the connecting tubule, cortical collecting duct, and outer medullary collecting duct. Type B and b intercalated cells are found only in the cortical collecting ducts and secrete HCO3 ions. They are able to do so because their polarity is reversed as compared

with type a intercalated cells (i.e., the Hþ-ATPase is in the basolateral membrane, and the Cl- HCO3 countertransporter is in the luminal membrane). Potassium is reabsorbed from the last portion of the outer medullary collecting duct and throughout the inner medullary collecting duct. In these segments of the nephron, potassium is reabsorbed by the paracellular route despite a lumen-negative transepithelial potential difference, because reabsorption of water increases the chemical concentration gradient sufficiently to overcome the unfavorable electrical gradient.

DETERMINANTS OF URINARY POTASSIUM EXCRETION Three main factors affect potassium secretion in the distal nephron: the magnitude of the chemical concentration gradient for potassium between the tubular cells and tubular lumen, the tubular flow rate, and the transmembrane potential difference across the luminal membranes of the tubular cells. Gastrointestinal absorption of a potassium load increases the ECF concentration of potassium. This results in an increase in the number of Kþ ions available for uptake at the basolateral membranes of the distal tubular cells by Naþ, Kþ-ATPase, and the resulting increase in intracellular potassium concentration increases the chemical concentration gradient for diffusion of Kþ ions out of the tubular cells across their luminal membranes. Aldosterone is the most important hormone affecting urinary potassium excretion. Its secretion by the zona glomerulosa of the adrenal gland is stimulated directly by hyperkalemia and angiotensin II (produced in response to volume depletion), whereas adrenocorticotropic hormone (ACTH), hyponatremia, and decreased a Intercalated cell

Tubular fluid

3Na+

Na+ ATP

+

3Na

Inhibited by K+-sparing diuretics

ATP

2K+

K+ K+

TEPD Lumen negative

H

+

+

ATP

2K+

K H+ + OH–

TEPD Lumen negative

Interstitial fluid

ATP

+

H+

CA

Tubular fluid

Interstitial Principal Cell fluid

97

HOH

+CO2 HCO3–

K

K+

Cl–

Cl–

K+

Figure 5-8 Renal tubular transport mechanisms for potassium in the principal cells of the late distal tubule and collecting duct. TEPD, Transepithelial potential difference. (Drawing by Tim Vojt.)

Figure 5-9 Renal tubular transport mechanisms for potassium in the a intercalated cells of the late distal tubule and collecting duct. CA, Carbonic anhydrase; TEPD, Transepithelial potential difference. (Drawing by Tim Vojt.)

98

ELECTROLYTE DISORDERS

extracellular pH play permissive roles in promoting aldosterone secretion. Aldosterone release is inhibited by dopamine and atrial natriuretic factor, both of which are released in response to volume expansion. Aldosterone increases reabsorption of Naþ and secretion of Kþ and Hþ ions in the distal nephron. Its primary effect is to increase the number of open Naþ channels in the luminal membranes of the principal cells. Sodium reabsorption via these luminal Naþ channels is electrogenic (i.e., it generates electronegativity in the tubular lumen). This electronegativity can be dissipated either by Kþ or Hþ ion secretion or by Cl reabsorption in the distal nephron. Aldosterone increases the activity and number of Naþ, Kþ-ATPase pumps in the basolateral membranes of the principal cells, and this effect may occur as a result of increased entry of Naþ ions across the luminal membranes. Increased Naþ, Kþ-ATPase activity in turn increases the intracellular Kþ concentration and facilitates Kþ secretion across the luminal membranes. Aldosterone also increases the number of open Kþ channels in the luminal membrane, thus facilitating Kþ exit into tubular fluid. Aldosterone can influence Hþ secretion in two ways. It directly promotes Hþ ion secretion in Hþ-secreting type a intercalated cells by stimulation of the Hþ-ATPase present in their luminal membranes. Aldosterone also promotes Hþ secretion in the distal tubule by stimulating electrogenic Naþ reabsorption in principal cells and increasing lumen electronegativity, which favors enhanced Hþ secretion. An increase in distal tubular flow enhances potassium secretion by rapidly moving secreted Kþ ions downstream and providing new tubular fluid from upstream in the nephron. This allows maintenance of a high chemical concentration gradient for potassium secretion and provides a “sink” for movement of Kþ ions into tubular fluid. A decrease in distal tubular flow has the opposite effect and promotes dissipation of the chemical gradient for diffusion of Kþ ions from principal cells into tubular fluid. Lumen electronegativity is generated by sodium reabsorption through Naþ channels in the luminal membranes of principal cells. Normally, some of this electronegativity is dissipated by passive Cl reabsorption. If a large concentration of a relatively nonresorbable anion (e.g., SO42, HCO3, penicillin) is present in distal tubular fluid, less dissipation of the electronegativity occurs, and Kþ secretion is enhanced. This factor contributes to the pathophysiology of metabolic alkalosis. In this setting, there is less Cland more HCO3 in the distal tubular fluid, and HCO3 is relatively nonresorbable in the cortical collecting duct. This is one reason metabolic alkalosis promotes urinary Kþ excretion. Amiloride is a diuretic that impairs luminal Naþ entry into principal cells by decreasing the number of open Naþ channels. This in turn reduces lumen electronegativity and impairs Kþ secretion. Thus, the magnitude of distal tubular lumen electronegativity has an important effect on urinary Kþ excretion.

Tubular fluid

Interstitial fluid

Cell

Slow Na+ 3Na+

ATP

2K+

K+

Luminal flow Cl–

–

Negative

+

Positive

Fast

Figure 5-10 Factors affecting urinary excretion of potassium. (Drawing by Tim Vojt.)

The Naþ and Cl concentrations of distal tubular fluid usually have little effect on Kþ secretion. When the luminal Naþ concentration is very low ( dogs) • Diet-induced hypokalemic nephropathy (cats)

No Hyperchloremic metabolic acidosis? Yes

Renal tubular acidosis

No Mineralocorticoid excess (hyperaldosteronism)

Figure 5-11 Algorithm for the clinical approach to hypokalemia. (Drawing by Tim Vojt.)

BOX 5-1

Causes of Hypokalemia

Decreased Intake Alone unlikely to cause hypokalemia unless diet is aberrant Administration of potassium-free (e.g., 0.9% NaCl, 5% dextrose in water) or deficient fluids (e.g., lactated Ringer’s solution over several days) Bentonite clay ingestion (e.g., cat litter)

Translocation (ECF ! ICF) Alkalemia Insulin/glucose-containing fluids Catecholamines Hypothermia Hypokalemic periodic paralysis (Burmese cats) Albuterol overdosage

Increased Loss Gastrointestinal (FEK 4%-6%) Chronic renal failure in cats Diet-induced hypokalemic nephropathy in cats Distal (type I) renal tubular acidosis (RTA) Proximal (type II) RTA after NaHCO3 treatment Postobstructive diuresis Dialysis Mineralocorticoid excess Hyperadrenocorticism Primary hyperaldosteronism (adenoma, adenocarcinoma, hyperplasia) Drugs Loop diuretics (e.g., furosemide, ethacrynic acid) Thiazide diuretics (e.g., chlorothiazide, hydrochlorothiazide) Amphotericin B Penicillins Unknown mechanism Rattlesnake envenomation

Effects of Potassium Depletion on Acid-Base Balance Hypokalemia often is said to be associated with metabolic alkalosis, but early studies used diuretics or mineralocorticoids to induce potassium depletion. These methods

Disorders of Potassium: Hypokalemia and Hyperkalemia probably caused disproportionate urinary loss of chloride relative to the chloride concentration of ECF, and chloride depletion presumably was the major factor responsible for the development of metabolic alkalosis (see Chapter 10). Pure potassium depletion apparently does cause metabolic alkalosis in rats, but in dogs it leads to metabolic acidosis.20,32,77 When potassium depletion was produced during a 2- to 4-week period in dogs, and care was taken to prevent chloride depletion, metabolic acidosis developed.32,77 When potassium was restored to the diet, metabolic acidosis resolved within 5 days. The observed reduction in net acid excretion and metabolic acidosis that accompany dietary potassium depletion in the dog appear to be caused by a distal renal tubular acidification defect, which is promptly reversed by potassium repletion.77 This acidification defect is at least partially related to decreased aldosterone secretion.97 Chronic potassium depletion also appears to lead to metabolic acidosis in cats. Adult cats were fed a potassium-restricted (0.2% potassium), 32% protein diet with or without 0.8% NH4Cl.61 Serum potassium concentrations decreased from 4.3 to 4.5 mEq/L to 3.1 to 3.5 mEq/L in the NH4Cl-treated cats and to 3.6 to 3.8 mEq/L in the cats not receiving NH4Cl. Urinary FEK was appropriately decreased to 3% to 6% in both groups of cats. Potassium balance was decreased in both groups but became negative only in the NH4Cl-treated cats. Metabolic acidosis developed in both groups but was more severe in cats treated with NH4Cl. Metabolic acidosis resolved in both groups during potassium repletion.

Effects on Muscle Muscle weakness develops when serum potassium concentration decreases to less than 3.0 mEq/L, increased creatine kinase concentration develops when serum potassium concentration decreases to less than 2.5 mEq/L, and frank rhabdomyolysis may occur when serum potassium concentration decreases to less than 2.0 mEq/L.106 Rear limb weakness may be observed in dogs and cats with hypokalemia. In cats, weakness of the neck muscles with ventroflexion of the head is commonly observed.57,59,180 Forelimb hypermetria and a broad-based hind limb stance also may be observed in hypokalemic cats. Respiratory muscle paralysis required ventilatory support in two cats with potassium depletion and was thought to be the cause of death in an experimental study of potassium depletion in dogs.59,147 Acute onset of hypokalemia and muscular weakness also have been reported in hyperthyroid cats.140 Three of the four cats in this study received fluid therapy with lactated Ringer’s solution and were treated by surgical thyroidectomy, but one cat developed hypokalemia before treatment. Serum potassium concentration was less than normal in only 5% of hyperthyroid cats in an early study,149 but in a recent study hypokalemia was

103

present in almost 30% of hyperthyroid cats before treatment by thyroidectomy.137 The effects of progressive potassium depletion on skeletal muscle were studied in dogs and rats.20 In both species, a progressive increase in ICF sodium concentration and a progressive decrease in ICF potassium concentration were observed during potassium deficiency. In rats, hyperpolarization of the cell membrane (as predicted by the Goldman-Hodgkin-Katz equation) was detected by direct measurement at all stages of potassium depletion. In dogs, there was an initial hyperpolarization of the cell membrane (mean measured Em, 92.4 mV) during moderate potassium deficiency because [Kþ]O decreased proportionately more than [Kþ]I. There was a dramatic decrease in Em (mean measured value, 54.8 mV) at the onset of muscle weakness and paralysis in dogs with severe potassium deficiency (serum potassium concentration, 1.6 mEq/L). In rats with potassium deficiency, predicted and measured Em values were similar during both moderate and severe potassium deficiencies, and paralysis was not observed. The inability to predict resting Em in dogs with severe potassium depletion could be explained by an increase in the sodium permeability of the muscle cell membrane. This study also demonstrated the development of metabolic acidosis in dogs (pH, 7.29; HCO3, 17.0 mEq/L) and metabolic alkalosis (pH, 7.54; HCO3, 37.0 mEq/L) in rats with severe potassium deficiency. Potassium is released from muscle cells during exercise, causing Vasodilatation and increased blood flow.106 This release of cellular potassium is impaired in states of potassium depletion, resulting in muscle ischemia. Muscle blood flow and potassium release increased markedly during exercise in normal but not in potassium-depleted dogs (serum potassium concentration, 2.3 mEq/L), and exercise caused rhabdomyolysis characterized by focal necrosis and inflammatory cell infiltration in potassiumIncreased creatine kinase depleted dogs.107 concentrations and electromyographic abnormalities have been observed in cats with hypokalemic polymyopathy, but histopathologic lesions usually are mild or absent.59,180 In dogs with experimentally induced potassium depletion, electromyographic changes were not observed, and increased serum creatine kinase concentration and muscle histopathology were observed only in dogs that had experienced extremely rapid potassium depletion induced by administration of desoxycorticosterone acetate in addition to a potassium-deficient diet.147 Intestinal ileus has been described in human patients with potassium depletion but usually is not recognized clinically in dogs and cats.

Effects on the Cardiovascular System Electrocardiographic changes and cardiac arrhythmias may develop, because hypokalemia delays ventricular repolarization, increases the duration of the action

104

ELECTROLYTE DISORDERS

potential, and increases automaticity. The electrocardiographic changes associated with hypokalemia in human patients (e.g., decreased amplitude T waves, ST segment depression, and U waves) are not consistently observed in dogs and cats, but supraventricular and ventricular arrhythmias may occur. Prolongation of the QT–interval and U waves have been reported in a dog with severe hypokalemia (2.0 mEq/L) caused by chronic vomiting and in dogs with experimentally induced potassium depletion (serum potassium concentration, 2.2 mEq/L).14,86,91 In another study, development of hypokalemia in dogs over a 5-day period was associated with ST segment deviations, decreased amplitude T waves, and the appearance of U waves.68 The appearance of T waves in normal dogs is variable (e.g., positive, negative, and biphasic), and interpretation of the effects of hypokalemia on ventricular repolarization is difficult unless a baseline electrocardiogram has been obtained previously. Hypokalemia potentiates the toxic effects of digitalis on cardiac conduction and may potentiate premature contractions. Hypokalemia also renders the myocardium refractory to the effects of class I antiarrhythmic agents (e.g., lidocaine, quinidine, procainamide). Therefore, serum potassium concentration should be measured and hypokalemia should be corrected in dogs with ventricular arrhythmias unresponsive to antiarrhythmic therapy.

Effects on the Kidneys Potassium depletion produces functional and morphologic abnormalities in the kidneys, referred to as hypokalemic nephropathy. Renal vasoconstriction leads to decreases in renal blood flow and glomerular filtration rate (GFR). Polyuria and polydipsia are observed in potassium depletion and result from impaired responsiveness of the kidneys to ADH. Defective collecting duct responsiveness to ADH is associated with decreased medullary tonicity, increased medullary blood flow, and impaired cyclic adenosine monophosphate (cAMP) generation in response to ADH. The urinary concentrating defect in potassium depletion results from decreased expression of ADH-regulated aquaporin2 water channels in the luminal membranes of the renal epithelial cells of the cortical and medullary collecting ducts.10,125 In one study, potassium depletion in dogs over an average of 51 days led to a decrease in total exchangeable potassium from 47.1 to 35.3 mEq/kg and a decrease in serum potassium concentration from more than 4.0 mEq/L to approximately 2.5 mEq/L.1 These dogs experienced decreases in GFR, renal blood flow, and urinary concentrating capacity (UOsm after 20 hours of water deprivation) of approximately 25%. In another study, potassium depletion (serum potassium concentration, 2.1 mEq/L) in dogs had little effect on GFR but caused a 45% reduction in maximal UOsm (1902 to 1055 mOsm/kg).19 In a clinical report, a dog with chronic

vomiting and hypokalemia (2.0 mEq/L) developed polyuria, polydipsia, and a urinary concentrating defect that persisted after correction of hypokalemia.86 These abnormalities were attributed to medullary washout of solute and were corrected by partial water restriction and dietary supplementation with NaCl and KCl. In yet another study, dogs subjected to potassium depletion (serum potassium concentration, 2.9 mEq/L) experienced a doubling of urine volume (596 to 1202 mL per 24 hours) and a 40% reduction in maximum urine osmolality (2006 to 1187 mOsm/kg).167 Potassium depletion increases renal ammoniagenesis and urinary net acid excretion, whereas potassium loading tends to have the opposite effect.189 In the rat, increased ammoniagenesis during potassium depletion occurs primarily via enhanced phosphate-dependent glutaminase activity and increased mitochondrial ammoniagenesis in the proximal tubular cells of the renal cortex. The decrease in ammoniagenesis during potassium loading may occur in renal tubular cells from the outer medullary region. Many experimental studies on potassium depletion and renal regulation of acid-base balance have been performed in rats. The renal response of the dog to acute acidosis is known to differ somewhat from that of the rat, and care must be taken in extrapolating data about the renal response to potassium depletion in the rat to dogs.190 Proximal renal tubular sodium reabsorption is increased during potassium depletion, possibly as a result of an increase in the activity of the proximal Naþ-Hþ antiporter. However, distal sodium reabsorption is decreased during potassium depletion. This presumably occurs as a result of decreased aldosterone secretion and is a direct effect of decreased ECF potassium concentration on the zona glomerulosa of the adrenal glands. Decreased distal sodium reabsorption decreases Kþ and Hþ ion secretion by decreasing luminal electronegativity. This decreases potassium loss in the urine but also tends to impair renal acid excretion. Thus, increased renal ammoniagenesis during potassium depletion may represent a mechanism for enhancing urinary excretion of fixed acid (as NH4þ) at a time when distal Hþ ion secretion is impaired. Consequently, derangements in acid-base balance are minimized. The cytoplasmic and mitochondrial enzyme activity profile of renal tubular cells during potassium depletion is strikingly similar to that observed during chronic metabolic acidosis.189 This similarity suggests the possibility of a common effector mechanism for stimulation of renal ammoniagenesis. Intracellular pH would be a logical candidate for such an effector. As Kþ ions leave cells to maintain ECF potassium concentration during potassium depletion, Hþ ions enter cells and presumably lower intracellular pH. Reduced intracellular pH may in turn be the signal for increased renal ammoniagenesis from glutamine. Some studies have demonstrated reduced

Disorders of Potassium: Hypokalemia and Hyperkalemia intracellular pH in renal tubular cells during potassium depletion, whereas others have found no change.2,177 Increased ammonia concentrations may activate the third component of complement (C3) and contribute to development of chronic tubulointerstitial disease by recruitment of immune cells.138,196 Vacuolization of proximal tubular cells is observed in human patients, whereas similar lesions are observed in the distal nephron, mainly in the medullary collecting ducts, in potassiumdepleted rats. Vacuolization of proximal tubular epithelial cells has also been reported in potassium-depleted dogs.1

SPECIFIC CAUSES OF HYPOKALEMIA IN DOGS AND CATS Hypokalemia arises from decreased intake, translocation of potassium from ECF to ICF, and excessive loss of potassium by either the gastrointestinal or urinary route. Decreased intake of potassium alone is unlikely to cause hypokalemia, but it may be a contributing factor. In chronically ill animals, for example, prolonged anorexia, loss of muscle mass, and ongoing urinary potassium losses probably combine to cause hypokalemia. A specific cause for mild hypokalemia in hospitalized dogs and cats often cannot be identified. Such hypokalemia may resolve with successful treatment of the primary disease process. Iatrogenic hypokalemia may develop when potassiumdeficient fluids are administered to anorexic patients in a hospital setting. For example, lactated Ringer’s solution (potassium concentration, 4 mEq/L) is a replacement solution and does not provide sufficient potassium for maintenance needs in most animals. Solutions used for maintenance fluid therapy should contain 15 to 30 mEq/L potassium (see Chapter 14). Ingestion of certain types of clay has been associated with hypokalemia in humans because the clay can bind potassium in the gastrointestinal tract and impair its absorption, and hypokalemia has been reported in a cat after ingestion of clay cat litter containing bentonite.18,84,96 Translocation of potassium into cells may occur with alkalemia, insulin release, and catecholamine release. Alkalemia contributes to hypokalemia as Kþ ions enter cells in exchange for Hþ ions. Insulin promotes uptake of glucose and potassium by hepatic and skeletal muscle cells and may contribute to hypokalemia when glucose-containing fluids are administered. The stress of illness and the associated epinephrine release may also contribute to hypokalemia. Severe hypokalemia has been reported in dogs that have ingested the b2-adrenergic agonist albuterol.129,200 The mechanism of hypokalemia was presumably rapid uptake of extracellular potassium by liver and muscle cells. Hypokalemia has been associated with hypothermia, possibly as a result of potassium entry into cells.164 Mild hypokalemia was reported in 78% of dogs suffering from rattlesnake envenomation.28 Affected dogs also had transient echinocytosis that was not consistently associated with the observed hypokalemia.

105

A syndrome characterized by recurrent episodes of limb muscle weakness, neck ventroflexion, increased creatine kinase concentrations, and hypokalemia has been reported in related Burmese cats 4 to 12 months of age.22,103,112,127,128 This syndrome may represent an animal model of hypokalemic periodic paralysis in humans, a familial disorder characterized by episodes of sudden translocation of potassium from ECF to ICF. Gastrointestinal loss of potassium (e.g., vomiting of stomach contents) is an important cause of hypokalemia in small animals. In one study, hypokalemia was present in 25% of dogs with gastrointestinal foreign bodies and occurred in association with hypochloremia, metabolic alkalosis, and hyponatremia.23 Chloride depletion and sodium avidity related to volume depletion contribute to perpetuation of potassium depletion and metabolic alkalosis in this setting by enhancing urinary losses of Kþ and Hþ ions. The effects of metabolic alkalosis on potassium balance are discussed further in Chapter 10. Urinary loss of potassium is another important cause of hypokalemia, and hypokalemia is common in cats with chronic renal failure. Approximately 20% to 30% of cats with chronic renal failure have hypokalemia at presentation, and in one study, chronic renal disease was the most common associated disorder observed in a survey of cats with hypokalemia.54,60,64,116 Most dogs with chronic renal failure have normal serum potassium concentrations. For example, fewer than 10% of dogs with chronic renal failure caused by renal amyloidosis had hypokalemia at presentation.55 Hypokalemia also commonly occurs during the postobstructive diuresis that follows relief of urethral obstruction in cats with idiopathic lower urinary tract disease. Renal tubular acidosis may be associated with hypokalemia (see Chapter 10). In distal (type I) renal tubular acidosis, hypokalemia is usually present before treatment, and urinary potassium losses may result in part from increased aldosterone secretion. Hypokalemia has been reported in distal renal tubular acidosis in cats.62,197,202 In proximal (type II) renal tubular acidosis, correction of acidosis requires large doses of NaHCO3, and hypokalemia usually appears during therapy. This is a result of the increased delivery of Naþ and HCO3 ions to the distal nephron. These factors enhance urinary potassium excretion by increasing distal tubular flow and lumen electronegativity (HCO3 is a relatively nonresorbable anion in the cortical collecting duct). Finally, hypokalemic nephropathy characterized by chronic tubulointerstitial nephritis may develop in cats fed diets low in potassium and containing urinary acidifiers.30,56,57,58,60 Stimulation of aldosterone secretion by chronic metabolic acidosis and decreased gastrointestinal absorption of potassium may contribute to potassium depletion in this syndrome.61,170 Mutations in genes that encode epithelial transport proteins and channels have been associated with rare

106

ELECTROLYTE DISORDERS

disorders of renal tubular function that cause hypokalemia in humans. One of these familial disorders, Bartter syndrome, is especially complex. It can be caused by mutations in the NKCC2 gene that codes for the luminal Naþ/- Kþ-2Cl cotransporter found in the thick ascending limb of Henle’s loop (type I), in the gene for the ROMK protein component of renal tubular potassium channels (type II), in the CLCNKB gene that codes for the ClC-Kb chloride channel in the basolateral membranes of tubular cells in the thick ascending limb (type III), or in the BSND gene that codes for barttin, a subunit protein of chloride channels that is required for proper insertion of the channels into the basolateral membrane (type IV).94,176 Other rare diseases can arise from a loss of function mutation in the NCCT gene for the thiazide-sensitive Naþ-Cl cotransporter in the luminal membranes of the distal convoluted tubule (Gitelman’s variant of Bartter syndrome) or a gain of function mutation in the SCNN1 gene for the luminal sodium channel (ENaC) of the principal cells of the collecting duct (Liddle syndrome). None of these rare tubular disorders has yet been recognized in veterinary medicine. Mineralocorticoid excess is a relatively uncommon cause of urinary potassium loss and hypokalemia in dogs and cats. One report described hyperaldosteronism in a dog thought to be caused by adrenal hyperplasia of the adrenal zona glomerulosa.26 Older dogs with hyperaldosteronism as a result of aldosterone-producing adenomas or adenocarcinomas are presented for evaluation of weakness and polyuria.67,102,207 Mild to moderate hypertension may be detected, and hypokalemia, hypernatremia, mild metabolic alkalosis, dilute urine, and extremely high serum aldosterone concentrations (>3000 pmol/L; normal, 14 to 957 pmol/L) are present on laboratory evaluation. Dogs with adenomas respond well to surgical adrenalectomy, but those with adenocarcinomas experience recurrence of clinical signs if metastasis occurs. One affected dog had polyuria and hyperaldosteronism associated with a very small (2 mm) adrenal adenoma that initially was undetected by computed tomography, and the dog responded to treatment with spironolactone.159 Ultimately, the tumor was identified and removed, and the dog recovered completely. Another dog with an adrenal adenocarcinoma had clinical and laboratory features of mineralocorticoid excess, but serum aldosterone concentration was undetectable.156 Serum desoxycorticosterone concentration was measured and found to be abnormally high (288 ng/mL; normal, 16 to 46 ng/mL). After surgical removal of the tumor, serum potassium concentration normalized, but serum desoxycorticosterone concentration remained high, and the dog was treated with spironolactone. Clinical and laboratory features of hyperaldosteronism also may be observed in dogs with hyperadrenocorticism caused by adrenal tumors that have been documented to

produce a variety of steroid hormones, including aldosterone, deoxycorticosterone, and corticosterone in addition to cortisol.17,46,115,117,130 Several cats (5 to 20 years of age) with hyperaldosteronism have been reported.* Affected cats are presented for evaluation of muscle weakness (sometimes with ventroflexion of the neck), ataxia, weight loss, polyuria, polydipsia, and ocular abnormalities (e.g., mydriasis, blindness, and retinal detachments) associated with hypertension. Laboratory features consist of hypokalemia, hypernatremia, mild metabolic alkalosis, increased serum creatine kinase activity, dilute urine, extremely high serum aldosterone concentrations (usually >1000 pmol/L; normal, 194 to 388 pmol/L), and plasma renin activity that is low or at the low end of the normal reference range (0.2 to 0.5 ng/L/sec; normal, 0.2 to 1.4 ng/L/sec). In two affected cats, urinary FEK was more than 50% and consistent with inappropriate kaliuresis.75 In these two cats, chronic renal disease also was present and might have contributed to hypertension and increased FEK. Hyperaldosteronism may be associated with hyperglycemia caused by insulin resistance, and one affected cat had diabetes mellitus that persisted despite removal of the adrenal tumor.75 Aldosterone-producing adrenal tumors in cats often are 1 to 3 cm in diameter and can be visualized on abdominal ultrasonography. Cytologic evaluation of fine needle aspirates is consistent with neuroendocrine neoplasia, and histologically these tumors are adenomas or adenocarcinomas. Invasion of the caudal vena cava by the tumor can result in thromboembolism.11,75,163 Removal of adrenal tumors causing hyperaldosteronism can be successful in affected cats, but surgery carries a high risk of life-threatening intraoperative or postoperative hemorrhage.11 Medical treatment with potassium supplementation (2 to 6 mEq/day), spironolactone (2 to 4 mg/kg/day) to antagonize the effects of aldosterone, and amlodipine (0.625 to 1.25 mg/day) to control hypertension and can be used to manage affected cats. Survival times of 1 to 5 years have been reported for surgical or medical management.11 Concurrent hyperaldosteronism and hyperprogesteronism have been reported in a cat that also had clinical signs of hyperadrenocorticism and diabetes mellitus.27 See Chapter 10 for further discussion of states of mineralocorticoid excess. Administration of loop or thiazide diuretics may cause hypokalemia as a result of increased flow rate in the distal tubules and increased secretion of aldosterone secondary to volume depletion. In one study, dogs with heart failure receiving furosemide had significantly lower mean serum potassium concentrations (mean serum potassium concentration, 3.9 mEq/L) than did normal dogs (mean serum potassium concentration, 4.4 mEq/L) or

*References 11, 63, 75, 119, 134, 135, 157, 159, 160, 163.

Disorders of Potassium: Hypokalemia and Hyperkalemia untreated dogs with arrhythmias (mean serum potassium concentration, 4.3 mEq/L).40 Of the dogs treated with furosemide, 17% had serum potassium concentrations less than 3.0 mEq/L. In another study, 10 dogs with congestive heart failure treated with captopril, furosemide, and a sodium-restricted diet did not develop significant changes in serum electrolyte concentrations.165 Penicillin derivatives may cause hypokalemia by acting as nonresorbable anions in the distal tubule and increasing secretion of potassium into tubular fluid. Amphotericin B may cause increased loss of potassium by binding to sterols in cell membranes and increasing permeability. Peritoneal dialysis can be complicated by hypokalemia if potassium-free dialysate is used for an extended time period.45

TREATMENT Preparations available for parenteral use include KCl (2 mEq Kþ/mL) and a potassium phosphate solution containing K2HPO4 and KH2PO4 (4.36 mEq Kþ/mL). Potassium chloride is the additive of choice for parenteral therapy because chloride repletion is essential if vomiting or diuretic administration is the underlying cause of hypokalemia. Replacement of chloride is also essential for resolution of the metabolic alkalosis often present in such settings (see Chapter 10). When administered intravenously, KCl generally should not be infused at rates greater than 0.5 mEq/kg/hr to avoid potential adverse cardiac effects. A scale such as that shown in Table 5-2 may be used to estimate the amount of KCl to add to parenteral fluids based on serum potassium concentration.88 Infusion rates greater than 0.5 mEq/kg/hr may be required to normalize serum potassium concentration in hypokalemic patients with diabetic ketoacidosis treated with insulin. In hypokalemic human patients, potassium infusion rates up to 0.9 mEq/kg/hr were used safely in one study.89 Careful mixing of potassium chloride after addition to flexible bags of fluids is extremely important to prevent the patient from receiving a high concentration of potassium that could be life threatening. In one

TABLE 5-2

study, inadequate mixing of potassium chloride added to flexible bags of fluid was demonstrated to result in up to a fourfold increase in the concentration of potassium in the fluids.51 For determination of serum potassium concentration, when submitting blood samples that have been drawn from intravenous catheters in patients receiving potassium-supplemented fluids, the initial volume of blood withdrawn should be discarded, and a second sample should be submitted to the laboratory to avoid results that may be spuriously high. Infusion of potassium-containing fluids initially may be associated with a decrease in serum potassium concentration as a result of dilution, increased distal renal tubular flow, and cellular uptake of potassium, especially if the infused fluid also contains glucose.60 This effect may be minimized by using a fluid that does not contain glucose, administering fluids at an appropriate rate, and beginning oral potassium supplementation as soon as possible. The concentration of potassium in the infused fluid generally should not exceed 60 mEq/L, because higher concentrations of potassium may cause pain and sclerosis of peripheral veins.162 Parenteral fluids containing up to 35 mEq/L have been used safely by the subcutaneous route.72 Careful potassium supplementation is important when using insulin to treat diabetic ketoacidosis. Chronic potassium depletion is usually present in affected patients as a result of loss of muscle mass, anorexia, vomiting, and polyuria. However, serum potassium concentrations are sometimes normal or even increased because of the effects of insulin deficiency and hyperosmolality on serum potassium concentration. Because blood glucose concentration decreases with insulin treatment, marked hypokalemia may develop if supplementation is not adequate. Potassium gluconate (e.g., Kaon and Tumil-K) is recommended for oral supplementation. In one study, orally administered KCl and KHCO3 were not palatable to cats.59 Dogs may require 2 to 44 mEq potassium per day, depending on body size.92 In cats with hypokalemic

Guidelines for Routine Intravenous Supplementation of Potassium in Dogs and Cats

Serum Potassium Concentration (mEq/L) 1,000,000/µL? • Hemolysis of HK RBC (e.g., Akita, Shiba)? Yes

No Drug administration? • ACE inhibitors • NSAIDs • Heparin • AII receptor blockers • K+-sparing diuretics • Trimethoprim

Pseudohyperkalemia

Yes

• Cyclosporin A, tacrolimus

No

Consider iatrogenic hyperkalemia, especially when drug administration is combined with decreased renal function or potassium supplementation Thrombolytic therapy?

Translocation of K+ ICF Æ ECF

Yes

No

Reperfusion injury

Hyperglycemia, glucosuria, ketonuria? Yes

No

Diabetic ketoacidosis

Azotemia? Yes

No Drug administration?

Cancer chemotherapy? Yes

Yes

No

• Acute mineral acidosis (e.g., NH4Cl) • b -blockers • Cardiac glycosides • Lysine or arginine infusion

Acute tumor lysis syndrome

Large bladder? Unable to urinate? Yes

No

Urethral obstruction

Trauma? Yes

Ruptured bladder

No

• Hypoadrenocorticism-like syndrome with GI losses (e.g., trichuriasis, salmonellosis) • Pleural or peritoneal effusion (third space loss) • Late pregnancy No

Pre and post-ACTH cortisol concentrations < 1.0 µg/dL? Yes

Hypoadrenocorticism

No Oliguric acute renal failure

Figure 5-13 Algorithm for the clinical approach to hyperkalemia. (Drawing by Tim Vojt.)

renal disease have reduced ability to tolerate an acute potassium load and may require 1 to 3 days to reestablish external potassium balance when intake of potassium is abruptly increased. Dogs with experimentally induced renal disease demonstrate decreased ability to excrete a potassium load. In the first 5 hours after a potassium load, dogs with experimentally induced renal disease excreted 30% to 37% of administered potassium, whereas control dogs excreted 56% to 67%.24,25 Kaliuresis was blunted in the dogs with remnant kidneys despite exaggerated hyperkalemia and increased secretion of aldosterone, and approximately 24 hours were required for complete excretion of the potassium load. Episodes of moderate (>5.3 mEq/L) or severe (6.5 mEq/L) hyperkalemia

occurred commonly in a population of dogs with naturally occurring chronic renal disease and responded well in 18 of 26 dogs managed by feeding them a potassium-reduced, home-prepared diet.184 Oliguria or anuria with hyperkalemia is more likely to occur in acute renal failure (e.g., ethylene glycol ingestion), but these findings may be observed terminally in chronic renal failure. Acute renal failure with oliguria or anuria is associated with hyperkalemia for several reasons. First, there has been insufficient time for renal adaptation to nephron loss, as occurs with chronic renal failure. Severe reductions in GFR and urine output result in inadequate distal tubular flow for effective urinary excretion of potassium. Finally, increased release of potassium from tissues during this

ELECTROLYTE DISORDERS

112 150 120 105

FEK (%)

90 75 60 45 30 15

15

30

45

60

75

90

105 120 135 150

GFR (mL/min)

Figure 5-14 Nomogram relating fractional potassium excretion (FEK) to glomerular filtration rate (GFR). Values for patients with an intact hormonal and renal tubular secretory mechanism for potassium (closed triangles) are used to delineate the hatched area. The open squares and circles indicate patients with selective aldosterone deficiency and renal tubular secretory defects, respectively. (From Batlle DC,Arruda JA, Kurtzman NA. Hyperkalemic distal renal tubular acidosis associated with obstructive uropathy. N Engl J Med 1981;304:373–380.)

catabolic state and acute metabolic acidosis may contribute to translocation of potassium from ICF to ECF. Hyperkalemia, hyponatremia, and Na/K ratios less than 27:1 are usually, but not always, found in dogs and cats with hypoadrenocorticism.* In dogs with hypoadrenocorticism, hyperkalemia has been reported in 74% to 96%, hyponatremia in 56% to 100%, and Na/K ratios less than 27:1 in 85% to 100% of cases. Hyperkalemia was found in 9 of 10 cats with hypoadrenocorticism, whereas hyponatremia and Na/K ratios less than 27:1 were found in all 10 affected cats.150 Treatment is begun immediately after a presumptive diagnosis of hypoadrenocorticism is made, but conclusive diagnosis requires results of an ACTH stimulation test. If sodium intake is sufficient to maintain normal ECF volume and distal tubular flow rate, an animal with hypoadrenocorticism may be able to maintain potassium balance. Treatment of dogs with hypoadrenocorticism with fluids alone also often decreases serum potassium concentration into the normal range. However, usually these animals are presented with anorexia and vomiting that contribute to decreased ECF volume and urine output, and without adequate endogenous mineralocorticoids, they are unable to excrete sufficient potassium to prevent frank hyperkalemia. Electrolyte abnormalities similar to those found in dogs with hypoadrenocorticism (i.e., hyponatremia and

*References 113, 150, 151, 154, 161, 168, 174, 192, 206.

hyperkalemia) can occur in dogs with gastrointestinal disease related to trichuriasis, salmonellosis, or perforated duodenal ulcer.53,123 Hyperkalemia in affected dogs with trichuriasis is not caused by a deficiency of aldosterone because plasma aldosterone concentrations have been found to be normal or high.87 Hyperkalemia and hyponatremia also have been observed in dogs and cats with chylous pleural and peritoneal effusions, in a dog with pleural effusion caused by a lung lobe torsion, in a dog with a neoplastic pleural effusion, in a dog with portal hypertension and peritoneal effusion associated with Bartonella henselae infection, and in cats with peritoneal effusion caused by neoplasia or feline infectious peritonitis.{ The hyperkalemia observed in these situations is thought to arise from decreased renal excretion of potassium as a consequence of volume depletion (e.g., gastrointestinal fluid loss and third-space loss of fluid) and decreased distal renal tubular flow. Hyperkalemia and hyponatremia also have been reported in three female Greyhounds late in pregnancy.175 The underlying mechanism was unknown, but all of the dogs had a history of vomiting or diarrhea. Hyporeninemic hypoaldosteronism is an important cause of unexplained asymptomatic hyperkalemia in human patients, but this disorder has rarely been recognized in veterinary medicine.47 Many affected human patients have mild to moderate renal insufficiency caused by diabetic glomerulosclerosis or interstitial renal disease. Most of them have low plasma renin and aldosterone concentrations. Even in patients with normal plasma aldosterone concentrations, the concentration of this hormone must be considered abnormal in light of the hyperkalemia. Resting plasma cortisol concentrations and response to ACTH are normal. Hyperchloremic metabolic acidosis and hypertension may be observed. It is unclear whether low aldosterone concentrations are a consequence of diminished renin secretion and lack of trophic effect of angiotensin II on the zona glomerulosa of the adrenal cortex or whether there is a primary adrenal defect in aldosterone secretion. To document this syndrome in veterinary patients would require demonstration of subnormal plasma renin and aldosterone concentrations or a subnormal increase in aldosterone after volume contraction or ACTH administration. Normally, aldosterone concentrations increase in response to ACTH in the dog.207 In this study, one dog with diabetes mellitus was suspected to have hyporeninemic hypoaldosteronism based on a subnormal aldosterone response to ACTH. Several drugs may contribute to hyperkalemia, especially when used in combination with one another, in conjunction with potassium supplementation, or in patients with renal sufficiency.148 Nonspecific b-blockers (e.g., propranolol) interfere with catecholaminemediated uptake of potassium by liver and muscle by {

References 21, 105, 111, 193, 206, 210.

Disorders of Potassium: Hypokalemia and Hyperkalemia blocking b2-adrenergic stimulation of cell membrane Naþ, Kþ-ATPase. Similar to digoxin, cardiac glycoside toxins found in the plant oleander (e.g., oleandrin, digitoxigenin, and Nerium) inhibit Naþ, Kþ-ATPase and can cause hyperkalemia and arrhythmias. The deleterious effects of oleandrin are blocked by infusion of fructose1,6-diphosphate.126 Angiotensin-converting enzyme inhibitors (e.g., enalapril) and angiotensin II receptor blockers (e.g., losartan) contribute to hyperkalemia by decreasing production of aldosterone by the adrenal glands and blunting glomerular efferent arteriolar constriction, which potentially can decrease delivery of sodium and water to the distal nephron and impair renal potassium excretion. Prostaglandins stimulate renin release, and use of nonsteroidal anti-inflammatory drugs may contribute to development of hyperkalemia. These drugs also may impair the stimulatory effect of prostaglandins on potassium channels in the luminal membranes of renal tubular cells. Heparin impairs aldosterone production by decreasing the number and affinity of angiotensin II receptors in the zona glomerulosa of the adrenal glands and may contribute to hyperkalemia in the presence of other predisposing factors.144 Potassiumsparing diuretics (e.g., spironolactone, amiloride, and triamterene) reduce urinary excretion of potassium and can cause hyperkalemia. Spironolactone competitively inhibits binding of aldosterone to its cytoplasmic receptor in the principal cells of the collecting duct. Amiloride and triamterene block sodium channels in the luminal membranes of the principal cells. Trimethoprim is similar in structure to amiloride and also inhibits sodium channels in the luminal membranes of the principal cells. Trimethoprim is most likely to cause hyperkalemia at high dosages, when urine pH is low (5 mEq/L or UAstrong >5 mmol/L Hyperchloremic acidosis

AG or SIGsimplified 10%) dehydration. For patients with moderate to severe dehydration, inadequate oral intake, electrolyte imbalance, or signs of hypovolemic or endotoxic shock, intravenous fluid administration is necessary. The rate of fluid administration depends on the presence or absence of shock, the extent of dehydration, and the presence of cardiac or renal disease that may predispose the patient to volume overload. Patients with a history of vomiting that are mildly dehydrated are usually responsive to crystalloids (e.g., lactated Ringer’s solution or 9% NaCl) at a rate that provides maintenance needs and replaces existing deficits and ongoing losses over a 24-hour period. Patients with signs of shock require more aggressive support. The volume deficit can be replaced with crystalloids at an initial rate of 60 (cat) to 90 (dog) mL/kg/hr, which is then tailored to maintain tissue perfusion and hydration. Central venous pressure monitoring and evaluation of urine output are necessary for patients with severe gastrointestinal disease, especially those with third-space losses of fluid into the gut or peritoneum. Colloids and hypertonic solutions can also be used to reduce the amount of crystalloid required (e.g., 5 mL/kg of 7% NaCl in 6% dextran intravenously, 10 to 20 mL/kg/day of degraded gelatin [Haemaccel] intravenously). Colloids are also useful in hypoproteinemic patients. Endotoxic shock is a common complication of severe gastrointestinal disease. Warning signs of endotoxemic shock include fever or subnormal body temperature, tachycardia, increased respiratory rate, slow capillary refill time, hyperemic or pale mucous membranes, transient leukopenia followed by leukocytosis with a left shift and toxic neutrophils, low-normal central venous pressure, and bounding pulses. Patients with endotoxic shock must be treated aggressively with fluid therapy, broad-spectrum antibiotics, glucocorticoids, oxygen, glucose, and bicarbonate as indicated.47

The effect of vomiting and diarrhea on acid-base balance is difficult to predict, and therapeutic intervention to correct acid-base imbalance should be based on blood gas analysis. Patients with normal acid-base status or mild metabolic acidosis may be given lactated Ringer’s solution at a rate sufficient to correct fluid deficits and provide for maintenance and ongoing losses for a 24-hour period. Potassium depletion may be a consequence of prolonged diarrhea, vomiting, or anorexia, but most polyionic replacement fluids contain only small amounts of potassium. Consequently, KCl is usually added to parenteral fluids and adjusted based on serum potassium concentrations. When severe metabolic acidosis is present (pH 50% of kcal)—caloric density is about 1.5 kcal/mL; use for high energy needs, volume restriction, or diarrhea caused by high (>50% of kcal) carbohydrate products b. Fiber containing—improved feces consistency over high-carbohydrate diets 2. Defined-formula diets—diets specifically modified for use in patients with impaired organ function a. Impaired gastrointestinal function (including anorexia for >2 weeks); contains peptides, mediumchain triglyceride, and glucose polymers versus intact macronutrient sources; may cause diarrhea b. Impaired liver function—high branched-chain/ aromatic amino acid content c. Impaired kidney function—supplemented with B-ketoacids of some essential amino acids d. “Stress”—high branched-chain amino acid content, high caloric density (>1 kcal/mL); efficacy disputed B. Energy 1. Total calories a. Restricted (1 kcal/mL) because of diseaseinduced increases in requirements; severe trauma, burns, sepsis, hyperthyroidism 2. Carbohydrate

A local nutrition support service dietitian may also be of help; these professionals may have nutritional products and delivery apparatus available, and possess extensive training and experience in nutritional support. Blenderized gruels made with commercial canned dog or cat diets are suitable for use in feeding tubes larger than 14 Fr in size. Specific products marketed for critically ill dogs or cats have been recently published.86 Table 26-3 presents the nutrient compositions of the liquid enteral products currently used in the Critical Care Unit at Michigan State University and in the Small Animal Intensive Care Unit at The Ohio State University. None of the human enteral products appear to contain adequate arginine for cats. Signs compatible with arginine deficiency have been reported after prolonged feeding of enteral diets designed for humans and those designed specifically for cats. In the absence of data to support other recommendations, we add 1 mg arginine/kcal (approximately 200 to 300 mg/cat/day) to human enteral liquid

a. Increased (>50% of kcal) due to diarrhea (high carbohydrate feedings) or increased energy needs b. Decreased ( 2 times in 24 hours

12-hour cessation of EN ordered

1 or more episodes of vomiting or regurgitation

No vomiting or regurgitation in 12-hour period

Additional 12-hour cessation of EN ordered

Resume feeding at last recorded caloric volume/rate

Continued vomiting or regurgitation

No vomiting or regurgitation

Alternative method of nutrient delivery

Resume feeding at lowest rate (1/3 RER)

Figure 26-17 Suggested Rescue Protocol for canine patients admitted to the critical care unit receiving enteral nutrition support through nasoenteric feeding tubes.47

642

SPECIAL THERAPY

Figure 26-18 Kangaroo Enteral Feeding Pumps for continuous or intermittent feeding.

In most instances, feedings can be increased to the total calculated dose over 24 to 72 hours, unless anorexia has been prolonged. Most anorexic patients begin voluntary food consumption after the second or third day. If fluid losses are present, they can be replaced by administration of additional water along with the diet.

GASTRIC RESIDUAL VOLUMES (GRVS)

Figure 26-19 Medfusion 2010i syringe pump for continuous or intermittent enteral feeding of cats and small dogs.

(Figure 26-19) to be suitable for either bolus or CRI feedings. This allows delivery of a precise volume over a given time frame, minimizing the possibility of “overdosing” a patient with an excess volume or giving boluses too rapidly. If the animal has been anorexic for more than 72 hours, the volume administered over the first 12 to 24 hours may be reduced by one third or one half of the total RER to reduce the risk of gastrointestinal intolerance.

For human patients with feeding tubes that reside within the stomach, gastric residual volumes (GRVs) are checked every 4 to 6 hours, regardless of the method of feeding (bolus or continuous).65,73 The nasoenteric feeding tube should be aspirated and the GRV should be recorded. New evidence in veterinary medicine indicates that GRVs can be slowly administered back through the nasogastric tube over 5 minutes and flushed with 5 mL of water before the next scheduled feeding with no overt gastrointestinal complications observed.47 In many human hospitals, intolerance to enteral feeding is assessed in part by serially measuring gastric residual volumes (GRVs).71,72 High GRVs have been reported as a reason to stop feeding human patients. Gastric aspirates greater than 150 to 250 mL in a 4-hour period were considered a mark of intolerance in adults, and cessation of EN had been recommended to minimize risk of aspiration pneumonia.65 In a study of low birth weight infants, delayed gastric emptying was defined as GRVs greater than 5 mL/kg in any 4-hour period.48 However, routine suspension of EN due to large GRVs has been questioned in human medicine, as there is no consistent relationship demonstrated between aspiration and gastric residual volumes.53,69,73,74,93 The incidence of fourth hourly GRV was not different between the continuous and

Enteral Nutrition intermittent feeding groups,48 which is similar to the recent findings of Holahan et al. It has been previously thought in veterinary and human medicine that high GRVs correlated with a higher incidence of vomiting, regurgitation, and the incidence of aspiration pneumonia. However, no significant correlation between average GRVs (mL/kg) and occurrence of vomiting or regurgitation was found in the Holahan study, and only two patients on EN had aspiration pneumonia, both of which had radiographic evidence of pneumonia documented before the start of EN. There is a lack of evidence in veterinary medicine to suggest what an acceptable GRV might be, or whether this measurement is a reliable means of assessment of gastric intolerance. The authors conclude that termination of enteral feedings due to elevated GRVs in critically ill dogs may not be warranted, particularly in patients not exhibiting signs of vomiting or regurgitation.

COMPLICATIONS OF ENTERAL FEEDING Mechanical, gastrointestinal, technical, and metabolic complications can occur with enteral tube feeding. Mechanical problems relate to placement (inadvertent placement of the tube into the trachea, regurgitation or vomiting up the tube, or removal of the tube by the patient) and maintenance of the tube, including clogged or blocked tubes. Tubes can become clogged if pills are crushed and forced into the tube, or if food is not adequately flushed out of the tube after bolus feeding or the tube is not capped to leave a column of water in the tube. Flushing clogged tubes with a variety of solutions (cranberry juice and cola beverages have been recommended in the past) may be irrational if the pK of the clogging material is not considered.12 Another potential problem that should be considered is drug incompatibility with enteral feeding formulas. Potassium chloride elixir is the most commonly reported incompatibility, causing formula precipitation and blocked tubes.4 Tube clogging is best prevented by prohibiting its use for the administration of nonliquid materials and by properly flushing and capping the tube after each use. To decrease the incidence of tube occlusion, we recommend flushing the tube with a small amount (3 to 5 mL) of warm water every 4 to 6 hours regardless of the delivery method; this becomes especially important in small dogs and cats. Most of the gastrointestinal problems caused by enteral feeding result from too rapid administration of the solution, or feeding solutions of high osmolality (>600 mOsm). Solutions entering the duodenum too quickly can cause vomiting, cramps, and diarrhea by overwhelming the normal neural and endocrine control mechanisms of the gastrointestinal tract.18 Hyperosmolar solutions cause rapid fluid and electrolyte influx into the gut lumen, leading to abdominal distention and

643

cramping. These problems can be managed by reducing the feeding rate or solution concentration, or by feeding diets that delay gastric emptying. Fiber-enriched human enteral products are marketed to normalize GI transit times, but there is little evidence to suggest that these products minimize or reduce diarrhea in tube-fed patients.91 Cold feeding solutions are thought to be a major cause of diarrhea, but research available to support this opinion is inconclusive. We have not observed problems with diets fed at room temperature and recommend that diets stored in a refrigerator be brought to room temperature before being tube fed or offered for oral consumption. Bacterial contamination of formulas during reconstitution and administration can also cause diarrhea,9 so diets should be handled carefully and not allowed to remain in the feeding tube system for more than 24 hours.40 Patient problems that may result in gastrointestinal intolerance to feeding include protein malnutrition, hypoalbuminemia from any cause, fat malabsorption, and associated gastrointestinal disorders. Hypoalbuminemia can lead to malabsorption and diarrhea because intravascular osmotic pressure required for nutrient uptake is greatly reduced.38 Pancreatic disease, biliary obstruction, ileitis, bacterial overgrowth, gastric surgery, and intestinal resection all can lead to varying degrees of fat malabsorption. Many medical disorders are associated with diarrhea,50 including enteric infections, malabsorption syndromes, gastroenteritis, mucosal defects, diabetes mellitus, carcinoid syndromes, hyperthyroidism, and immunodeficiency states. The most important cause of diarrhea in tube-fed human patients is concomitant antibiotic or drug use because it is the most difficult factor to change. Development of diarrhea often results in discontinuation of tube feedings in an attempt to reduce the diarrhea that nursing staff and clinicians associate with liquid feedings. However, the problem may be the antibiotic used or related to the underlying disease process. By altering the normal gastrointestinal flora, antibiotic usage can also change the fatty acid composition of gut contents, which may adversely affect sodium and water absorption by the colon. Intestinal bacteria also may ferment undigested nutrients, forming organic acids, hydrogen and carbon dioxide gas, and ultimately causing diarrhea. Many antibiotics and drugs have been associated with diarrhea including penicillins, aminoglycosides, cephalosporins, chloramphenicol, clindamycin, aminophylline, cimetidine, potassium chloride, digoxin, and magnesium-containing antacids.44 The oral suspensions of some antibiotics, electrolytes, and other medications are hyperosmolar, suggesting that these drugs may play a role in the pathogenesis of diarrhea.78 A study investigating the cause of diarrhea in tube-fed patients found that medicinal elixirs containing theophylline, as well as liquid preparations of acetaminophen, codeine, cimetidine,

644

SPECIAL THERAPY

isoniazid, and vitamins all contained sorbitol. Although commonly added to improve the palatability of some medications, sorbitol is also known to have an osmotic effect in sensitive individuals and should be considered a potential cause of diarrhea in patients receiving enteral diets.33 Technical complications include: feeding stopped for diagnostic or therapeutic procedures, feeding stopped for owner visitation, feeding stopped for extended walks outside, equipment malfunction, syringe pump off, syringe pump disconnected, or accidental rate change. Care should be taken when patients are receiving enteral nutrition to minimize the time lost due to temporary discontinuation of enteral feedings. Mechanical complications were evaluated at Michigan State University. Dogs fed continuously lost a median of 0.5 hours of enteral nutrition, where dogs fed intermittently had a median of 0 hours of EN lost due to technical complications. The 0.5 hours lost in the continuous group equates to 2% of the 24 hour PPND. This discrepancy accounts for the lower average PPND (approximately 2%) of continuously fed dogs (98.4%) when compared with the intermittently fed group (100%).47 Metabolic problems include rapid absorption of highcarbohydrate solutions, which could result in hyperglycemia; osmotic diuresis; and ultimately nonketotic, hyperosmolar coma. This is referred to as the refeeding syndrome;90 we do not routinely observe this complication with enteral feedings. Metabolism of glucose also results in the production of CO2 and metabolic water; excess CO2 production can further compromise patients with pulmonary disease. If metabolic water is retained, it can contribute to hyponatremia and edema. Most of these complications can be prevented by acclimatization of the patient to the feeding solution, slow rates of administration, and not overfeeding. Additional monitoring and potential supplementation of electrolytes such as potassium, magnesium, calcium, and phosphate, may be required in critically ill patients. Urine or blood glucose concentrations should be monitored at regular intervals if hyperglycemia is suspected. Hyperglycemia can be managed by reducing nutrient flows or by giving insulin, as discussed in Chapter 25. Until there is more documentation (controlled studies) in the veterinary literature, evidence-based guidelines for enteral feeding in human adult patients41,91 should be reviewed by practitioners and their staff interested in providing nutrition support to veterinary patients.

CONCLUSIONS • Early enteral nutrition may be beneficial to the veterinary patient, although further studies are needed in clinical populations. • Early initiation of nutrition should be an integral part of the patient’s treatment plan consideration to supply enterocyte nutrition and protect the mucosal barrier.

• Both the continuous and intermittent methods of nasoenteric tube feeding can facilitate adequate nutrient intake with minimal GI complications in critically ill dogs. • Additional research is needed to determine optimal diet formulations for different patient populations.

REFERENCES 1. Abood S, Buffington CA. Improved nasogastric intubation technique for administration of nutritional support in dogs. J Am Vet Med Assoc 1991;199:577–9. 2. Abood S, Mauterer JV, McLoughlin MA, et al. Nutritional support of hospitalized patients. 2nd ed. Philadelphia: WB Saunders Co; 1993. 3. Abrams J. The nutrition of the dog. In: Boca Raton, Fla: CRC Press; 1977. p. 1–27. 4. Altman E, Cutee AJ. Compatibility of enteral products with commonly employed drug additives. Nutr Supp Serv 1984;4:8–17. 5. Anderson R. Water balance in the dog and cat. J Small Anim Pract 1982;23:588–98. 6. Beal M, Jutkowitz LA, Brown AJ. Development of a novel method for fluoroscopically-guided nasojejunal feeding tube placement in dogs. J Vet Emerg Crit Care 2007;17(3):S1. 7. Beal M, Mehler SJ, Staiger BA, et al. Technique for fluoroscopic nasojejunal tube placement in dogs, In: Proceedings of the American College of Veterinary Internal MedicineCanada: Montreal, Quebec; 2009 June 3–6th. 8. Beal M, Mehler SJ, Staiger BA, et al. Technique for percutaneous radiologic gastrojejunostomy in the dog, In: Proceedings of the American College of Veterinary Internal MedicineCanada: Montreal, Quebec; 2009 June 3–6th. 9. Belknap D, Davison U, Flournoy DJ. Microorganisms and diarrhea in enterally fed intensive care unit patients. J Parenter Enter Nutr 1990;14:622–8. 10. Bone S, Mann FA, Backus RC, et al. Evaluation of Polymeric Diets delivered directly into the small intestine through surgically placed jejunostomy tubes, In: Proceedings of the American College of Veterinary Internal MedicineMontreal, Quebec; 2009 CAN; June 3-ttgh. 11. Bright R. Percutaneous tube gastrostomy. 2nd ed. Philadelphia: Lea and Febiger; 1990. 12. Buffington C, Blaisdell J. Effect of citrate solutions on enteral formula clot formation. J Parenter Enter Nutr 1989;13:14S. 13. Buffington C, Holloway C, Abood SA. Normal dogs. In: St Louis, MO: Elsevier; 2004. p. 24–5. 14. Buffington C. Nutritional support of the hospitalized patient. Video Forum for Small Anim Clin 1989. 15. Butterworth CJ. The skeleton in the hospital closet. Nutr Today 1974;9:4–8. 16. Campbell J, Jutkowitz LA, Santoro K, Hauptman J, Holahan ML, Brown AJ. Continuous versus intermittent delivery of nutrition via nasoenteric feeding tubes in critically ill canine and feline patients. J Vet Emerg Crit Care 2010;20:232–6. 17. Carr C, Ling KD, Boulos P, et al. Randomised trial of safety and efficacy of immediate postoperative enteral feeding in patients undergoing gastrointestinal resection. BMJ 1996;312:869–71.

Enteral Nutrition 18. Cataldi-Betcher E, Seltzer MH, Slocum BA, et al. Complications occurring during enteral nutrition support: A prospective study. J Parenter Enter Nutr 1983;7:546–52. 19. Center S, Wilkinson E, Smith CA, et al. 24-Hour urine protein/creatinine ratio in dogs with protein-losing nephropathies. J Am Vet Med Assoc 1985;187(8):820–4. 20. Chandler M, Guilford WG, Lawoko CRO. Comparison of continuous versus intermittent enteral feeding in dogs. J Vet Intern Med 1996;10(3):133–8. 21. Chandra R, Scrimshaw NS. Immuno-competence in nutritional assessment. Am J Clin Nutr 1980;33:2694–7. 22. Chyka P, Seger D, Krezelok EP, Vale JA. Position paper: single-dose activated charcoal. Clin Toxicol 2005;43 (2):61–87. 23. Ciocon J, Galindo-Ciocon DJ, Tiessen C, Galinda D. Continuous compared with intermittent tube feeding in the elderly. J Parenter Enter Nutr 1992;16:525–8. 24. Clark C. Fluid Therapy. 2nd ed. Santa Barbara, Calif: American Veterinary Publications; 1975. p. 601–9. 25. Crane S. Placement and maintenance of a temporary feeding tube gastrostomy in the dog and cat. Comp Cont Ed Vet Pract 1980;10:770–80. 26. Crowe D. Enteral nutrition for the critically ill or injured patient: Part II. Comp Cont Ed Vet Pract 1986;8:719–30. 27. Crowe D. Methods of enteral feeding in the seriously ill or injured patient: Part I. J Vet Emerg Crit Care 1985;3:1–6. 28. Daye R, Huber ML, Henderson RA. Interlocking box jejunostomy: A new technique for enteral feeding. J Am Anim Hosp Assoc 1999;35:129–34. 29. de Aguilar-Nascimento J, Kudsk KA. Early nutritional therapy: the role of enteral and parenteral routes. Curr Opin Clin Nutr Metab Care 2008;11(3):255–60. 30. Debraekeleer J, Gross KL, Zicker SC. Feeding nursing and orphaned puppies from birth to weaning. 5th ed. Topeka, Kan: Mark Morris Institute; 2010. 31. Della-Ferra M, Baile CA, McLaughlin CL. Feeding elicited by benzodiazepine-like chemicals in puppies and cats: Structure-activity relationships. Pharm Biochem Behav 1980;12:195–200. 32. DeNovo R, Churchill J, Faudskar L, et al. Limited approach to the right flank for placement of a duodenostomy tube. J Am Anim Hosp Assoc 2001;37:193–9. 33. Edes T, Walk BE, Austin JL. Diarrhea in tube-fed patients: Feeding formulas not necessarily the cause. Am J Med 1990;88:91–3. 34. Edney AT, Smith PM. Study of obesity in dogs visiting veterinary practices in the United Kingdom. Vet Rec 1986;118:391–6. 35. Ford R. Nasogastric intubation in the cat. Comp Cont Ed for the Ann Health Tech 1980;1:29–33. 36. Fulton RJ, Dennis JS. Blind percutaneous placement of a gastrostomy tube for nutritional support in dogs and cats. J Am Vet Med Assoc 1992;201:697–700. 37. Gianotti L, Nelson JL, Alexander JW, Chalk CL, Pyles T. Post injury hypermetabolic response and magnitude of translocation: prevention by early enteral nutrition. Nutrition 1994;10(3):225–31. 38. Gottschlich M, Warden GD, Michek M, et al. Diarrhea in tube-fed burn patients: incidence, etiology, nutritional impact, and prevention. J Parenter Enter Nutr 1988;12:338–45. 39. Gross K, Becvarova I, Debraekeleer J. Feeding nursing and orphaned kittens from birth to weaning. 5th ed. Topeka, Kan: Mark Morris Institute; 2010.

645

40. Grunow J, Christenson JW, Moutos D. Contamination of enteral nutrition systems during prolonged intermittent use. J Parenter Enter Nutr 1989;13:23–5. 41. Guidelines for the use of parenteral and enteral nutrition in adult and pediatric patients. J Parenter Enter Nutr 2002;26S:1SA–6SA. 42. Hacker R, Harvey-Banchik LP. Prospective randomized control trial of intermittent versus continuous gastric feeds for critically ill trauma patients. Nutr Clin Pract 2008;23:564–5. 43. Hadfield R, Sinclair DG, Houldsworth PE, et al. Effects of enteral and parenteral nutrition on gut mucosal permeability in the critically ill. Am J Respir Crit Care Med 1995;152(5):1545–8. 44. Hayes-Johnson V. Tube feeding complications: Causes, prevention, therapy. Nutr Supp Serv 1988;6:17–24. 45. Hiebert J, Brown A, Anderson RG, et al. Comparison of continuous vs. intermittent tube feedings in adult burn patients. J Parenter Enter Nutr 1981;5:73–5. 46. Hill RaS, KR. Energy requirements and body surface area of cats and dogs. J Am Vet Med Assoc 2004;225:689–94. 47. Holahan M, Abood SA, Hauptman J. Intermittent and continuous enteral nutrition in critically ill dogs: A prospective randomized trial. J Vet Intern Med 2010;24:520–6. 48. Horn D, Chaboyer W, Schluter P. Gastric residual volumes in critically ill pediatric patients: A comparison of feeding regimens. Aust Crit Care 2004;17:98–103. 49. Ireland L, Hohenhaus AE, Broussard JD, et al. A comparison of owner management and complications in 67 cats with esophagostomy and percutaneous endoscopic gastrostomy feeding tubes. J Am Anim Hosp Assoc 2003;39:241–6. 50. Jergens A. Diarrhea. Philadelphia: WB Saunders; 1995. 51. Kaneko J. Serum proteins and the dysproteinemias. In: Kaneko J, editor. Clinical biochemistry of the domestic animals. 4th ed. New York: Academic Press; 1989. p. 142–65. 52. Kaur N, Gupta MK, Minocha VR. Early enteral feeding by nasoenteric tubes in patients with perforation peritonitis. World J Surg 2005;29(8):1023–7. 53. Kenny D, Goodman P. Care of the patient with enteral tube feeding: an evidence-based practice protocol. Nurs Res 2010;59:S22. 54. Kleiber M. Body size and metabolic rate. Physiol Rev 1947;27:511. 55. Kocan M, Hickisch SM. A comparison of continuous and intermittent enteral nutrition in NICU patients. J Neurosci Nurs 1986;18:333–7. 56. Kompan L, Kremzar B, Gadzijev E, Prosek M. Effects of early enteral nutrition on intestinal permeability and the development of multiple organ failure after multiple injury. Intensive Care Med 1999;25(2):157–61. 57. Koretz RaM, JH. Elemental diets: facts and fantasies. Gastroenterology 1980;78:393–410. 58. Kuehn N. North American companion animal formulary. North American Compendiums, Inc.; 2008. 59. Laflamme DPKG, Lawler DF, et al. Obesity management in dogs. J Vet Clin Nutr 1994;1:59–65. 60. Laflamme DPKR, Schmidt DA. Estimation of body fat by body condition score. J Vet Intern Med 1994;8:154. 61. Lantz G, Cantwell HD, Van Vleet JF, et al. Pharyngostomy tube induced esophagitis in the dog: An experimental study. J Am Vet Med Assoc 1983; 19:207–12. 62. Levenson S. Nutritional assessment-present status, future direction, and prospects. In: Levenson S, editor. Report of

646

63.

64.

65.

66. 67. 68.

69. 70.

71.

72. 73. 74. 75.

76.

77. 78.

79. 80.

81.

SPECIAL THERAPY the second ross conference in medical research. Columbus, Ohio: Ross Laboratories; 1981. Levine P, Smallwood LJ, Buback JL. Esophagostomy tubes as a method of nutritional management in cats: A retrospective study. J Am Anim Hosp Assoc 1997; 33:405–10. Long C, Schaffel N, Geiger JW, et al. Metabolic response to injury and illness: Estimation of energy and protein needs from indirect calorimetry and nitrogen balance. J Parenter Enter Nutr 1979;3:452–6. MacLeod J, Lefton J, Houghton D, Roland C, et al. Prospective randomized control trial of intermittent versus continuous gastric feeds for critically ill trauma patients. J Trauma 2007;63:57–61. MacyDaR SL. Cause and control of decreased appetite. In: Philadelphia: WB Saunders; 1989. p. 18–24. Marik P, et al. Early enteral nutrition in acutely ill patients: a systematic review. Crit Care Med 2001;29(12): 2264–70. Mauterer J, Abood SK, Buffington CA, et al. New technique and management guidelines for percutaneous nonendoscopic tube gastrostomy. J Amer Vet Med Assoc 1994;205:579. Mcclave S, Lukan JK, Stefater JA, et al. Poor validity of residual volumes as a marker for risk of aspiration in critically ill patients. Crit Care Med 2005;33:324–30. McClave S, Martindale RG, Vanek VW, McCarthy M, et al. Guidelines for the provision and assessment of nutrition support therapy in the adult critically ill patient. JPEN J Parenter Enteral Nutr 2009;33(3):277–316. McClave S, Snider HL, Lowen CC, et al. Use of residual volume as a marker for enteral feeding intolerance: prospective blinded comparison with physical examination and radiographic findings. JPEN J Parenter Enteral Nutr 1992;16:99–105. McClave S. Clinical use of gastric residual volumes as a monitor for patients on enteral tube feeding. JPEN J Parenter Enteral Nutr 2002;26:S43–S50. Metheny N, Schallom L, Oliver DA, et al. Gastric residual volume and aspiration in critically ill patients receiving gastric feedings. Am J Crit Care 2008;17:512–9. Metheny N. Residual volume measurement should be retained in enteral feeding protocols. Am J Crit Care 2008;17:62. Mohr A, Leisewitz AL, Jacobson LS, Steiner JM, Ruaux CG, Williams DA. Effect of early enteral nutrition on intestinal permeability, intestinal protein loss, and outcome in dogs with severe parvoviral enteritis. J Vet Intern Med 2003;17:791–8. Moore F, Feliciano DV, Andrassy RJ, McArdle AH, et al. Early enteral feeding, compared with parenteral, reduces postoperative septic complications. The results of a meta-analysis. Ann Surg 1992;216(2):172–83. National Research Council . Nutrient requirements of the dog and cat. Washington DC: National Academies; 2003. Neimiec PJ, Vanderveen TW, Morrison JI, et al. Gastrointestinal disorders caused by medications and electrolyte solution osmolality during enteral nutrition. JPEN J Parenter Enteral Nutr 1983;7:387–9. Orton E. Needle catheter jejunostomy. 2nd ed. Philadelphia: Lea and Febiger; 1990. Pa’pa K, Psa’der R, Steczer A, et al. Endoscopically guided nasojejunal tube placement in dogs for short-term postduodenal feeding. J Vet Emerg Crit Care 2009;19(6):554–63. Pisters P, Ranson JHC. Nutritional support for acute pancreatitis. Surg Gynecol Obstet 1992;175:275–84.

82. Prasad A. Zinc in human nutrition. Boca Raton, Fla: CRC Press; 1979. 83. Prentiss P, Wolf AV, Eddy HE. Hydropenia in the cat and dog: Ability of the cat to meet its water requirements solely from a diet of meat or fish. Am J Physiol 1959;196:626–32. 84. Ragins H, Levenson SM, Singer R, et al. Intrajejunal administration of an elemental diet at neutral pH avoids pancreatic stimulation. Am J Surg 1973;126:606–14. 85. Rawlings C. Percutaneous placements of a midcervical esophagostomy: New technique and representative cases. J Am Anim Hosp Assoc 1993;29:526–30. 86. Saker K, Remillard RL. Enteral assisted feeding. In: 5th ed. Topeka, Kan: Mark Morris Institute; 2010. p. 469. 87. Schaeffer M, Rogers QR, Morris JG. Protein in the nutrition of dogs and cats. Cambridge, UK: Cambridge University Press; 1989. p. 159–205. 88. Serpa L, Kimura M, Faintuch J, Ceconello I. Effects of continuous versus bolus infusion of enteral nutrition in critical patients. Rev Hosp Clin Fac Med S Paulo 2003;8(1):9–14. 89. Silk D. The continuing journey towards the optimization of enteral nutrition 1978–2000-The Nutricia Research Foundation Award acceptance lecture. Clin Nutr 2001;20:5. 90. Solomon S, Kirby DF. The refeeding syndrome: a review. JPEN J Parenter Enteral Nutr 1990;14:90–7. 91. Stroud M, Duncan H, Nightingale J. Guidelines for enteral feeding in adult hospital patients. Gut 2003;52: VII1–VII12. 92. Swann H, Sweet DC, Holt DE, Michel K. Placement of a low-profile duodenostomy and jejunostomy device in five dogs. J Small Anim Pract 1998;39:191–4. 93. Taylor B, Krenitsky J. Nutrition in the intensive care unit: year in review 2008–2009. JPEN J Parenter Enteral Nutr 2010;34:21. 94. Teeter S, Collins DR. Intragastric intubation of small animals. Vet Med Small Animal Clin 1966;61:1067–76. 95. Tyler J. Hepatoencephalopathy. Part II. Pathophysiology and treatment. Comp Cont Ed Vet Pract 1990;12:1260–70. 96. VonWerthern CaW G. A new technique for insertion of esophagostomy tubes in cats. J Am Anim Hosp Assoc 2001;37:140–4. 97. Walton R, Wingfield WE, Oglivie GK, et al. Energy expenditure in 104 postoperative and traumatically injured dogs with indirect calorimetry. J Vet Emer Crit Care 1996;6:71–9. 98. Warner C, Bobo W, Reid S, Rachal J. Antidepressant discontinuation syndrome. Am Fam Physician 2006;74:449–56. 99. Windsor A, Kanwar S, Li A, Barnes E, Guthrie J, et al. Compared with parenteral nutrition, enteral feeding attenuates the acute phase response and improves disease severity in acute pancreatitis. Gut 1998;42(3):431–5. 100. Wohl J. Nasojejunal feeding tube placement using fluoroscopic guidance: technique and clinical experience in dogs. J Vet Emerg Crit Care 2006;16:S27–S33. 101. Wolfe R, Durkot MJ, Wolfe MH. Effect of thermal injury on energy metabolism, substrate kinetics, and hormonal concentrations. Circ Shock 1982;9:383–94. 102. Wolman L, Anderson GH, Marless EB, et al. Zinc in total parenteral nutrition: Requirements and metabolic effects. Gastroenterology 1979;76:458–67. 103. Ziegler E, Fomon SJ. Fluid intake, renal solute load, and water balance in infancy. J Pediatr 1971;78:561–8.

CHAPTER • 27

Fluid Therapy with Macromolecular Plasma Volume Expanders Dez Hughes and Amanda Boag

“Those who fill our professional ranks are habitually conservative. This salutary mental attitude expresses itself peculiarly in our communal relations; namely, when a new idea appears which is more or less subversive to old notions and practices, he who originates the idea must strike sledge hammer blows in order to secure even a momentary attention. This must then be followed by a long, patient, propaganda and advertising until in the grand finale, the public, indifferent at first, is aroused, proceeds to discuss, and finally accepts the iconoclastic proposal as a long-accepted fact of its own invention and asks wonderingly, ‘Why such a bother? What after all is new about this? We knew it long ago!” Howard A. Kelly, MD. Electrosurgery in gynaecology, Ann Surg 93:323, 1931.

In the late nineteenth century, Ernest Starling proposed the concept that the balance between hydrostatic and osmotic pressure gradients between the intravascular and interstitial fluid compartments governed transvascular fluid exchange.151 He postulated that a hydrostatic pressure gradient in excess of the osmotic gradient at the arterial end of the capillary bed results in a net transudation of fluid into the interstitium. At the venous end of the capillary bed, plasma proteins (which do not normally pass out of the blood vessels) exert an osmotic force in excess of the hydrostatic gradient, resulting in a net fluid flux into vessels. More than a century of research has confirmed that Starling’s hypothesis provides the foundation for microvascular fluid exchange but also has revealed that the anatomy and physiology of the microvasculature, interstitium, and lymphatic system are much more complex. Consequently, a much deeper understanding of transvascular fluid dynamics is necessary for a logical and rational approach to intravenous therapy with fluids containing macromolecules. This chapter assumes the reader is familiar with the information given in Chapter 1 explaining the fluid compartments of the body and the mechanisms of water and solute flow among compartments. Although this chapter discusses the anatomy, physiology, and biophysics of transvascular fluid dynamics in some depth, comprehensive reviews and texts are available on the subject

for a more complete discussion of solute and solvent exchange among the microvasculature, interstitium, and lymphatics.5,127,154 The main aim of this chapter is to objectively address the complexities and controversies of colloid therapy while avoiding the tendency toward bias apparent in some articles dealing with the crystalloid-colloid controversy. A deeper appreciation of the relevant issues should ensure a more rational approach when deciding whether colloid therapy is appropriate. The present chapter is exhaustive in its dealing with some issues but not all-inclusive, and the reader also is referred to several reviews of colloid fluid therapy available in the veterinary31,54,98,135 and human medical literature.55,56,103,129,177

THE MICROVASCULAR BARRIER In simple terms, the healthy microvascular barrier is a capillary wall that is relatively impermeable to protein. In addition to the endothelial cell and the capillary basement membrane, a luminal surface layer (the glycocalyx) and the interstitial matrix also contribute to the selective permeability of the microvascular barrier.5,127,180 The glycocalyx coats the luminal aspect of the endothelial cell and is composed of proteins, glycoproteins, and glycolipids that modify the permeability of the

647

648

SPECIAL THERAPY

microvessel by occupying spaces within the wall or via electrostatic attraction or repulsion.93 Plasma proteins, especially albumin and orosomucoid, are thought to contribute significantly to maintaining the selective permeability of the endothelium.45–47,74,102 On a morphologic basis, capillary walls may be continuous, fenestrated, or discontinuous.122,158 Continuous capillaries, which are found in the majority of tissues and organs of the body, are so called because the wall is composed of a continuous endothelial cell and basement membrane. They are freely permeable to water and small solutes such as sodium but are relatively impermeable to macromolecules. The passage of smaller plasma proteins, such as albumin (molecular radius of 3.5 nm), is less restricted than the passage of larger plasma proteins. Fenestrated capillaries have a continuous basement membrane with regions that are only covered by thin endothelial diaphragms or are entirely devoid of endothelium. They are found in tissues characterized by large fluxes of water and small solutes such as the glomerulus and the intestine. Interestingly, the permeability of fenestrated capillaries to macromolecules is similar to that of continuous capillaries. This feature has been shown to be a result of a net negative charge of the basement membrane.12,146 Discontinuous capillaries are found in the liver, spleen, bone marrow, and some glands. They have gaps up to 1 mm between endothelial cells with no basement membrane and are therefore freely permeable to protein. The permeability of the microvascular barrier has been explained by the presence of pores of differing sizes.111 These pore sizes often are extrapolated from experimental data regarding fluid and solute fluxes and do not always correlate with morphologic studies such as electron microscopy, implying that they represent functional rather than anatomic entities. The majority of experimental data suggest there are two effective pore sizes in the microvascular barrier in most tissues, with a high frequency of small pores that restrict efflux of macromolecules and a low frequency of large ones through which macromolecules can pass freely.127 Rather than being a free fluid space, the interstitium represents a dynamic environment that may contribute to the permeability characteristics of the microvascular barrier and modify the flow of fluid and macromolecules from the blood vessels to the lymphatics.5,10,11 The interstitium is composed of a collagen framework that contains a gel phase of glycosaminoglycans (of which hyaluronan is the most common), along with protein macromolecules and electrolytes in solution. The relative proportions of these constituents differ widely among organs and tissues, resulting in variations in the permeability and mechanical properties of the interstitium. Glycosaminoglycans are extremely long chains of repeating disaccharide subunits wound into random coils and entangled with each other and the collagen framework. They have molecular weights of the order of 107, and each molecule bears many

thousand anionic moieties.5 This interstitial structure has been suggested to mechanically oppose distention (i.e., edema formation) and resists contraction during dehydration because of repulsion between the anionic moieties.71 The interstitial matrix itself is differentially permeable to macromolecules, and a colloid osmotic gradient also can exist from the perimicrovascular space across the interstitium to the lymphatics. Although the collagen network and many of the glycosaminoglycans are fixed in the interstitium, hyaluronan may be mobilized and removed via lymphatic drainage, thereby altering the permeability of the interstitium.5 Increased microvascular permeability may occur during inflammatory states thereby exacerbating macromolecule extravasation.

TRANSVASCULAR FLUID DYNAMICS Although not stated implicitly in his seminal article, Starling’s hypothesis was subsequently formalized to state simply that the hydrostatic pressure gradient between the capillary and the interstitium (Pc
Pi) is equal to the osmotic pressure gradient between the plasma and the interstitium (pp
p i). This expression can be expanded to describe fluid flux (Jv) across the microvascular barrier: Fluid flow ¼ hydrostatic gradient
osmotic gradient or Jv ¼ ðPc
Pi Þ
ðpp
pi Þ For a solute to exert its full osmotic pressure across a membrane, the membrane must be impermeable to the solute. If the membrane is partially permeable to the solute molecule, the equilibrium concentration gradient is lower, and the solute exerts only part of its potential osmotic pressure. The realization that the microvasculature was only partially impermeable to smaller macromolecules led to the inclusion of the reflection coefficient (s) in the fluid flux equation.155 Jv ¼ ðPc
Pi Þ
sðpp
pi Þ In descriptive terms, the reflection coefficient is the fraction of the total potential osmotic pressure exerted by the solute in question. Conceptually, one also can consider it as the proportion of the solute molecules reflected from the microvascular barrier. If a membrane is completely impermeable, no solute molecules pass through, the concentration gradient is maximal, and the solute exerts its full osmotic pressure (i.e., the reflection coefficient ¼ 1). If the membrane is completely permeable to the solute in question, it passes through freely, no concentration difference exists, and no osmotic pressure can be exerted (i.e., the reflection coefficient ¼ 0).

Fluid Therapy with Macromolecular Plasma Volume Expanders Further research showed that fluid flow from vessels differed among tissues depending on the surface area of the capillary beds in the organ and the hydraulic conductance (i.e., the ease of fluid flow) through the microvascular barrier. To account for this variability, the fluid flux equation is modified by the filtration coefficient (Kfc). This term simply implies that fluid flow is equal to a fraction of the effective hydrostatic and osmotic pressure gradients. Jv ¼ Kfc ½ðPc
Pi Þ
sðpp
pi Þ Each different constituent of plasma may differ in its rate of efflux from a vessel depending on such factors as its molecular radius, shape, and charge, and the permeability of the microvascular barrier to the constituent in question. The two major groups of molecules with respect to transvascular fluid flux are termed the solvent phase and the solute phase, and expressions were developed to predict the egress of both major groups of molecules from the microvasculature.83,92,110,115 The solvent phase includes water and those molecules that are not significantly impeded in their passage through the microvascular barrier, whereas the solute flux equation describes the passage of molecules that do not flow freely from the vasculature. The solvent flow equation remains the same as the previous expression of fluid flow except that the filtration coefficient is subdivided into the hydraulic conductance (Lp) and the membrane surface area (S), and the hydrostatic and osmotic gradients are expressed as DP and Dp, respectively: Jv ¼ Lp SðDP
sDpÞ The two major mechanisms of solute flow through the microvascular barrier are convection (i.e., carriage in a bulk flow of fluid) and diffusion (i.e., random motion resulting in net movement of molecules from an area of high concentration to an area of lower concentration).127 An analogy to illustrate the two mechanisms would be a wave breaking on a beach. Some of the sodium molecules in the wave will be moving away from the beach by diffusion; however, the forward convective flow of the wave carries them in the opposite direction. The solute flow equation is the most relevant expression with respect to intravenous therapy with fluids containing macromolecules. It states that the rate of solute flux (Js) is equal to the sum of the convective flow and the diffusional movement. Solute flowðJs Þ ¼ convective flow þ diffusion Convective flow is equal to the product of fluid flow (Jv), the fractional permeability of the membrane (1
s ), and the mean intramembrane solute concentration,C. Diffusion is equal to the product of the solute permeability (P), the surface area of the microvascular barrier (S), and the

649

solute concentration gradient across the membrane (DC). Therefore the expression representing macromolecular flux becomes: Js ¼ Jv ð1
sÞ
C þ PSDC Solute flow ¼ convective flow þ Diffusion At normal lymph flow rates, convection has been estimated to account for approximately 30% of the total flux of albumin into lymph.123 An important point that warrants further emphasis is that the rate of solute efflux is dependent on the rate of solvent efflux. Any condition that increases the rate of fluid flow across a membrane can increase the extravasation of macromolecules. Hence, intravenous fluid therapy with crystalloid or colloid can increase albumin loss into the interstitium.124 These mathematical expressions give the impression of a constant hydrostatic pressure gradient acting across a single membrane of static and uniform conductivity and permeability (homoporous), with filtration opposed by an osmotic pressure resulting from a single impermeant solute, the plasma “protein.” In fact, the hydrostatic pressure and osmotic pressure gradients vary among different tissues and at different levels of the capillary bed within the same tissue.121,156,159 In disease states, the differences among organs may be significant and the clinician must consider the possibility of individual organ edema (e.g., pulmonary, myocardial, or intestinal edema) even if there are no overt signs of a systemic edematous state. The total osmotic gradient is a summation of all the impermeant solutes present within plasma, which all have unique reflection coefficients and efflux rates.156 Furthermore, the surface area of the capillary bed may change depending on precapillary sphincter activity and the permeability of the microvascular barrier and interstitium may also vary physiologically and in disease states.8,71,113,180,181

NORMAL STARLING FORCES AND THE TISSUE SAFETY FACTORS PLASMA COLLOID OSMOTIC PRESSURE Although in popular usage colloid is interpreted most often as referring to a macromolecule that cannot pass through a membrane, the strict definition refers to the dispersion in a gas, liquid, or solid medium of atoms or molecules that resist sedimentation, diffusion, and filtration. This definition is in contradistinction to crystalloids, which are freely diffusible. Oncotic pressure is defined as the osmotic pressure exerted by colloids in solution (hence it is redundant to use the phrase colloid oncotic pressure). Proteins in plasma are truly in solution, but

650

SPECIAL THERAPY

they closely resemble a colloid solution and thus are referred to and treated as such. The osmotic pressure exerted by the naturally occurring colloids in plasma is higher than that calculated for an ideal solution in vitro. One of the main reasons for this discrepancy is that negatively charged proteins (such as albumin, which has a net negative charge of 17 at physiologic pH) retain cations within the intravascular space by electrostatic attraction (termed the Donnan effect).71 These cations contribute to the effective plasma protein osmotic pressure because osmotic pressure is proportional to the number of molecules present rather than their size. Therefore colloid osmotic pressure (COP) is the most correct term when referring to the osmotic pressure exerted by plasma proteins and their associated electrolyte molecules. For comparison, the oncotic pressure exerted by an albumin solution of 7 g/dL is 19.8 mm Hg, whereas the in vivo COP is actually 28 mm Hg, and the total osmotic pressure of all plasma solutes is 5400 mm Hg.71 By virtue of its relatively high concentration in the vascular space, albumin usually accounts for 60% to 70% of the plasma COP with globulins making up the remainder.108,168,176 Interestingly, the variation in COP in dogs may be because of differences in globulin concentration than in albumin concentration.65,108 Red blood cells and platelets do not contribute significantly to plasma COP.118 Serum albumin concentration is determined by the relative rates of synthesis, degradation, and loss from the body and its distribution between the extravascular and interstitial spaces. Albumin synthesis, which is unique to the liver, appears to be regulated, at least in part, by the hepatic plasma COP.53,117,130 Increases of plasma COP independent of albumin concentration, such as in hyperglobulinemia, are associated with decreased serum albumin concentration.18,131,132 The main site of albumin degradation is uncertain, but the reticuloendothelial system has been suggested. Equations have been derived to estimate plasma COP from plasma protein concentrations,108,160 but direct measurement with a colloid osmometer is more accurate.7,28,160,176 COPs measured in normal dogs and cats are given in Table 27-1.44,108,186

INTERSTITIAL COLLOID OSMOTIC PRESSURE Capillaries are permeable to protein, despite the fact that the microvascular barrier greatly restricts macromolecular flux. Of the total quantity of albumin present in the body, 40% is intravascular and 60% is extravascular.133 Furthermore, all of the albumin present in plasma circulates through the interstitium every 24 hours.114 The interstitial COP varies from tissue to tissue depending on such factors as the permeability of the capillary wall to protein, the rate of transvascular solvent flow, the retention of protein in the interstitial matrix, and the rate of lymphatic clearance of protein. The microvascular barrier of skeletal

TABLE 27-1

Species Canine (plasma) Canine (plasma) Canine (whole blood) Feline (plasma) Feline (whole blood)

Colloid Osmotic Pressure in Normal Cats and Dogs Colloid Osmotic Pressure Mean  SD (mm Hg)

Reference Number

20.8  1.8 17.5  3.0 19.9  2.1

185 108 44

19.8  2.4 24.7  3.7

185 44

muscle or subcutaneous tissue is relatively impermeable to protein, whereas the pulmonary capillary endothelium is more permeable with a reflection coefficient to albumin of approximately 0.5 to 0.64.113 Consequently, the normal protein concentration in lymph from skin or skeletal muscle is approximately 50% that of plasma compared with 65% in pulmonary lymph.113 Hyaluronan and its associated cations also may contribute to interstitial COP.5 Because of the volume occupied by the interstitial matrix, interstitial albumin is distributed in a volume that is less than the total interstitial volume. This phenomenon is called the volume exclusion effect, and the “excluded volume” with respect to albumin may be as high as one half to two thirds of the total interstitial volume.13,112,175 Consequently, in a normally hydrated interstitium, much less protein is required to exert a given osmotic pressure, and relatively smaller volumes of extravasated fluid result in greater decrements in interstitial COP. This effect maintains the intravascular-to-extravascular COP gradient in early edema formation. Conversely, when interstitial volume is overexpanded by fluid in edematous states, a dramatic increase occurs in the volume available for albumin sequestration.71 The increase in interstitial COP that occurs with dehydration acts to restrict mobilization of interstitial fluid.76

INTRAVASCULAR HYDROSTATIC PRESSURE Intravascular hydrostatic pressure is the main force that determines fluid egress from the vasculature. It may vary in different tissues and at different levels within each capillary bed. The normal hydrostatic pressure in the capillary bed is controlled by local myogenic, neurogenic, and humoral modulation of the arterial and venous resistances. Precapillary arteriolar constriction may reduce flow, and therefore hydrostatic pressure, through a capillary bed or shunt flow away from that bed, resulting in changes in the total surface area available for

Fluid Therapy with Macromolecular Plasma Volume Expanders transvascular fluid movement. The hydrostatic pressure within a blood vessel at any particular site depends in part on where resistance to flow occurs, with hydrostatic pressures decreasing most across the areas of major resistance. In most tissues, the majority of resistance has been attributed to small arterioles, but experimental studies of the lung suggest that a significant pressure decrease may occur across the capillary bed itself.15,16,143

INTERSTITIAL HYDROSTATIC PRESSURE As with all the other Starling forces, normal interstitial pressure also varies among tissues. Interestingly, in many tissues the resting pressure is slightly negative (subatmospheric), tending to favor rather than oppose fluid filtration from the microvasculature.179 This finding has been postulated to be the result of the molecular structure of the interstitial matrix, such that with normal hydration the biomechanical stresses on the molecules and the repulsion among like electrostatic charges act to expand the interstitium.5 In encapsulated organs, such as the kidney, normal interstitial pressures are positive. Interstitial pressures can also change depending on the functional state of the organ. For example, interstitial pressures in the nonabsorbing intestine are negative to slightly positive, whereas intestinal interstitial pressures are positive in the absorptive state.70 As mentioned before, the molecular structure of the interstitium mechanically opposes distention. Conventionally, it is said that one third of the total body water is found in the extracellular space and that the interstitium constitutes three fourths of the extracellular space. These figures are averages for the whole body, and the relative sizes of the intravascular and interstitial spaces vary among tissues. Tissues vary in their capacity to accommodate interstitial fluid depending on the size of the interstitial space relative to the total volume of the tissue and the nature of the interstitial matrix itself, especially its distensibility. The distensibility of an organ or tissue is termed its compliance, and depending on the nature of the tissue, the compliance of the interstitium may vary widely. Extreme examples would be tendon (which is relatively noncompliant) and loose subcutaneous connective tissue (which is relatively distensible). The accumulation of edema fluid in the peribronchovascular interstitium in canine lungs is likely the result of the higher compliance of this region of the pulmonary interstitium. An extremely important concept related to the interstitial hydrostatic pressure is that of stress relaxation. In a normally hydrated animal, the interstitium in most tissues is relatively noncompliant. Small increases in volume caused by increased fluid extravasation result in large changes in interstitial hydrostatic pressure that act to oppose further extravasation of fluid and increase lymphatic drainage pressure—two of the tissue safety factors described later.72,157 As the interstitium becomes

651

gradually more distended, it continues to oppose distention until a critical point is reached (suggested to correspond to the disordering of the interstitial matrix). Abruptly, the resistance to distention decreases (i.e., compliance increases), and fluid then can accumulate without a corresponding protective increase in interstitial pressure and lymph flow. At this point, the distended interstitium no longer opposes the movement of fluid and protein, resulting in increased extravasation and self-perpetuation of the edemagenic process. Furthermore, the greatly increased interstitial space provides a large volume for protein sequestration.

TISSUE SAFETY FACTORS From the previous discussion, it should be apparent that there are three main homeostatic mechanisms that prevent or limit accumulation of fluid in the interstitium. First, extravasation of fluid into a relatively nondistensible interstitium results in an increased interstitial hydrostatic pressure that opposes further extravasation. Second, after extravasation of low-protein fluid, interstitial COP decreases because of dilution and washout of protein, thereby maintaining or even enhancing the COP gradient between the intravascular space and interstitium. Third, the increased interstitial pressure results in an increased driving pressure for lymphatic drainage. These alterations in Starling forces that act to limit interstitial fluid accumulation have been termed the tissue safety factors.72,157 Their relative importance varies depending on the characteristics of the tissue.5,33 In a tissue that is relatively nondistensible (e.g., tendon), an increase in interstitial pressure may be the most important means by which to counteract filtration. In a tissue with moderate distensibility and with a relatively impermeable microvascular barrier (e.g., skin), the decrease in interstitial COP assumes more importance in protecting against interstitial fluid accumulation. In a distensible tissue that is quite permeable to protein (e.g., lungs), increased lymph flow appears to be the most important safeguard against interstitial edema.183

PHARMACOKINETICS AND PHARMACODYNAMICS OF MACROMOLECULAR PLASMA VOLUME EXPANDERS Transvascular fluid dynamics are extremely complex. The balance of the hydrostatic and osmotic pressure gradients between the intravascular and interstitial fluid compartments forms the basis for microvascular fluid exchange. However, this simple concept is belied by the great heterogeneity in Starling forces and transvascular fluid dynamics that exists among and within tissues in both healthy and diseased states. The relative importance of the different tissue safety factors also varies among

652

SPECIAL THERAPY

tissues, and the potential for self-regulation of transvascular fluid fluxes often is underestimated. When considering fluid therapy with macromolecular volume expanders, a great deal of emphasis has been placed on the manipulation of individual Starling forces (such as intravascular COP) in isolation rather than addressing the system in its entirety. Maintenance of intravascular volume depends on an intricate and dynamic interaction between the intravascular and interstitial Starling forces and the structure and function of the microvascular barrier, interstitium, and lymphatic system. Infusion of intravenous fluids can change all of the Starling forces, modify the permeability of the microvascular barrier, change the volume and composition of the interstitium, and increase lymphatic flow. Furthermore, the magnitude and relative significance of these changes vary among and within tissues. Consequently, it is a gross and potentially dangerous oversimplification to view the body as the homogenous sum of its individual parts when contemplating intravenous fluid therapy. From a clinical standpoint, the differences between the lungs and the systemic circulation are of the utmost importance. For example, in a dog with systemic inflammatory response syndrome and aspiration pneumonia causing pulmonary edema by means of increased microvascular permeability, colloid therapy may be effective in limiting subcutaneous edema at the expense of worsening pulmonary fluid extravasation. Despite this great heterogeneity, the concept that net fluid extravasation depends on the balance between intravascular COP and capillary hydrostatic pressure forms the basis for intravenous colloid therapy.64,73,90,174 By virtue of their larger molecular size, and in the absence of an increase in microvascular permeability, colloid molecules are retained within the vasculature to a greater degree than are crystalloids. Consequently, smaller volumes of colloid result in greater plasma volume expansion compared with crystalloid,51,144,145 and crystalloid is expected to leak into the interstitium to a greater degree than colloid and cause more interstitial expansion or edema.27 This may be beneficial if the animal has an interstitial fluid deficit or deleterious if there is interstitial edema. One hour after infusion of a crystalloid solution, as little as 10% of the infused volume may remain in the intravascular space.145 Some evidence indicates that tissue perfusion is better after volume expansion with colloids than with crystalloids, even when resuscitation is titrated to physiologic endpoints.63 Unfortunately larger colloids may reduce tissue perfusion by increasing plasma viscosity.22 Many factors influence the volume and duration of intravascular expansion associated with artificial colloids, including the species of animal, dose, specific colloid formulation, preinfusion intravascular volume status, and the microvascular permeability. These factors likely explain the great variability in intravascular persistence and volume expansion in published studies.

Artificial colloids are polydisperse; that is, they contain molecules of different molecular weight. In contrast, in a monodisperse colloid such as albumin, molecules are all the same size. The artificial colloids have extremely complex pharmacokinetics in part because of this large range of molecular sizes.84 The smaller molecules pass rapidly into the urine and interstitium, whereas the larger molecules remain in circulation and gradually are hydrolyzed by amylase or removed by the monocyte phagocytic system.161 This initial rapid excretion of small, osmotically active molecules followed by gradual elimination of large molecules results in an exponential decline in intravascular expansion. Manufacturer data sheets can be misleading because they may imply that a major proportion of the volume expansion lasts for 24 to 36 hours. Estimates of the degree of initial plasma volume expansion for hetastarch and dextran 70 vary from 70% to 170% of the infused volume.67,77,87,91,124 This decreases to approximately 50% of the infused volume after 6 hours. Volume expansion with hydroxyethyl starch then declines gradually from 60% to 40% of the infused volume during the next 12 to 18 hours, whereas with dextran 70 it decreases gradually from 40% to 20% of the infused volume.161 In experimental dogs, blood volume was increased by approximately 25% both immediately and 4 hours following infusion of 20mL/kg of both dextran 70 and hetastarch.147 In dogs with hypoalbuminemia of various causes receiving hydroxyethyl starch, COP was not significantly different from baseline 12 hours after infusion.106 In the authors’ experience, the duration of volume expansion with artificial colloids can be even shorter, especially with capillary leak syndromes. This relatively short duration of action and the high cost of artificial colloids have led some authors to question the costeffectiveness of colloid infusions in veterinary patients.173 The duration of action of colloids may be expressed in terms of plasma colloid concentrations, plasma COP measurements, or degree of volume expansion. The initial volume of intravascular expansion is the result of the COP of the infused colloid, which is determined by the number of molecules, not their size. This concept is extremely important because the distribution of molecular weights is narrowed after intravenous infusion.57,58 The smaller molecules that are responsible for a large part of the COP and intravascular volume expansion are extravasated or excreted within hours. The intravascular colloid concentration (i.e., mass per unit volume) is still high due to the large molecules, but the COP is relatively low. COP and degree of volume expansion tend to decrease faster than does the plasma concentration of colloid. Data from an experimental study of euvolemic human volunteers given twice the usual dose of a high molecular weight form of hydroxyethyl starch may therefore have little bearing on the effects of commercially available hydroxyethylstarch in a dog with systemic inflammatory response syndrome in hypodynamic, septic shock.

Fluid Therapy with Macromolecular Plasma Volume Expanders To reiterate, the osmotic effect of macromolecules is because of their number rather than their size. Consequently, if more than 50% of the molecules leak into the interstitium, a net reduction in intravascular volume is likely as water leaves the intravascular space by osmosis along with the colloid. Therefore the difficulty is how to determine the magnitude of increase in permeability (i.e., the size of the “gaps” in the microvascular barrier). Although experimental techniques exist to detect an increase in microvascular permeability,14,25 they are not currently applicable in a clinical setting. A growing body of evidence suggests that hydroxyethyl starches can mitigate increases of microvascular permeability in several capillary leak states.34,97,109,184 The optimal molecular weight for this effect seems to be between 100 and 300 kDa.185 Unfortunately, relatively few products with molecules in this size range are available in the United States. Only 35% of the molecules in one preparation of hetastarch fall within this optimal size range.184 European formulations of hydroxyethyl starch (e.g., Haes-steril, Fresenius Kabi, Bad Homburg, Germany) contain more molecules in the optimal molecular size range.

COLLOID PREPARATIONS The artificial colloids used most commonly worldwide fall into three major groups: the hydroxyethyl starch derivatives, the dextrans, and the gelatins. Availability varies among countries. The hydroxyethyl starches are synthesized by partial hydrolysis of amylopectin (the branched form of plant starch), the dextrans from a macromolecular polysaccharide produced from bacterial fermentation of sucrose, and the gelatins from hydrolysis of bovine collagen followed either by succinylation or linkage to urea. The preparations used most commonly in the United States are hydroxyethyl starch preparations and dextran 70, both of which are available as 6% (6 g/ dL) solutions in 0.9% saline. Several gelatin-based products are available in Europe and Australia (Haemaccel, Intervet/Schering Plough Animal Health, Milton Keynes, UK; Gelofusine, Dechra Veterinary Products, Shrewsbury, UK). Hydroxyethyl starches are manufactured by a complex process and are described using standardized pharmacologic terminology. An understanding of this terminology gives the clinician information about their molecular structure and allows estimation of their likely pharmacokinetics and pharmacodynamics. Amylopectin is a polysaccharide, which along with amylose, forms the plant structural polysaccharide, starch. Amylopectin is very similar in structure to glycogen and contains short chains of a-1,4-linked glucose units linked to other chains by a-1,6-links. Solutions of native starch would be unstable if injected as they are rapidly hydrolyzed by plasma amylases. Chemical modification is required to resist degradation and thereby increase intravascular persistence. This is achieved by substitution of the hydroxyl

653

(-OH) groups on the glucose units with hydroxyethyl (-OCH2CH2OH) groups. The terms “hetastarch,” “pentastarch,” and “tetrastarch” are nonspecific terms used to describe different preparations of hydroxyethyl starches. The term “hetastarch” and the abbreviation “HES” are sometimes used interchangeably, but this should be avoided; hetastarch is just one of the hydroxyethyl starches. The abbreviation, HES, may be correctly used as an umbrella term for all hydroxyethyl starches, which are then subclassified on the basis of their molecular structure. The HES family is most precisely described by reference to their molecular weight and their degree of substitution (e.g., HES 450/0.7 or HES 130/0.4). These characteristics are described more fully later. The C2/ C6 hydroxyethylation ratio is another important pharmacologic characteristic that may be used as a descriptor but it is not routinely included in product descriptions at this time.177

Molecular Weight (MW) In general, the molecules in HES preparations show great polydispersity. The molecules can range in size from a few thousand to a few million Daltons and in any one solution will generally follow a bell-shaped distribution. Hydroxyethyl starches have been arbitrarily divided into high molecular weight (>400 kDa), medium MW (200 to 400 kDa) and low MW (15,000 Da).178,179,180,182 The foundations for this arbitrary classification have been based primarily on their characteristics for dialytic removal.182 The volume of distribution of each of these substances further determines its compartmentalization and accessibility for dialytic removal.23,97,178,182 Hundreds of solutes have demonstrated intrinsic toxicity that mimics or reproduce particular aspects of the uremic syndrome, and thousands of retained solutes have now been demonstrated by mass spectroscopy in uremic subjects.133,177,182 Some retained solutes, such as urea, have minimal inherent toxicity but serve as markers for retention of similar but unidentified solutes with greater clinical significance.49,180 Small water-soluble solutes have demonstrated significance in the expression of uremia because both the morbidity and mortality of uremia can be corrected by their removal with conventional dialysis.179,182 Extensive prospective studies in human patients with kidney failure confirm significant outcome benefits associated with the extent of small-molecular-weight solute removal (i.e., dialysis dose).71,73,106,120,124 However, uremic toxicity is more complex than can be explained by retention of small-molecular-weight solutes and attention has refocused on retention of middle molecules and protein-bound solutes that are poorly removed by dialysis.74,77,133,177,182 There is an empirical link between the appearance of uremic signs and the accumulation of nitrogenous endproducts of protein (amino acid) oxidation. Urea is a small-molecular-weight (60 Da) nitrogenous metabolite whose plasma concentration exceeds that of all other uremic solutes. It contributes minimally to the clinical manifestations of uremia86 but has remained fundamentally associated with the morbidity and outcome of uremic syndrome because of its abundance and its link to the metabolism of dietary and endogenous nitrogen.49,64 No single retention solute (including urea) has been shown to explain the major consequences of the uremic syndrome. Azotemia must be viewed as a marker for the collective appearance of numerous small water soluble compounds, protein carbamylation, redirected metabolic pathways, or other small-molecular-weight solutes coupled to nitrogen metabolism and/or bound to body proteins. The proven correlation of urea removal by hemodialysis with outcome in renal failure has prompted the designation of urea as a surrogate index for all putative

small-molecular-weight retention solutes that remain unidentified or unmeasured.49,82,120 Reduction of urea appearance and the extrarenal removal of urea are used to prescribe the therapy for uremia and to monitor the efficiency and adequacy of these therapies. 50,76,172 This designation is both rational and problematic. Urea is uncharged, present at high concentration, readily detected, and readily diffused across all body fluid compartments and the dialysis membrane. As such, it serves as an excellent solute to document dialyzer performance and whole body clearance of low-molecularweight solutes. However, these unique features and its minimal uremic toxicity question whether it appropriately or accurately reflects the dialytic behavior of other solutes with more profound uremic toxicity and thus may overrepresent removal of these solutes.68,180,181 Dietary protein intake directly influences the generation rate (appearance) of urea, and dialytic clearance and residual renal function influence its removal from the body. Thus serum urea concentration is poised to reflect renal function and dialytic and nutritional adequacy. The individual contributions of urea generation, its removal, and its distribution volume to steady-state serum urea concentration cannot be differentiated by routine urea measurement; however, perturbations of the steady-state induced by dialysis allow kinetic dissection of these independent parameters by formal urea kinetic analysis in patients undergoing hemodialysis (Figure 29-1).48,64,141 The kinetics of urea generation and removal have become the bellwether of the adequacy assessment of dialysis delivery and nutritional status in uremic subjects. 76 The role of urea to function as a global surrogate for uremic toxicity remains controversial in light of the broader recognition and assessment of middle molecules and protein-bound solutes as retained uremia solutes. Similarly, urea assessment provides an incomplete appraisal of dialysis delivery despite its documented utility and evidence as a predictor of dialysis adequacy. Nevertheless, the clinical assessment of urea and urea kinetic modeling remain the recommended and established indices for determining adequacy and delivery of therapeutic hemodialysis.73,76,82,106,120 A variety of manipulation and mathematical models have been developed to characterize the kinetics of urea during dialysis and its relationship to adequacy.* Of these, the fractional clearance of the urea distribution volume (Kt/V) has become the standard measure for the dose of dialysis delivered during a dialysis session.76 From the same analysis, the generation rate of urea (G) can be derived to estimate the protein catabolic rate (PCR) of the patient as a measure of the adequacy of dietary protein intake, and the volume of distribution of urea (V) can be computed to better define hydration and adjustment to the dose (Figure 29-1). *References 47, 48, 64, 141, 153, 162, 172.

Hemodialysis and Extracorporeal Blood Purification

683

70 (61.3) (57.9)

60

Liver

spKt/V ⫽ 1.6

G (3.6 mg/min)

Kd (166 mL/min)

Urea pool V (31 L)

BUN (mg/dl)

50 40

TAC⫽ 35.8 mg/dl

30 20 (14.0)

Kr (0.4 mL/min) 10 Kidney

Td

Ti

0 0

Single-pool, fixed-volume model

A

B

24

48

72

96

Time (h)

Fig. 29-1 A, Single-pool, fixed-volume kinetic model of the urea metabolism and representative modeled kinetic parameters determined in a 33-kg dog on intermittent maintenance hemodialysis consuming approximately 56 g of dietary protein. Urea is generated in the liver as the major end product of protein metabolism. The urea generation rate, G (mg of urea/min), determines the accumulation of urea in the urea pool with a volume, V (L). Its removal from the urea pool is determined by the continuous residual renal clearance, Kr (mL/min), and intermittently by hemodialysis via the urea clearance of the dialyzer, Kd (mL/min). B, Graphic illustration of a three-point BUN profile (before and after hemodialysis values in parentheses) that can be fitted to the single-pool model in the right panel. With direct measurement of renal and dialyzer urea clearances (Kr, Appendix, Equation 6 and Kd, Appendix, Equation 5, respectively), kinetic modeling allows computation of the urea generation rate (G, Appendix, Equation 9), the urea distribution volume (V, Appendix, Equation 10), and the time-average concentration of BUN (TAC, Appendix, Equation 1). The dose of dialysis expressed as the fractional clearance of the urea distribution volume using single-pool kinetics (spKt/V, Appendix, Equation 11) also can be calculated. Td is the duration of dialysis, and Ti is the duration of the interdialytic interval. AUC is the area under the BUN versus the time curve and can be estimated using a trapezoidal method or ideally calculated by fitting the changes in BUN to the kinetic model.

DIALYSIS ADEQUACY The optimal outcome for animals with acute renal failure is survival until renal function has recovered, but secondary goals may vary qualitatively depending on the nature of the underlying disease. An optimal outcome additionally should promote physiologic and metabolic stability to facilitate recovery and promote an acceptable quality of life while minimizing secondary injury to the recovering kidneys or other organs (heart, lungs, gut, brain). As an outcome, survival is multifactorial and predicated on the diversity of the underlying cause and comorbidities, in addition to the delivered therapy. As such, outcome assessment by survival alone may be disassociated from recovery of renal function or adequate delivery of dialysis.34,149 Consequently, more sensitive and predictive outcome measures should be considered for assessment of dialysis adequacy, including recovery of renal function, improvement of the systemic manifestations of uremia, and reduction of complications attending uremia or its therapy.19 Survival is the optimal outcome for animals with endstage renal disease because there is no prospect for

recovery of renal function. Realistic outcomes for these animals that are treated with hemodialysis vary depending on age, chronicity of the disease, comorbidities, and residual renal function. Appropriate markers for dialysis adequacy include length of survival, owner perceived quality of life (e.g., activity, social interaction, appetite), elimination of uremic symptomatology (hypertension, hyperphosphatemia, anemia), nutritional adequacy, and elimination of dialysis-associated complications. For both acute and chronic dialysis, survival is a difficult outcome parameter to correlate specifically to dialytic interventions. Yet, despite these constraints, the kinetically modeled dose of dialysis (Kt/V) has been shown to correlate independently with survival as an outcome in humans undergoing maintenance hemodialysis,73,106,120,124 and it is likely to be linked similarly to the success of dialysis in animals. The empirical use of proven standards of dialysis adequacy and clinical experience in human patients are useful first approximations for appropriate veterinary standards of dialysis adequacy until evidence-based standards are determined in animals.34,59

SPECIAL THERAPY

684

QUANTIFICATION OF HEMODIALYSIS DELIVERY AND UREA KINETIC MODELING

80

80 Urea reduction ratio (%)

The delivery (dose) and efficacy of hemodialysis can be expressed in a variety of ways with differing degrees of complexity and utility. Predialysis and immediate postdialysis concentrations of routine serum chemistries (e.g., urea nitrogen, creatinine, phosphorus, bicarbonate, electrolytes) are the simplest expression of efficacy and can be applied similarly to their use in conventional therapy (Figures 29-1 and 29-2).48,105,120,172,183 Although useful to document the instantaneous outcome of the treatment, these assessments do not facilitate the uniform prescription of dialysis to animals of differing size or metabolic status. Nor do they help to clarify the excretory impact of the therapy beyond the intradialytic interval. The predialysis and postdialysis concentrations of plasma urea and creatinine can be expressed further as reduction ratios (URR and CrRR, respectively), which are used routinely to evaluate the intensity of therapy (Appendix, Equations 2 and 3).* Urea reduction ratio (URR) can be expressed either as the fractional or percent in change in urea during the treatment and is the most universally used predictor of adequacy for dialysis sessions in animals (Figures 29-3 and 29-4; Table 29-2). The average dialysis treatment in cats and most dogs will achieve a URR approaching 95%. This high level of treatment intensity

100

60

40

20 N ⫽ 517 HD treatments in 72 dogs (2002–2004) 0 0

1

3 2 Blood processed (L/kg BW)

4

5

Figure 29-3 Prediction of dialysis treatment intensity (urea reduction ratio [URR]) as a function of the volume of blood processed in 72 dogs undergoing hemodialysis. URR was computed from predialysis and postdialysis BUN concentration (Appendix, Equation 2). The volume of blood processed (Qb
Td) was indexed to body weight to compare dogs of different sizes. The relation (modeled as URR ¼ 1ea(Qb
Td/BW), r2 ¼ 0.69) is displayed as a thick solid line with its 95% confidence interval (CI; thin lines). To achieve a low-efficiency treatment with URR equal to 30%, a volume of 0.3 L of blood/kg of body weight must be processed during the treatment (e.g., 6 L in a 20-kg dog). The variation in resulting URR (95% CI, 15% to 45%) underscores the necessity for close monitoring of the delivered (and not prescribed) dose of dialysis for each treatment. Similarly, a URR of 80% is obtained with 1.4 L (95% CI, 0.9 to 2.9) of blood processed per kilogram of body weight (e.g., 28 L in a 20-kg dog).

Kr⫽0.4 mL/min Kr⫽3.0 mL/min 100 Kr⫽4.5 mL/min

40

20

0 0

2 4 Time post-dialysis (days)

6

8

Figure 29-2 Changes in BUN during and after 5 hours of hemodialysis treatments in a 33-kg dog presented for acute antifreeze poisoning at varying degrees of residual urea clearance. The predialysis and immediate postdialysis BUN concentrations reflect a simple assessment of treatment intensity (dose). The eKt/V (approximately 2.9 per session) for the dialysis treatments was identical for each level of urea clearance, yet the rate of increase and the equilibrated BUN concentration after stopping dialysis increased inversely with residual urea clearance.

*References 34, 38, 40, 96, 105, 120, 156.

Urea reduction ratio (%)

BUN (mg/dl)

60

80

60

40

20 N ⫽ 271 HD treatments in 66 cats (2000–2004) 0 0

1 2 Blood processed (L/kg BW)

3

Figure 29-4 Prediction of dialysis treatment intensity (urea reduction ratio [URR]) as a function of the volume of blood processed in 66 cats undergoing hemodialysis. Other conventions are as described for Figure 29-4. The closer correlation (r2 ¼ 0.85) between volume of blood processed and URR in cats compared with dogs is probably because of the more uniform body shape and sizes in this species.

Hemodialysis and Extracorporeal Blood Purification is due to a combination of relative long treatment time and relatively smaller patient size (5 to 40 kg) compared with humans in which a URR target is 60% to 65%. In very large animals (50 to 70 kg), this degree of treatment intensity is often difficult to obtain, and a URR of 80% to 85% is typical. Reduction ratios are convenient for clinical assessment but do not account for all aspects of solute transfer. Uremic toxicity and patient well-being are not predicted necessarily by the highest or lowest concentration or the intermittent change of specific retained uremia solutes.65 The integrated exposure to uremia toxins over time is considered by some a more realistic determinant of well-being and therapeutic adequacy.63,104,106,115,143 For urea, this is expressed as the time-averaged concentration (TACurea), which is calculated as the area under the BUN profile (curve) divided by the duration of the dialysis cycle (Figure 29-1; Appendix, Equation 1). TACurea has been shown to predict morbidity and outcome in human patients undergoing hemodialysis and provides an integrated overview of urea dynamics (and presumably uremia toxicity) during a single or over multiple dialysis cycles. It has been highly predictive of dialysis adequacy and outcome for survival but remains nonspecific and fails to distinguish the multifactorial contributions to urea metabolism during the dialysis cycle, including dialysis dose, urea generation, nutritional adequacy, residual clearance, and distribution volume.48,98,104,107,120 At face value, neither predialysis BUN nor TACurea are adequate surrogates to characterize the adequacy of dialytic therapy or urea metabolism. An animal with a low predialysis BUN can represent effective dialysis (high dialysis delivery), recovering renal function (increased residual renal clearance), inadequate nutrition (low urea generation rate or PCR), or volume overload (expanded urea distribution volume). Conversely, under dialysis, worsening renal function, high catabolic rate, or volume contraction can all be reflected by a high predialysis BUN. The dose of dialysis delivered to the patient can be defined alternatively by the amount of clearance provided by the hemodialyzer during the dialysis session. Using the measured (instantaneous) clearance of the dialyzer for urea (Kd, mL/min) and the dialysis session length (Td, minutes), the dose of dialysis can be defined as Kd
Td or the volume of the patient cleared of urea (depurated volume) during the treatment (mL). This value can be indexed further to the total reservoir or distribution volume of urea in the patient (V, mL) to compare treatment efficacy among patients of different body sizes as V is equal to the patient’s total body water. This expression is analogous to conventional dosing of drugs as milligrams per kilogram of body weight. The value obtained with this kinetic expression, Kt/V, (Appendix, Equation 11) is unitless and represents the fractional clearance of the urea distribution volume.48,50,64,157,172

685

Kt/V has become the international reference for dialysis dosing and delivery.76 This assessment of dialysis dose and intensity advances our understanding of the delivery of dialysis during individual treatments but requires the additional measurement of Kd (Appendix, Equation 5) and the imprecise estimation of V from the patients weight and hydration status. These predictions of dialysis dose are limited by simplifying assumptions regarding urea generation, fluid removal, and solute transference during the session, which require more extensive evaluation. A more fundamental understanding and precise description of solute dynamics during dialysis can be derived from kinetic modeling of the intradialytic and interdialytic changes in BUN similar to pharmacokinetic profiles used to describe drug metabolism.48,64,141 Urea kinetic modeling (UKM) is fundamental to understanding the prescription, monitoring, and quality assurance of hemodialysis procedures and must be familiar to all practitioners of this therapeutic modality. It dissects the mutually independent influences of dialysis, residual renal function, nutrition, catabolism, and distribution volume on the intermittent perturbations in urea concentration during and between the dialysis sessions. This kinetic approach to urea metabolism also yields the fractional clearance of urea (Kt/V) as a measure of dose in addition to urea generation rate (G), protein catabolic rate (PCR), and the distribution volume of urea (V) that are ionic dialysance otherwise beyond clinical assessment. The simplest kinetic assessment of urea during intermittent hemodialysis is represented by a single-pool, fixed-volume model, in which the entire volume of distribution of urea (i.e., total body water) is presumed to behave as a single pool with no change in volume or urea input during the treatments (see Figure 29-1). 47,48,141– 143 In this simplified model, the only kinetic variable is total urea clearance (K), which represents the sum of residual renal clearance (Kr) and the clearance of the dialyzer (Kd) (see Figure 29-1; Appendix, Equations 5 and 6).186 The absolute removal of urea in this system will be reflected by the change in urea concentration at any time during dialysis such that: Ct ¼ C0 eKt=V ,

ð1Þ

where Ct is the urea concentration at time ¼ t; C0 is the predialysis urea concentration at t ¼ 0; K is the total urea clearance; and V is the volume of urea distribution. Rearrangement of Equation 1 provides Equation 2 for single-pool (sp) conditions, sp Kt=V

¼ lnðC0 =Ct Þ:

ð2Þ

Equation 2 is the fundamental kinetic expression for the fractional clearance of urea (dialysis dose) during a single dialysis session. In the simplified single-pool model,

686

SPECIAL THERAPY

the kinetic prediction of dialysis dose can be derived very simply from the measured predialysis and postdialysis BUN concentrations. It must be emphasized, that this expression represents a gross oversimplification of the events and kinetic variables during therapeutic hemodialysis and should be used only to provide a rough estimate of the dialysis dose. During a therapeutic dialysis session, the relationships between G, V, and K (illustrated in Figure 29-1) are more complex, highly interdependent, and cannot be described mathematically by a single simple relationship. Mathematical description of each variable, however, can be defined in terms of the other two with formal urea kinetic modeling (Appendix, Equations 8 through 10). When one of the variables (G, V, or K) is known, the others can be resolved by simultaneous iterative solution of the equations to yield a unique solution for the unknowns when residual renal clearance (Kr), instantaneous dialyzer clearance (Kd), ultrafiltration volume, and the measured changes in BUN during and after the treatment are known.47,48,141–143 These computations are performed easily with commercially available software or can be programmed into routine spreadsheet applications. The simplified single-pool, fixed-volume model presumes conditions not generally valid in therapeutic dialysis sessions and loses accuracy if total body water (TBW) changes during or between treatments. The model also loses accuracy during high-efficiency treatments of short duration, when the urea distribution does not behave as a single homogenous compartment. Delayed diffusion from the intracellular compartment or variations in diffusion among discrete fluid compartments (e.g., skin, muscle, gut) with different perfusion and transference characteristics creates a solute disequilibrium between compartments that promotes a postdialysis rebound of urea that is not predicted by immediate postdialysis blood sampling.47,55,126,144 Deviations in the assumptions for single-pool, fixed-volume kinetics can be minimized by measurement of the postdialysis urea at 45 to 60 minutes after the end of the dialysis treatment rather than immediately postdialysis. By this time, intercompartmental shifts (or rebound) have reestablished solute equilibrium, and the plasma concentration reflects the equilibrated concentration of urea across all body compartments.47,151,163As stated previously, therapeutic hemodialysis deviates considerably from the single-compartment model illustrated in Figure 29-1. Retained solutes, including urea, can be distributed in multiple compartments, which are partially secluded from the dialyzer by delayed transfer or differences in regional perfusion. Most dialysis treatments also require ultrafiltration, and urea generation proceeds throughout the session which further deviate the serum urea concentration from single-pool predictions. These deviations from single-pool, fixed volume assumptions can be improved to provide greater accuracy to urea

kinetic analysis by using more mathematically complex double-pool142 or noncompartmental kinetic modeling methods (Figure 29-5). The double-pool variable volume kinetic model accounts for intercompartmental solute diffusion during and after completion of hemodialysis, and dpKt/V is regarded as the standard for dialysis dose. Optionally, correction algorithms that account for these compartmental deviations have been applied to single-pool assessments using additional blood sampling and appropriate software in human patients.41,48,69 These correction formulas minimize many of the limitations of single-pool estimates but have not been validated in animals. More accurate predictions of dialysis dose also can be obtained using single-pool kinetic calculations by incorporating an equilibrated BUN obtained 45 to 60 minutes after cessation of the treatment as the enddialysis value. Use of the equilibrated BUN in the single-pool calculations yields eKt/V as a measured dialysis dose that closely approximates the dpKt/V and better reflects whole patient clearance. Both the eKt/V and the dpKt/V assessments of dialysis dose will be lower than dose predicted as the spKt/V. Online measurement of these kinetic determinants of dialyzer performance and dialysis dose can be computed in real-time with ionic dialysance techniques that advance

Liver G

Kd

Central compartment V1

Kc

Peripheral compartment V2

Kr Kidney Double-pool variable-volume model

Figure 29-5 Graphic illustration of the double-pool variablevolume kinetic model of the urea metabolism during high efficiency hemodialysis. In this model, the urea generation rate (G), the renal clearance (Kr), and the dialyzer clearance (Kd) are the major determinants of urea content in the central compartment (volume V1). An additional peripheral compartment (volume V2) continuously exchanges solutes and water with the central pool. The bidirectional rate constant for urea transference between the two pools is indicated by Kc. When Kc ¼ 1, urea diffuses freely between the compartments and the system reverts to a single-pool model. A lower Kc implies a slower diffusional component into and out of the peripheral compartment. If the peripheral compartment remains unaccounted for, single-pool kinetic modeling results in a lower apparent V, a more rapid decrease of the urea concentration in the central pool, a greater postdialysis rebound, and overestimation of the dose of dialysis, Kt/V. Anatomically, the two compartments can represent the extracellular and intracellular spaces, respectively, or body areas with different perfusion characteristics.

350

600

300

500

250

400

200

300

150

200

100

100

50 0:00

2:00

3:00 Time

4:00 5:00 h:min

350

600

300

500

250

400

200

300

150

200

100

100

50 0:00

B

0 1:00

Blood flow (Qb, mL/min)

A

Ionic dialysance (Kd-ionic, mL/min)

Hemodialysis and Extracorporeal Blood Purification

0 1:00

2:00

3:00 Time

4:00 5:00 h:min

Figure 29-6 Screen shots of the ionic dialysance display of the Gambro Phoenix illustrating the ionic dialysance (solid line, left axis) and blood flow (dashed line, right axis) throughout a dialysis session. A, Demonstrates constant dialyzer performance and extraction ratio during the treatment with a Kd-ionic of approximately 195 mL/ min at a Qb of 300 mL/min (extraction ratio, 0.65). B, Illustrates a marked and progressive decrease in Kd-ionic after 1.5 hours of treatment associated with extensive clotting of the dialyzer necessitating termination of the treatment.

the monitoring of individual dialysis sessions and ensure adequate dialysis delivery (Figure 29-6). Automated, bloodless kinetic modeling systems using ionic clearance are available on many modern delivery systems and provide kinetic assessments for each dialysis treatment as an alternative to blood-based modeling techniques.24,67,108,114,118 Dialysance of a dialyzer is a measure of solute mass transfer across the dialysis membrane when the solute is present in both the blood and dialysate. Ionic dialysance is a kinetic assessment of the transfer characteristics of the ionic solutes in the blood and dialysate. The collective concentration ionic solutes in solution can be measured by the conductivity of the solution, which is proportional to the electric current conducted through the solution. The conductivity of both plasma and the dialysate is influenced primarily by the concentration of sodium and chloride and will change with perturbations of these solutes.67,114 The clearance of a solute by the dialyzer is equal to its dialysance when the solute is present only in the blood and is absent in the dialysate. The collective dialysance of small-molecularweight ions (such as sodium) is considered equivalent to the dialysance of urea, and consequently ionic dialysance can be used as a reasonable surrogate for the dialysance of urea. In conventional single-pass hemodialysis, circuits in which the dialysate contains no urea, urea dialysance becomes equal to urea clearance, and ionic

687

dialysance becomes an acceptable predictor of the urea clearance of the dialyzer, Kd-urea. Analogous to measurement of blood-based dialyzer clearance, ionic dialysance is computed from measurements of dialysate conductivity (concentration of ionic solutes) at the inlet and outlet ports of the dialyzer in response to transient changes in inlet dialysate conductivity and the instantaneous dialysate and blood flow rates.* When ionic dialysance is programmed sequentially during the dialysis treatment, serial updates of the instantaneous clearance (Kd-ionic) of the dialyzer can be monitored, and the depurated volume for treatment (Kd-ionic
t) is predicted at the end of the session. The ionicKt/V, as a surrogate for spKt/V, provided when the ionic dialysance is indexed to urea distribution volume, V. The availability and simplicity of ionic dialysance to predict dialysis delivery at every treatment should promote a better understanding of the kinetics of dialytic therapy and the efficacy of dialysis prescriptions. Sudden or progressive decreases of Kd-ionic during the treatment can alert possible clotting in the dialyzer or development of access recirculation that may compromise the adequacy of the treatment. It is also possible to make interim projections of the ionicKt/V for the session to ensure the treatment targets will be met by the end of the scheduled session time. If therapeutic targets will not be met under current circumstances, adjustments to treatment time, blood flow, and dialysate flow, access repositioning, or dialyzer exchange can be initiated to modify the forecast treatment to ensure adequacy.29 Routine animal hemodialysis is provided intermittently three times weekly based on human convention. As for humans, this schedule represents a compromise between clinical benefits, time constraints, and financial burden. However, recent experience in human patients with daily dialysis schedules has demonstrated marked theoretical and clinical benefits to the increased dialysis frequency.{ Because diffusion is a first order process, dialysis becomes more efficient as the frequency of dialysis increases.28,46,66,99 Critical analysis of varying dialysis schedules has shown the total weekly dose calculated as the sum of the individual treatments is not equivalent among dialysis schedules with differing frequencies. Daily treatment schedules have equivalent clinical outcomes to traditional three times a week hemodialysis schedules even when delivered at a lower total weekly dose. For example, six treatments per week at a spKt/V of 1.0 per treatment are more efficient than three conventional treatments per week with a spKt/V of 2.0 per treatment. To reconcile these differences, the concept of standard Kt/V (stdKt/V) has been proposed to compensate for the differences in efficiency when comparing schedules with different intermittence.50,64–6699 Standard Kt/V is *References 53, 54, 67, 92, 128, 130. { References 28, 46, 63, 66, 72, 100, 171, 174.

688

SPECIAL THERAPY

an hypothetical continuous urea clearance that would achieve a constant blood urea concentration identical to the average predialysis urea concentration for all intermittent treatments provided during the week. This theoretical concept allows comparisons among dialysis schedules with differing dialysis times and intervals, including the extreme case of continuous therapy. A dialysis schedule with three 4-hour treatments per week with a spKt/V of 2.0 per treatment is equivalent to a stdKt/V of 2.7. Increasing the schedule to six 2-hour treatments per week (spKt/V, 1.0 per treatment) with the same total 12 hours of weekly dialysis substantially increases the amount of dialysis delivered to the equivalent stdKt/Vof 3.9 (Appendix, Equation 12). Stated differently, a three times a week, 240-minute treatment schedule (stdKt/V, 2.7) requiring 12 hours of treatment could be provided with equivalent efficacy by considerably shorter treatments of 70 minutes per session if provided six times weekly for a total weekly dialysis time of 7 hours. Although reduction of the individual treatment time is possible according to this analogy and for illustrative purposes, this recommendation would not be clinically prudent.52,55,66,113 Conversely, decreasing the frequency of dialysis to two treatments per week would require extension of each treatment to almost 24 hours to achieve an equivalent stdKt/V. These quantitative predictions illustrate the marked benefits to increased frequency of therapy and are in accordance with recent clinical observations, suggesting it is difficult to compensate for decreased frequency of therapy with longer treatment times.47,50,172 As an alternative to sdtKt/V for comparing the equivalency of intermittent and continuous therapies, including residual renal function, the intermittent kinetics of hemodialysis can be converted to a continuous equivalent clearance (EKR).25,28,50,183,186 This concept is more intuitive for most clinicians as the relative contribution of dialysis can be compared directly with residual renal function and with other intermittent or continuous dialytic therapies (Appendix, Equation 7). Total patient clearance (renal clearance, Kr, and dialyzer clearance, EKR) is expressed in the familiar term (milliliters per minute) of clearance, similar to the glomerular filtration rate, and the resulting total clearance can be used to predict the expected uremic morbidity, similar to patients with earlier stages of chronic kidney disease. A prerequisite for the validity of most urea kinetic modeling algorithms is the presumption of steady-state urea metabolism (i.e., constant food intake (quality and quantity), constant endogenous nitrogen metabolism and catabolism, stable body weight, and a regular dialysis schedule).48 These conditions rarely exist for most veterinary applications of hemodialysis that are prescribed for acute kidney failure; however, classic double-pool, equilibrated, and EKR analyses appear valid under these conditions in human patients if careful attention is paid to the accuracy of all input variables.26,44,87

The rationale to scale dialysis dose to the nebulous index (V) that cannot be readily measured has kinetic justification and historical acceptance. The first order kinetics of urea removal by dialysis proceeds with an elimination constant equal to Kd/V and is a measure of the intensity of the treatment. Even though V is not measured directly, it is derived mathematically to yield the expression, Kt/V, with kinetic modeling. Recently, however, the universality of scaling dialysis dose to the urea distribution volume has been questioned in human patients as the relative distribution volume varies independently of body size, between genders, and in patients of differing body composition.43 Consequently, scaling dialysis dose to V may promote under treatment in some individuals and relative overtreatment in others. The comparative significance of this issue has not been addressed in animals, but it is likely the diversity of size, species, and breed, in addition to gender, in animal patients that could impose even greater variance in the relative urea distribution volume than seen in humans. The effect of dose of dialysis on outcome has been demonstrated in humans with end-stage chronic kidney disease in several large-scale clinical studies.* The dose of dialysis that is adequate to manage dogs and cats with either acute or chronic kidney failure needs to be established using appropriate tools for treatment quantification. However, until these parameters are established, routine application of UKM extends therapeutic insights of dialysis delivery far beyond reliance on routine chemistry tests and clearly benefits the assessment and clinical management of uremic animals. Kinetic parameters and quantitation of dialysis delivery are important tools for quality assurance of dialytic therapy in animals; however, they are not therapeutic goals per se.186 The provision of a yet-to-be-defined minimal dose of dialysis is only one of the requirements of therapeutic adequacy, and management of uremia necessitates an individually tailored global approach to the animal.

USE OF HEMODIALYSIS TO CORRECT UREMIA The major application of dialytic therapy is the transient elimination of innumerable and unspecified solutes and fluid retained during renal failure that would otherwise be cleared by healthy kidneys. The benefits of intermittent dialysis are transient, and with cessation of dialysis, the concentrations of urea and all retained uremia solutes with continued generation increase immediately until a new steady state is achieved or until the next dialysis session (Figures 29-1 and 29-3). It is firmly established that uremia is associated with retention of a myriad of lowmolecular-weight solutes that are effectively predicted by the blood urea concentration; dialytic removal of these *References 28, 56, 71, 73, 106, 120, 124, 171.

Hemodialysis and Extracorporeal Blood Purification solutes to minimize the time-averaged urea concentrations mitigates the associated morbidity and mortality of uremia but does not resolve all uremic symptomatology.* It is equally established that additional classes of retention solutes including protein-bound, lowmolecular-weight solutes, secluded solutes, and so called middle molecules with a molecular weight between 500 Da and 60,000 Da are poorly dialyzed by conventional high-flux diffusive and hemofiltration techniques, limiting the efficacy of extracorporeal techniques.{ The diffusive removal of urea and small-molecular-weight solutes is exceptionally efficient in animals because of their small size (volume) relative to the surface area and clearance capabilities of the hemodialyzer. Theoretically, these solute and the fluid abnormalities attending uremia could be corrected temporarily during a single hemodialysis session, but clinical sequelae associated with abrupt excursions in the solute and fluid content of the patient limit the rate and magnitude that they can be altered. The change in solute concentration (e.g., urea) during dialysis is influenced by the size of the animal and the interactive parameters defining the dialysis prescription (see Appendix, Equation 8). The intensity of dialysis can be adjusted by altering blood flow rate (Qb), dialysate flow rate (Qd), clearance of the hemodialyzer (Kd), rate of ultrafiltration (UF), or length of the dialysis session (Td) to accommodate the size and therapeutic needs of the animal. After dialysis, BUN (and other retained uremia solutes) increases in proportion to urea generation from dietary nitrogen and endogenous protein catabolism (G) and inversely with residual renal function (Kr) (see Figures 29-1 and 29-2). Higher dietary protein intake, increased catabolism, and lower residual renal function will produce a steeper increase and higher steady-state concentration of urea after dialysis unless interrupted by an intervening dialysis treatment before achieving a steady state (Figure 29-2). The peak predialysis urea, time-averaged urea concentrations, and the exposure to urea and other uremic toxins will be lower the more frequently and effectively a patient is dialyzed.46,48,50,64,186 The hemodialysis session is defined by the dialysis prescription, which is formulated interactively with consideration of the physical and clinical condition of the patient and the alterations of body fluid volume and composition subject to dialytic correction. The prescription must accommodate the physiologic, hematologic, and biochemical status of the patient before dialysis and target the desired modifications at the end of the session (Box 29-2). Patient assessment includes (1) species, breed, weight; (2) degree of azotemia; (3) hemodynamic stability and predisposition to hypotension and hypovolemia (i.e., body weight, estimated blood volume, blood pressure, volemic status); (4) hematocrit and total *References 48, 49, 56, 57, 106, 120. { References 56, 113, 133, 181, 182, 185.

BOX 29-2

689

Clinical Considerations Influencing the Hemodialysis Prescription

1. Patient characteristics (species, size, age, body condition) 2. Severity of the azotemia and retained uremic toxins 3. Electrolyte and mineral disorders: sodium, potassium, chloride, bicarbonate calcium, magnesium, and phosphate 4. Acid-base imbalances and depleted or deficient solutes: bicarbonate, calcium, glucose 5. Exogenous intoxications (e.g., ethylene glycol) 6. Hydration status and fluid balance 7. Physiologic disturbances: blood pressure, body temperature, oxygenation, change in body weight, mental state 8. Coagulation status 9. Medications, surgical history, and comorbid clinical conditions 10. Dialysis treatment history

plasma solids; (5) electrolyte and acid-base abnormalities; (6) oxygenation capacity; and (7) bleeding potential. The prescription is individualized for each patient and every dialysis session by selecting dialytic options that best achieve the solute removal and ultrafiltration goals of the session without predisposing therapeutic risk. Specific factors regulating these processes are prescribed independently and are outlined in Box 29-3. Hemodialysis prescriptions for animals have been derived empirically as consensus-based guidelines for a diverse array of animal types and clinical conditions. There has been little validation or standardization of dialysis therapy based on outcome assessment. However, animal dialysis has advanced over the past 40 years, and dialysis prescriptions are based on a solid understanding of the physical and physiological principles governing dialysis and clinical aspects of uremia.

HEMODIALYSIS PRESCRIPTION FOR ACUTE KIDNEY INJURY (AKI) The rapid accumulation of retained solutes in acute uremia intensifies expression of the clinical signs and metabolic disturbances compared with the uremia of chronic kidney disease. Hemodialysis prescriptions are prioritized to resolve hyperkalemia, profound azotemia, fluid imbalance, metabolic acidosis, and persistent nephrotoxins and to accommodate ongoing therapies (e.g., parenteral feeding). The therapeutic efficiency of hemodialysis must be applied judiciously to prevent overtreatment when the risks of dialysis disequilibrium syndrome, hypovolemia, hypotension, and bleeding are high. Consequently,

690

SPECIAL THERAPY

BOX 29-3

Components of the Hemodialysis Prescription

1. Selection of the hemodialyzer (surface area, bundle volume, solute and ultrafiltration characteristics, hemocompatibility, and biocompatibility) 2. Selection of extracorporeal circuit and priming solution 3. Blood flow rate (Qb) 4. Dialysis time (Td) and scheduled bypass time 5. Dialysate composition and/or modeling 6. Dialysate flow rate and direction (Qd) 7. Treatment schedule 8. Access connection (“single needle,” reversed direction) 9. Anticoagulation (anticoagulant, target ACT, protocol) 10. Ultrafiltration (volume target, rate) 11. Ancillary medications 12. Monitoring schedule 13. Rinse back (solution, volume, air) 14. Catheter locking solution 15. Posttreatment (medications, monitoring)

dialysis goals for initial treatments in animals with AKI differ considerably from the goals and prescription for later dialysis treatments.

Hemodialyzers Selection of the hemodialyzer is based initially on its contribution to the extracorporeal volume and secondarily on its diffusive, convective, and biocompatibility properties according to guidelines in Table 29-1. The smallest neonatal hemodialyzer currently available has a 0.3 m2 surface area and a 28-mL blood volume compartment (F3, Fresenius Medical Care, Waltham, Mass.). For cats and dogs weighing less than 6 kg, a dialyzer with a surface area between 0.2 and 0.4 m2 and a priming volume less than 30 mL generally is tolerated. A synthetic dialyzer (neonatal or pediatric) with a surface area between 0.4 and 0.8 m2 and a priming volume less than 45 mL is appropriate for use in dogs weighing between 6 and 12 kg of body weight. Dialyzers with surface areas up to 1.5 m2 and priming volumes up to 80 mL can be used on dogs between 12 and 20 kg of body weight. Larger dialyzers with surface areas greater than 2.0 m2 and priming volumes greater than 100 mL can be used in dogs weighing more than 30 kg. A dialyzer with a smaller surface area (0.3 to 0.5 m2) than recommended may be chosen preferentially in dogs of all sizes for initial hemodialysis treatments when the BUN concentration is greater than 200 mg/dL to reduce the intensity of the treatment and risk of dialysis disequilibrium. Solute removal follows first order kinetics, and

animals with marked azotemia (BUN >250 mg/dL) will experience quantitatively greater urea removal per unit of time and blood flow than those with lesser degrees of azotemia. For patients with severe azotemia a low-efficiency dialyzer with a lower urea clearance may be more appropriate and safer than use of a high-efficiency device. A smaller dialyzer can be selected also for initial treatments with reduced blood flow rates to limit the resident time of blood in the dialyzer to minimize clotting. At a blood flow rate of 20 mL/min, the resident time of blood in a 28-mL dialyzer is only 1.4 minutes compared with 9 minutes in a 1.5-m2 dialyzer with a blood volume of 180 mL.

Treatment Intensity Initial dialysis treatments are prescribed to be less intensive (slower blood flow rate, smaller dialyzer surface area, and possibly shorter treatment time) than those prescribed for subsequent treatments. At slow blood flow rates, urea extraction across the dialyzer approaches 100%, and urea clearance (Kd-urea, in milliliters per minute) is approximately equal to extracorporeal blood flow (Qb, in milliliters per minute). When high-efficiency and high-flux dialyzers are used, Kd-urea increases quantitatively with Qb until blood flow exceeds 200 mL/min.47 At blood flow rates above 200 mL/min, the relationship flattens as urea clearance is influenced by membrane characteristics and dialysate flow in addition to Qb.47 At blood flow rates greater than 300 mL/min, dialyzer performance is influenced minimally by increased single-pass flow, but total solute removal during the treatment will increase as a function of the cumulative flow through the dialyzer. The total volume of blood passed through the dialyzer during the treatment (Qb
Td) has been established as a reasonable predictor of the intensity of the treatment as estimated by the URR (Figures 29-3 and 29-4).34,59,96 This relationship can be used as an operational parameter to guide the prescription and delivery of dialysis to the target URR for differing severities of uremia and phases of management (Table 29-2).

Dialysis Time Once the target URR is defined for the treatment, the approximate volume of blood requiring dialytic processing to achieve the goal can be determined (Figures 29-3 and 29-4). From this volume (Qb
t), appropriate combinations of blood flow rate (Qb) and dialysis time (t) can be derived. For patients with moderate to severe azotemia, a long dialysis session time (slow Qb) is preferable to a short session time (fast Qb) that yields the same volume of processed blood and prescribed URR. Prescription of a dialysis session time less than 180 minutes could promote use of inappropriate blood flow rates that induce rapid changes in BUN and life-threatening dialysis complications. Short treatments usually cause

Hemodialysis and Extracorporeal Blood Purification

TABLE 29-1

691

Recommended Extracorporeal Volumes and Characteristics of Hemodialyzers Used for Hemodialysis in Dogs and Cats Body Weight

Total Dialyzer Extracorporeal Volume Volume

% BV

Cats, dogs Cats Dogs Dogs Dogs Dogs

6 kg 6-12 kg 12-20 kg 20-30 kg >30 kg