Hematology Vet-1-120

Veterinary Hematology
A Diagnostic Guide and Color Atlas
John W. Harvey, DVM, PhD, DACVP

Professor and Chair
Department of Physiological Sciences
College of Veterinary Medicine
University of Florida
Gainesville, Florida
with 836 illustrations

3251 Riverport Lane
St. Louis, Missouri 63043
Veterinary Hematology: A Diagnostic Guide and Color Atlas ISBN: 978-1-4377-0173-9

Copyright © 2012 by Saunders, an imprint of Elsevier Inc.
All rights reserved. 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.
Library of Congress Cataloging-in-Publication Data
Harvey, John W.
Veterinary hematology : a diagnostic guide and color atlas / John W. Harvey.
p. ; cm.
Includes bibliographical references and index.
ISBN 978-1-4377-0173-9 (pbk. : alk. paper)
I. Title.
[DNLM: 1. Hematologic Diseases–veterinary–Atlases. 2. Animals, Domestic–blood–Atlases.
3. Hematologic Tests–veterinary–Atlases. SF 769.5]
636.089′615–dc23
2011039099
Vice President and Publisher: Linda Duncan

Acquisitions Editor: Heidi Pohlman
Associate Developmental Editor: Brandi Graham
Publishing Services Manager: Catherine Jackson
Senior Project Manager: Carol O’Connell
Design Direction: Jessica Williams
Working together to grow
libraries in developing countries
www.elsevier.com | www.bookaid.org | www.sabre.org
Printed in China
Last digit is the print number: 9 8 7 6 5 4 3 2 1
T his book is dedicated to the memory of Charles F. Simpson, DVM, PhD. Charlie
trained as a veterinarian at Cornell University and as a pathologist at the University of
Minnesota. He was a member of the University of Florida faculty for 33 years. Charlie’s
research focused on the cardiovascular system, including blood. When I joined the Uni-
versity of Florida faculty in 1974, he had already published hematology papers on babesio-
sis in horses, anaplasmosis in cattle, sickle cell formation in deer, the extrusion of the
metarubricyte nucleus, the maturation of reticulocytes, and the formation of Heinz bodies
in erythrocytes. His major research tool was the transmission electron microscope shown
in the image below. As a mentor and friend, Charlie opened not only his laboratory, but
also his home to me. During the years our careers overlapped, and we published 11 papers
together. Nineteen of his transmission electron microscope images are included in this
book. To me and others who knew him, Charlie is remembered as a warm, generous person
with an inquisitive mind and a wonderful sense of humor. His memory also lives on in
the form of the Charles F. Simpson Memorial Scholarship that is given each year to a
graduate student in the College of Veterinary Medicine at the University of Florida.
John Harvey
Dr. Charles F. Simpson (left) and his technologist Tom Carlisle in front of their transmission electron
microscope.
Foreword
Benefits for the Veterinarian and the Veterinary Clinical Pathologist
When the Atlas of Veterinary Hematology first appeared in the bibliography, outstanding in the 2001 Atlas, has been
2001, it was an instant classic. The atlas was well-organized, comprehensively updated.
concise, and extremely well-illustrated. The accompanying ref- From cover to cover, Veterinary Hematology: A Diagnostic
erence list was comprehensive. Furthermore, and perhaps Guide and Color Atlas reflects the commitment to excellence,
most important, the atlas provided readers with the singular attention to detail, and dedication to the discipline of veteri-
experience, insight, and perspective of one of the world’s most nary hematology that have made John Harvey such a credit
well-known and well-respected veterinary hematologists, to the specialty of veterinary clinical pathology. Although
John Harvey. It quickly became a standard reference for vet- future updates will surely be needed, this new book serves as
erinary clinical pathologists, pathology trainees, private prac- a timeless contribution to our knowledge and understanding
titioners, veterinary technicians, and veterinary students alike. of comparative hematology. As a contemporary colleague and
This long-awaited revision will do nothing but add to the friend of John’s, I have personally had the privilege to learn
original text’s legacy of excellence. Under the new title Veteri- from him directly through our numerous discussions, interac-
nary Hematology: A Diagnostic Guide and Color Atlas, this tions, and case reviews. Through this, his new master work,
edition is significantly expanded and updated. The outstand- his wisdom is preserved and made available for the ages.
ing illustrations that made the original Atlas of Veterinary
Hematology such a valuable contribution have been augmented
by additional illustrations that even include electron micro- Alan H. Rebar, DVM, PhD, DACVP
graphs of a number of significant hematologic disorders. The Senior Associate Vice President for Research
discussions of the physiology and pathophysiology of blood Executive Director of Discovery Park
and bone marrow have been very significantly enhanced Professor of Veterinary Clinical Pathology
throughout the text, thereby making the new volume even School of Veterinary Medicine
more valuable to specialists and students at all levels. Of note, Purdue University
iv
Foreword
Benefits for the Veterinary Technician
Veterinary professionals and students alike need comprehen- Veterinary Hematology: A Diagnostic Guide and Color Atlas
sive medical reference texts that may better enable us to provides additional information on techniques for performing
perform our jobs accurately and help our patients. Even the hematology testing. Veterinary technicians and students will
brightest minds are not walking encyclopedias. Therefore, especially appreciate the logical organization of the material
Veterinary Hematology: A Diagnostic Guide and Color Atlas is a as well as the additional information included regarding clini-
needed source of knowledge for both study and in practice. cal and diagnostic aspects of specific diseases.
I first met Dr. Harvey more than 30 years ago during an Veterinary Hematology: A Diagnostic Guide and Color Atlas
internship at the University of Florida, and I have attended enhances understanding of the material and serves as a
many of his continuing education lectures through the years. vital reference text for both veterinary technician students
Dr. Harvey has always been my ‘go-to’ specialist. If I am and practicing veterinary technicians. I am confident this
looking under the microscope, he’s always just a ‘flip of a page’ reference atlas and guide will be a welcomed addition to
away! every veterinary clinic and classroom throughout the veteri-
The original Atlas of Veterinary Hematology: Blood and Bone nary world.
Marrow of Domestic Animals has for years been a valuable
resource both as a bench-top reference for practicing veteri-
narians, veterinary students, and veterinary technicians and an
aid to technician students in their quest for mastery of this Elaine Anthony MA, CVT
complex topic. This exciting new book is a combination of the Associate Professor
Atlas with Veterinary Laboratory Medicine: Interpretation and School of Veterinary Technology
Diagnosis. St. Petersburg College
v
Preface
This reference presents images and information concerning Veterinary technologists will likely find the techniques and
the hematology of common domestic mammals including blood cell morphology sections most useful. Attempts were
dogs, cats, horses, cattle, sheep, goats, pigs, and llamas. The made to include both common and rare morphologic findings
hematology of nonmammalian species is presented superfi- in blood and bone marrow, including preparation and staining
cially, primarily focusing on comparisons with mammals. This artifacts. Often, more than one example of a cell type, parasite,
book updates and expands the material published in 2001 in or abnormal condition is shown, because the morphology can
the Atlas of Veterinary Hematology: Blood and Bone Marrow of be variable. The benefits and pitfalls of automated instrumen-
Domestic Animals by John Harvey and combines its morpho- tation are discussed, as is the importance of manual blood film
logic content with additional updated topics covered in the review as an important quality control measure for hematol-
introduction and hematology chapters of Veterinary Labora- ogy instrument-generated data and in the identification of
tory Medicine. Interpretation and Diagnosis, 3rd edition, which morphologic abnormalities and parasites that cannot be
was published in 2004 by Dennis Meyer and John Harvey. detected using automated instruments.
Even more information is provided concerning the clinical Practicing veterinarians and veterinary students should
and hematologic appearance of specific disorders, and electron benefit from this complete reference, even if they are not
microscopy images have been added to provide ultrastructual directly involved in bone marrow evaluation, because the bone
detail of cell morphology. marrow chapters provide the basis for understanding diseases
This new text and atlas covers all aspects of hematology that result in abnormalities in the peripheral blood. The added
except therapy. It contains concise discussions of hematopoiesis electron microscopy images and bone marrow aspirate smear
and the physiology of erythrocytes, leukocytes, and hemostasis cytology and core biopsy histology chapters will be most
that provide the foundation needed to understand disorders of useful to clinical pathologists, anatomic pathologists, and resi-
blood. These topics are presented in sufficient detail to be ben- dents in training for these disciplines. Readers interested in
eficial in the training of interns and residents, as well as veteri- learning more about a given topic will hopefully appreciate
nary students. The differential diagnoses of anemia, leukocyte the extensive bibliography provided.
disorders, and hemostatic disorders are presented in such a way
as to emphasize the pathophysiology underlying these pro-
cesses. The utilization and interpretation of both routine and
specialized diagnostic tests are also discussed. John Harvey
vi
Acknowledgments
I want to acknowledge those most responsible for my educa- sessions. Contributors are acknowledged in appropriate figure
tion as a clinical pathologist. Few people have the opportunity legends for photographs I have taken from these glass slides.
to receive training from the giants of their profession, but Likewise, I am extremely grateful for images enthusiastically
I was blessed in being trained by Jerry Kaneko, the father of provided by colleagues that are also acknowledged in the
veterinary clinical biochemistry, and Oscar Schalm, the father figure legends.
of veterinary hematology, during my graduate training at the Jennifer Owen and Heather Wamsley provided the con-
University of California, Davis. Many other colleagues have scientious reviews and helpful suggestions for which I will be
contributed to my development as a hematologist, with Alan forever grateful. Completion of text was only possible because
Rebar and Victor Perman being particularly noteworthy, as Glen Hoffsis, Dana Zimmel, John Haven, and Rachel DiSesa
we challenged one another with unknown hematology slides assumed some of my job duties during the last year. I also
in front of various national audiences. Past and present Uni- appreciate the strong support from Elsevier staff members,
versity of Florida faculty members, Charles Simpson, Dennis most notably Brandi Graham and Carol O’Connell.
Meyer, Rose Raskin, Mary Christopher, Rick Alleman, Finally, I especially want to thank my wife Liz for her
Heather Wamsley, Mark Dunbar, and many University of patience, understanding, and support, not only during the
Florida residents have advanced my understanding of veteri- many months required to produce this text, but during my
nary hematology. entire academic career. Joseph Campbell urged everyone to
University of Florida clinical pathologists and technolo- “Follow your bliss.” For me that has been teaching and research
gists, most notably Melanie Pate, Lane Pritchard, and Tina in veterinary hematology, and this pathway would not have
Conrad, have also helped me by identifying and preparing been possible without Liz’s many years of support.
material included in this text. I greatly appreciate all the col-
leagues who have submitted material to the annual American
Society for Veterinary Clinical Pathology slide review John Harvey
vii
C H A P T E R
1
Introduction to Veterinary
Hematology
Laboratory tests are done for a variety of reasons. Screening I N T ER N A L V ER S U S EX T ER N A L

tests, such as a complete blood count (CBC), may be done on L A B O R AT O R I E S
clinically normal animals when they are acquired to avoid a
financial and/or emotional commitment to a diseased animal, A variety of factors should influence the decision of whether
to examine geriatric patients for subclinical disease, or to a test will be done in an in-house laboratory or be sent to an
identify a condition that might make an animal an anesthetic external laboratory. A major concern is whether the necessary
or surgical risk. Screening tests are often done when an ill personnel, equipment, and supplies are available to perform
animal is first examined, especially if systemic signs of illness the test accurately. Considerations include personnel knowl-
are present and a specific diagnosis is not apparent from the edge of species differences and a willingness to conduct
history and physical examination. Tests are also done to quality-control tests to verify that the procedure is working
confirm a presumptive diagnosis. A test may be repeated or a properly. The costs per test (technician time, reagent costs,
different test may be done to confirm a test result that was equipment costs) must be compared to determine which
previously reported to be abnormal. Tests may be done to option is more economical. The stability of the test may deter-
assist in the determination of the severity of a disease, to help mine whether it will be done internally. The time it takes to
formulate a prognosis, and to monitor the response to therapy obtain results may be important, especially with critically ill
or progression of disease. patients. The hours of operation of the laboratories are impor-
Decisions to request hematology tests in animals are largely tant for test results that are needed at night or on the weekend.
based on the cost of the test versus the potential benefit of Commercial laboratories generally have better quality control
the result to the animal. A CBC is routinely done to establish than laboratories within private practices.
a database for patient evaluation, while other hematology tests Commercial veterinary laboratories are preferred to com-
may be done in an attempt to evaluate a specific problem. mercial human laboratories because errors can occur if tests
Examples of more specific hematologic tests that focus on a designed to evaluate human samples are used without modi-
problem identified during the diagnostic evaluation of an fication to test samples from animals. Hematology analyzers
animal include coagulation tests, such as prothrombin time; must be calibrated for species differences to obtain accurate
bone marrow biopsy and interpretation; and immunologic results. Technologists must be aware that blood cell morphol-
tests, such as the direct Coombs’ test. Although single tests ogy and blood parasites are different in various animal species.
may be done to address a specific problem (e.g., an erythrocyte Antibody-dependent immunology tests designed for humans
phosphofructokinase assay), multiple tests are often utilized are generally not valid in animals. Veterinary laboratories are
to provide a more comprehensive answer to a broader problem more likely to have established their own reference intervals
(e.g., a hemostasis panel is generally requested to evaluate a for various animal species (as opposed to extracting them from
bleeding animal). the literature) than are human laboratories. A knowledge of
Stat is an abbreviation for statim (Latin meaning “immedi- specific animal diseases and training in veterinary laboratory
ately”). Stat tests are tests that are given high priority and medicine is essential for the evaluation of hematologic speci-
begun immediately in situations where rapid results are needed mens and interpretation of laboratory data; consequently a
for the medical management of critically ill patients. Addi- veterinary clinical pathologist should be available to perform
tional fees may be charged for stat tests because they disrupt certain subjective tests and provide consultation concerning
the flow of work in the laboratory and result in inefficiency. all test results.
1
2 VETERINARY HEMATOLOGY
animals of both sexes and various breeds. Monogastric animals

R EF ER EN C E I N T ERVA L S should have been fasted overnight prior to blood sample
collection.
In order to be able to interpret laboratory data from ill animals,
it is essential that appropriate reference intervals be estab- Determination of Reference Intervals
lished from apparently healthy animals drawn from the same Specific reference intervals should be established for each
general population as the ill animals to be examined. The term instrument and each test evaluated. Ideally, each animal would
reference interval is preferred to the commonly used normal have its own reference intervals established by multiple assays
range. The latter term implies that it is the range of test results done over time when the animal was healthy. In some instances,
from all “normal” animals. In reality, a low percentage of limited numbers of baseline values are available for an animal
apparently healthy “normal” animals will have test values that can be helpful, but rarely are analytes measured often
outside the normal range, and, depending on the test, many enough to establish an accurate reference interval for an
abnormal (diseased) animals may have values within the individual animal. Consequently population-based reference
normal range. Healthy animals may have transient increases intervals are used.
or decreases in laboratory test results based on changes in When the frequency diagram of test results from a healthy
environment, emotional status, diet, and so on, and a low population is examined, many analytes exhibit a Gaussian or
percentage of healthy animals simply have values above or bell-shaped distribution (Fig. 1-1). When a Gaussian distri-
below the general population of healthy animals. Apparently bution is present, a minimum of 40 individuals (100 or more
healthy animals may also have occult disease that causes one is preferred) should be assayed for statistical validity.2 In this
or more abnormal laboratory test results, and sample collec- case, the reference interval is calculated using the mean ±2
tion, handling, and laboratory errors can result in artifactually standard deviations (SD). This interval approximates the 95%
high or low values from healthy animals. Consequently it is confidence interval. In other words, about 95% of healthy
not appropriate simply to use the actual range of values from animals have test values within this reference interval, with
all apparently healthy animals assayed. To develop useful ref- about 2.5% of healthy animals having values above and about
erence intervals, one must decide which animals will be 2.5% of healthy animals values below the reference interval.
assayed, how many animals need to be analyzed, and what A common mistake made by novices is to calculate the refer-
method or methods will be used to remove high or low outli- ence interval from the mean ±1 SD. When this is done, about
ers that would otherwise render the interval of limited value 32% of healthy animals will have values outside the calculated
as a reference. interval. If less than 40 healthy animals are available, the upper
and lower values measured should be used to create an esti-
Selection of Reference Animals mated reference interval.5
Specific reference intervals are needed for each species of Some analytes do not exhibit a Gaussian distribution. Most
animal being tested. Less often, a different reference interval commonly there is a skew toward the higher values. The use
is needed for an analyte from a specific breed of animal (e.g., of mean ±2 SD to calculate reference intervals results in
hematocrit values in greyhound dogs are higher than those in inappropriate reference intervals for skewed populations, as
most other dog breeds). Values may vary with the age of the shown in Figure 1-2. Data may be manipulated (e.g., log or
animal, with major changes occurring prior to puberty (e.g.,
3-week-old pups have lower hematocrits than adults). Con-
sequently some analytes need different reference intervals for Mean
different age groups. Some analytes also vary with sex, preg-
Relative frequency
nancy, emotional state, and activity level. The types of animals

sampled and environmental conditions present during the
establishment of a reference interval should be defined, along
with the methods and equipment used, so that the user can
make appropriate evaluations. Ideally, a reference interval
should be established using a population of healthy animals
with a composition (age, breed, sex, diet, etc.) like the popula-
tion of ill animals being evaluated. Homogeneous populations
generally have more narrow reference intervals than heterog- 16 18 20 22 24 26
enous populations. Establishing a reference interval for a mmol/L
blood analyte using a group of male foxhound dogs housed in
a research colony, fed the same diet, and conditioned to phle- FIGURE 1 -1
botomies would likely result in reference intervals too narrow Frequency diagram of a hypothetical plasma analyte with Gaussian dis-
tribution. The central (tallest) vertical line denotes the mean. Each addi-
for the population of dogs examined in a typical small-animal tional vertical line represents one standard deviation (SD) from the
practice. Reference intervals are generally established for a adjacent vertical line. The reference interval calculated using mean ±2 SD
species by utilizing samples from apparently healthy adult (21 ± 4 mmol/L) is 17 to 25 mmol/L.
C ha p ter 1 n Introduction to Veterinary Hematology 3
square root transformation) so that the frequency distribution conducting the test. Hematology indices such as the red
of the transformed data approximates a Gaussian distribution. cell distribution width (RDW) vary more between laborato-
The boundaries are determined as before and results are ries, making the use of published reference intervals less
retransformed to determine the reference interval. Alterna- acceptable.
tively, one can use percentiles to determine upper and lower The units used in reporting values can vary by laboratory
limits, especially if large numbers of healthy animals are evalu- and a conversion factor may be needed to compare a measured
ated. Values are listed in ascending order. The lower limit is value to a published reference interval. For example, blood
determined by the formula (n + 1) × 0.025, and the upper iron might be reported as 100 µg/dL or 18 µmol/L. Most U.S.
limit is determined by the formula (n + 1) × 0.975, where n laboratories continue to use conventional units, such as mg/
= the number of normal animals assayed.2 If 119 animals were dL; Canadian and European laboratories use the Interna-
used, the value for the 3rd lowest animal would be used as the tional System of Units (SI units), such as mmol/L. Where
lower limit and the value from the 117th animal (3rd from possible, moles are used rather than weight (e.g., mg) for SI
the top) would be used as the upper limit. units. This cannot be done for analytes, such as serum
protein concentration, where the molecular weight is variable
Interpretation of Test Results Relative to and/or unknown. For enzymes, an SI enzyme unit is defined
Reference Intervals as 1 µmol/min of substrate utilized or product formed. SI
The common usage of the 95% confidence interval to establish units are reported per liter.
reference intervals means that 5% of healthy animals will be For many wild animal species, reference intervals may not
reported as abnormal for a given test. When multiple tests are be published for some or all tests. The simultaneous measure-
done in laboratory medicine profiles, the probability of at least ment of a healthy “control” animal from the same species,
one test being abnormal increases with the number of tests preferably a cohort, can be used as a rough guideline reference
done. For example, there is a 64% chance that at least one value and therefore can aid interpretation of the patient’s
abnormal test result will be obtained when 20 analytes are results.
measured from a healthy animal.6 The degree to which a test
result is above or below the reference interval is generally
important in deciding whether a high or low value should be S EN S I T I V I T Y A N D S P E C I F I C I T Y
taken seriously. OF TESTS
Use of Published Reference Intervals Ideally analyte values obtained from a healthy animal popula-
Routine hematology test results are usually similar between tion would not overlap with values obtained form a diseased
laboratories; consequently published reference intervals for animal population. Unfortunately there is almost always some
values such as total leukocyte counts and hematocrits are often overlap in the distribution of individual analyte test results
used to interpret results from a species (e.g., wallaby) when between the two groups (Fig. 1-3). When the disease being
reference values have not been established in the laboratory considered has a major impact on an analyte, little overlap in
Mean
Healthy
Relative frequency
Relative frequency
Diabetes
0 0.2 0.4 0.6 0.8 1.0 1.2 0 50 100 150 200 250
Cells 103/L Glucose (mg/dL)
FIGURE 1-2 FIGURE 1 -3

Frequency diagram of hypothetical absolute blood cell counts with a Overlapping Gaussian distributions of a healthy dog population com-
skewed population. The central (tallest) vertical line denotes the mean. pared with a population of dogs with type 2 diabetes mellitus.
Each additional vertical line represents one standard deviation (SD) from
the adjacent vertical line. The use of mean ±2 SD to calculate the refer- The figure is redrawn from Farver TB. Concepts of normality in clinical bio-
ence interval is inappropriate, as demonstrated by the lower limit being chemistry. In: Kaneko JJ, Harvey JW, Bruss ML, eds. Clinical Biochemistry
an impossible negative value. of Domestic Animals. 6th ed. San Diego: Academic Press; 2008:1-25.
values will occur; however, extensive overlap occurs if the for which they are being tested. As can be seen in Figure 1-4,
analyte concentration is minimally altered by the disease being if one increases the reference interval of the healthy popula-
considered. True positives (TPs) are positive test results from tion in order to minimize the FPs, the number of FNs
animals with the disease for which they are being tested, false increases.
positives (FPs) are positive test results for animals without the A clinical test should be safe and practical, and should
disease for which they are being tested (Fig. 1-4), true nega- accurately indicate the presence or absence of a specific disease
tives (TNs) are negative test results from animals without the or pathology. Sensitivity, specificity, and predictive value con-
disease for which they are being tested, and false negatives stitute measures of a test’s utility for ruling in or ruling out a
(FNs) are negative test results from animals with the disease given disease.
Sensitivity is the likelihood of a positive or abnormal test
result occurring in animals with the disease being considered
(Box 1-1). For example, if 23 of 28 cats with feline infectious
peritonitis (FIP) are recognized to have a low absolute lym-
phocyte count in blood, the sensitivity of lymphopenia as a
Healthy diagnostic test for cats with FIP is calculated to be 82%
Relative frequency
(Tables 1-1 and 1-2).7

Specificity is the likelihood of obtaining a negative or
normal test result in nondiseased animals—that is, animals
FP without the particular disease under consideration. In other
TN
words, specificity represents the proportion of animals without
the disease in question that have normal tests. Specificity may
FN
TP be calculated in two distinctly different ways, either by assum-
Diabetes
ing that all of the nondiseased animals are healthy or by
A assuming that although nondiseased animals do not have the
particular disease for which the analysis is being performed,
they may have other diseases.
Determining the specificity of a test in a group of healthy
animals is of little value because reference intervals are gener-
ally established to include 95% of the total population of
Healthy healthy animals, with 2.5% of healthy animals having values
Relative frequency
above and 2.5% of healthy animals having values below the

reference interval. The specificity of a test is much more useful
when the population of animals typically evaluated in a vet-
FP erinary hospital setting is being used.1 In this approach, the
TN “nondiseased” group includes not only healthy animals pre-
sented for elective procedures but also animals with diseases
FN TP other than the disease being considered.
Diabetes
P R ED I C T I V E VA LU E S A N D
0 50 100 150 200 250 D I S E A S E P R EVA LEN C E
B Glucose (mg/dL)
Predictive values demonstrate how well a test performs in a
FIGURE 1-4 given population. In contrast to sensitivity determinations
Frequency diagrams of a healthy dog population compared with a popu- (which are made using only a population of animals with the
lation of dogs with type 2 diabetes mellitus. Graphs are redrawn from disease in question) and specificity determinations (which are
Figure 1-3 to demonstrate true-negative (TN), false-negative (FN), true- made using only a population of animals without the disease
positive (TP), and false-positive (FP) values used to calculate sensitivity,
specificity, and predictive values. The top graph (A) demonstrates the
under consideration), predictive value determinations are
effect of using the mean +2 standard deviations (SD) to set the upper made from populations that contain animal both with and
limit of the reference interval. The lower graph (B) demonstrates the without the disease in question.
effect of using the mean +3 SD to set the upper limit. The number of The predictive value of a positive test (PVPT) considers
FP tests are reduced but the number of FN tests are increased by using only animals in the population being studied that have a posi-
the higher reference limit.
tive test result and determines what percentage of animals
The figure is redrawn from Farver TB. Concepts of normality in clinical actually have the disease being considered (see Box 1-1). It
biochemistry. In: Kaneko JJ, Harvey JW, Bruss ML, eds. Clinical Biochemis- answers the question “How likely is it that an animal with a
try of Domestic Animals. 6th ed. San Diego: Academic Press; 2008:1-25. positive test will actually have the disease being considered?”
Table 1-1
Test Results from the Evaluation of 224 Cats with a History and Clinical Signs Consistent with Feline
Infectious Peritonitis (FIP) Resulting in the Inclusion of FIP in the List of Differential Diagnosesa
NUMBER OF CATS AFFECTED
Test Have FIP (N = 28) Do Not Have FIP (N = 196) Total Cats (N = 224)
Lymphopenia (<1.5 x 103 cells/µL) 23 43 66
Monocytosis (>0.9 x 103 cells/µL) 2 43 45
Hyperglobulinemia (>5.1 g/dL) 11 7 18
Coronavirus titer positive 22 84 106
N, Number of cats.
a
Data from Sparkes AH, Gruffydd-Jones TJ, Harbour DA. An appraisal of the value of laboratory tests in the diagnosis of feline infectious peritonitis. J Am Anim
Hosp Assoc. 1994;30:345-350.
Box 1-1 Table 1-2

Formulas for the Calculation of
Sensitivity, Specificity, Predictive Examination of Lymphopenia as a Diagnostic Test
Value of a Positive Test, for Feline Infectious Peritonitis (FIP)a,b
Predictive Value of a Negative NUMBER OF CATS AFFECTED
Test, and Prevalence
Have FIP Do Not Have FIP Total Cats
Test (N = 28) (N = 196) (N = 224)
TP × 100
Sensitivity (%) = Lymphopenia 23 43 66
TP + FN
True False positive Total
TN × 100
Specificity (%) = positive positive
TN + FP No lymphopenia 5 153 158
TP × 100 False True negative Total
Predictive value of a positive test (%) =
TP + FP negative negative
TN × 100
Predictive value of a negative test (%) =
TN + FN N, Number of cats.
a
( TP + FN) × 100 Cats with a history and clinical signs consistent with FIP were evaluated,
Prevalence (%) = resulting in FIP being included in the list of differential diagnoses. Lymphope-
TP + TN + FP + FN nia was defined as <1.5 x 103 lymphocytes per microliter of blood.
b
Data from Sparkes AH, Gruffydd-Jones TJ, Harbour DA. An appraisal of the
TP, true positive (the number of animals with the disease being tested for that
value of laboratory tests in the diagnosis of feline infectious peritonitis. J Am
have a positive test result); FP, false positive (the number of animals without
Anim Hosp Assoc. 1994;30:345-350.
the disease being tested for that have a positive test result); TN, true negative
(the number of animals without the disease being tested for that have a nega-
tive test result); and FN, false negative (the number of animals with the
disease being tested for that have a negative test result).
prevalence of a disease affects the predictive values of a test
used to diagnose the disease but not its sensitivity or specific-
Based on the selected population of cats presented in Tables ity. For most tests, the PVPT will be low and the PVNT will
1-1 and 1-2, there is a 23/66 or 35% chance that a cat with be high if the disease has a low prevalence in the population
lymphopenia in this population will have FIP.7 being studied. The PVPT will be low because low prevalence
The predictive value of a negative test (PVNT) considers magnifies the number of false-positive results—that is, most
only animals in the population being studied that have a nega- positive test results are false positives because few animals
tive or normal test result and determines what percentage of actually have the disease (see Box 1-1). The exception would
animals with negative test results do not have the disease be a test where false-positive results occur infrequently (e.g.,
being considered (see Box 1-1). It answers the question “How polymerase chain reaction tests for specific infectious agents
likely is that an animal with a negative or normal test result or inherited blood cell defects). The PVNT will be high
will be free of the disease being considered?” Based on the because few false-negative results are present in a population
selected population of cats presented in Table 1-2, there is a when the disease prevalence is low.
153/158 or 97% chance that a cat with a normal or increased To improve diagnostic accuracy, the prevalence (likelihood)
blood lymphocyte count will not have FIP. of the disease being considered can be increased by using the
The prevalence of a disease in a population is simply the history, physical examination, and adjunctive diagnostic tests
percentage of animals in a given population that have a certain to restrict the population, as described for cats in Table 1-1.
disease (see Box 1-1). The prevalence of FIP in the selected The prevalence of FIP in the general cat population is much
population presented in Table 1-1 is 28/224 or 12.5%. The lower than 12.5%. By ruling out one or more diseases that can
give the same positive test result as the disease being consid-
ered, a clinician decreases the size of the population being C U T O F F VA LU E S
studied, thereby increasing the prevalence of the disease in the
population and increasing the positive predictive value of the The PVPT may be increased by using a cutoff value above or
test for the disease being considered. below the standard reference interval, depending on whether
Laboratory tests are used to help rule in or rule out a spe- the disease under consideration results in an increase or a
cific disease. When significant hazards are associated with decrease in the analyte being measured. For example, low
treatment (e.g., amputation or high-risk chemotherapy) or mean cell volume (MCV) is a diagnostic test suggestive of
euthanasia is being considered, it is necessary to be as certain chronic iron deficiency in dogs. If we use 64 fL as the lower
as possible that the disease is actually present. Consequently limit of the reference interval to calculate its positive predic-
tests with high positive predictive values are needed for a tive value, the value would not be remarkably high because
rule-in strategy. When the penalty for missing a diagnosis is there are various other relatively common disorders that can
high, as with a disease for which therapy is effective if begun result in low MCVs in dogs, most notably inflammatory con-
quickly, tests with high negative predictive values are theoreti- ditions and portosystemic shunts. However, it is recognized
cally important as a rule-out strategy. A normal test result by that the other causes of microcytosis rarely result in MCV
virtue of its high negative predictive value would suggest that values below 52 fL. Consequently, if a dog has a MCV below
the disease is not present. Unfortunately many diseases have 52 fL, chronic iron deficiency anemia is highly likely and the
low prevalence, which by itself can result in a high negative PVPT using this cutoff value would approach 100%. However,
predictive value. The best evidence for ruling out a disease is 52 fL is not routinely used as a cutoff value for a positive test
finding a negative test result for an assay that has a high because many cases of chronic iron deficiency would be
sensitivity for recognizing the disease. Based on the selected missed. Nonetheless, it is important to realize that dogs with
population of cats presented in Tables 1-1 and 1-2, finding a especially low MCV values almost certainly have chronic iron
normal or increased blood lymphocyte count is more reliable deficiency anemia.
for ruling out FIP than is finding a low lymphocyte count for The effects of varying the cutoff value of a test on sensitiv-
making a diagnosis of FIP. ity, specificity, and predictive values are demonstrated in Table
Information is generally available concerning the sensitiv- 1-3, where plasma fibrinogen concentration was evaluated as
ity of routine tests for common diseases, but information is a diagnostic test for Rhodococcus equi pneumonia in 165 foals
often lacking concerning all diseases that may have a positive from a single farm.3 It is important to recognize that fibrino-
test result and the frequency of these diseases in the popula- gen is an acute-phase protein that often increases in associa-
tion being evaluated. Consequently, the specificity of a test can tion with various causes of inflammation in horses and that
vary when populations containing animals with other diseases the heat precipitation assay used to measure fibrinogen (while
are analyzed. PVPTs and PVNTs also vary considerably easily performed and clinically useful) is relatively imprecise.
depending on the population analyzed. Although accurate As the cutoff value for plasma fibrinogen concentration is
values are not usually available for PVPTs and PVNTs, clini- increased, the specificity and PVPT increase, but the sensitiv-
cians use their knowledge and experience, combined with the ity and PVNT decrease (see Table 1-3). Results from this
principles outlined above, to make informed judgments con- study also demonstrate that the PVPT increases and the
cerning the likelihood that a disease can be ruled in or ruled PVNT decreases as the prevalence of disease in a population
out of the differential diagnosis. These decisions are seldom increases. In choosing the most appropriate cutoff value for a
based on a single test result; instead, information in the history test, one must consider a number of factors including sensitiv-
is considered along with the clinical signs and results of the ity and specificity of the test, prevalence of disease in the
physical examination, diagnostic imaging, and other labora- population being tested, and consequences of false-positive
tory tests. The likelihood that a disease will be present increases and false-negative tests. In the example above, failure to iden-
if several findings are supportive of the diagnosis. For example, tify an infected foal (false-negative test) might result in the
in the FIP study discussed above, the PVPT was 35% for cats debilitation or death of the foal. Conversely, the treatment of
with lymphopenia, 77% for cats with lymphopenia and hyper- healthy foals based on false-positive test findings could result
globulinemia, and 89% for cats with lymphopenia, hyper- in unnecessary financial losses and potential injury to healthy
globulinemia, and a positive coronavirus titer. The PVNT foals as a result of the adverse side effects of antimicrobial
increased from 97% when lymphopenia alone was absent to therapy.
99% when all three findings were absent. Minimal change
occurs in the PVNT because the relatively low disease preva-
lence in the population is a major contributing factor to the ACC U R AC Y V ER S U S P R E C I S I O N
high negative PVNT. This contribution is most clearly dem-
onstrated by looking at blood monocyte data in the FIP study The accuracy of an analytical procedure is determined by
presented in Table 1-1. Only 7% of FIP cats have a monocy- how closely the result approaches the true value of the
tosis (sensitivity), and the PVPT for monocytosis is only 4%, analyte being measured. An accurate test is one where the
yet the PVNT for a cat lacking a monocytosis is 88%. average of several assay results is close to the true value (Fig.
Table 1-3
Sensitivity, Specificity, and Predictive Values of Plasma Fibrinogen Concentrations at Selected Cutoff Values
for the Early Identification of Foals with Rhodococcus equi Pneumonia, Assuming Two Different Prevalences
of Diseasea
PREDICTIVE VALUES
PREVALENCE 10% PREVALENCE 40%
Cutoff value (mg/dL) Sensitivity (%) Specificity (%) PVPT (%) PVNT (%) PVPT (%) PVNT (%)
300 100 6 11 100 42 100
400 91 51 17 98 55 89
500 71 68 20 96 60 78
600 38 96 51 93 86 70
700 29 97 51 92 86 67
800 12 100 100 91 100 63
PVPT, Predictive value of a positive test; PVNT, predictive value of a negative test.
a
Data from Giguère S, Hernandez J, Gaskin J, Miller C, Bowan JL. Evaluation of white blood cell concentration, plasma fibrinogen concentration, and an agar
gel immunodiffusion test for the early identification of foals with Rhodococcus equi pneumonia. J Am Vet Med Assoc. 2003;222:775-781.
Precise Accurate
40 accurate 40 imprecise
30 30
Test results (mmol/L)
Test results (mmol/L)
Precise Inaccurate
20 inaccurate 20
imprecise
10 10
0 10 20 30 40 0 10 20 30 40
Standard (mmol/L) Standard (mmol/L)

Plots comparing test results of triplicate assays of three standards (y-axis) Plots comparing test results of four replicate assays of three standards
to the known values of the standards (x-axis). The top plot is accurate (y-axis) to the known values of the standards (x-axis). The top plot is
with good precision. The bottom plot has good precision but is accurate but imprecise. The bottom plot is inaccurate and imprecise.
inaccurate.
1-5). Analytic procedures with low accuracy are said to have sample. The CV is the standard deviation (SD) of the repeated
a negative bias if results are below the true value or a positive measurements expressed as a percent of the mean of the
bias if results are above the true value. repeated measurements (SD/mean × 100). The CV indicates
The precision of a test reflects how reproducible the test the amount of random error (imprecision) that is present in
results are when the assay is replicated. Precision is indepen- an assay. A high CV value (e.g., more than 10%) indicates that
dent of accuracy (Fig. 1-6); consequently test results can be an assay lacks precision. A low CV value (e.g., less than 5%)
highly reproducible but erroneous (see Fig. 1-5, lower plot). indicates that assay results are reproducible, varying little with
Precision or, more accurately, the amount of imprecision repeated measurement. The degree of imprecision of an assay
present in an assay, is determined by calculating the coefficient can also be measured over time intervals to assess within-run,
of variation (CV) for repeated measurements made on a single between-run, or between-day variation.
selected blood films (100 cells per differential). CVs were also
Automated versus Manual Methods calculated from results from 20 pairs of randomly selected
As can be seen in Figures 1-7 and 1-8, manual leukocyte and slides (200 cells per differential). Last, CVs were calculated
platelet counts are less precise (CV 15% and 13%, respectively) from results from 20 quads of randomly selected slides (400
than automated leukocyte and platelet counts (CV 2% and cells per differential). As expected, the CVs for leukocyte types
4%, respectively). These values do not indicate whether manual that were numerous (e.g., neutrophils) were much lower than
or automated methods are more accurate. In fact, the mean CVs for leukocyte types that were present in low numbers
manual platelet count is probably more accurate (more near (e.g., basophils), and CVs decreased as the total number of
the true platelet count) than the mean automated platelet cells counted in the differential increased (Figs. 1-9 and 1-10).
count because platelets in small platelet clumps can be visual- The CVs for each of the five leukocyte types from this dog
ized and counted separately in a hemacytometer chamber but were plotted versus the mean percentage of each leukocyte
would be counted as one platelet or not counted at all in an type for 100-, 200-, and 400-cell manual differential counts
automated cell counter. and compared with a like plot with data determined by per-
For manual differential leukocyte counts, the CV varies forming 20 automated differential counts on blood from a
with the percentage of a given leukocyte type present in the single dog using an Advia 120 (Siemens Healthcare Diagnos-
blood film and the total number of leukocytes included in the tics, Inc., Tarrytown, NY) hematology analyzer (Fig. 1-11).
differential leukocyte count. For example, 100 cell differential Automated hematology analyzers have lower CVs for each
counts were performed by a single technologist on each of 80 percentage of leukocyte type present because they examine
stained coverslip blood films from a dog with a mild baso- thousands of leukocytes (assuming a normal leukocyte
philia. CVs were calculated from results of 20 randomly count) in performing the differential leukocyte count.
10
300
200
Leukocytes (103/L)
Platelets (103/L)
100
0 0
Manual Automated Manual Automated

Individual plots of total leukocyte counts performed 20 times each using Individual plots of platelet counts performed 20 times each using a
a manual method and an automated method on the same canine blood manual method and an automated method on the same canine blood
sample. The manual method utilized 20 separate dilutions (Unopette sample. The manual method utilized 20 separate dilutions (Unopette
365855, Becton Dickinson Co., Franklin Lakes, NJ), followed by the 365855, Becton Dickinson Co., Franklin Lakes, NJ), followed by the
counting of all leukocytes in 1 µL of 1/100 diluted blood in a hemacy- counting of all platelets in 1/25 µL of 1/100 diluted blood in a hema-
tometer chamber. A Cell-Dyn 3500 (Abbott Laboratories, North cytometer chamber. A Cell-Dyn 3500 (Abbott Laboratories, North
Chicago, IL) calibrated for canine blood was used to perform the auto- Chicago, IL) calibrated for canine blood was used to perform the auto-
mated cell counts. The mean and coefficient of variation (CV) for the mated cell counts. The mean and coefficient of variation (CV) for the
manual counts were 7.1 × 103/µL and 15% respectively. The mean and manual counts were 240 × 103/µL and 13% respectively. The mean and
CV for the automated counts were 6.7 × 103/µL and 2% respectively. CV for the automated counts were 219 × 103/µL and 4% respectively.
70 100
90 100 Cell manual
65 200 Cell manual
80
Coefficient of variation (%)

400 Cell manual
70 Advia 120
60
Neutrophils (%)
60
55 CV 6% 50
CV 5%
CV 9%
40
50 30
20
45
10
0
40 0 5 10 15 20 25 30 35 40 45 50 55 60
100 200 400
Leukocyte type (%)
Number of cells per differential count
FIGURE 1 -11
FIGURE 1-9
Mean coefficient of variation (CV) values for each of the five leukocyte
Box plots of neutrophil percentages and coefficients of variation (CVs) types from a single dog are plotted versus the mean differential counts
from manual differential counts from a single dog with 55.4% neutro- of each leukocyte type. Mean values were determined from 20 differential
phils. Values represent the results of 20 differential leukocyte counts each leukocyte counts each of 100, 200, and 400 nucleated cells. A like plot
of 100, 200, and 400 nucleated cells. A median line is shown. Boxes with data determined by performing 20 automated differential counts on
include 25th to 75th percentiles and error bars include 10th to 90th blood from a single dog using an Advia 120 hematology analyzer is
percentiles. included for comparison.
5
Critical Difference
4
The CV of an assay affects how the results are interpreted,
especially if an assay is being repeated to determine whether
3
a treatment is effective. For example, if the total leukocyte
Basophils (%)
count for a dog is 4600/µL before treatment and 5800/µL

after treatment, does this represent a real improvement or
2
might it reflect imprecision in the measurement of the total
CV 42% CV 33% leukocyte count? An additional confounding variable in this
1
example is the biological variability of the animal itself.
CV 94% Jensen et al.4 calculated the analytical CV for an automated
0
total leukocyte count in healthy laboratory beagles to be
3.7%, while the CV for repeated total leukocyte counts from
100 200 400
individual beagles (within dog CV) was 12.1%.4 From these
numbers, a critical difference of 35% was calculated. This
Number of cells per differential count
means that the total leukocyte count would have to increase
FIGURE 1-10 by more than 35% before the therapy could be assumed to
Box plots of neutrophil percentages and coefficients of variation (CVs) have an influence on this analyte. In the example above,
from manual differential counts from a single dog with 1.4% basophils. the automated total leukocyte count would have to exceed
Values represent the results of 20 differential leukocyte counts each of 4600/µL × 1.35, or 6200/µL, before a therapeutic effect might
100, 200, and 400 nucleated cells. A median line is shown. Boxes be assumed. A considerably greater difference would be
include 25th to 75th percentiles and error bars include 10th to 90th
percentiles.
required if total leukocyte counts were done using a manual
method because of its higher analytical CV. A greater critical
difference might also have been calculated in the above
example had client-owned animals been used for this study
However, they are not always more accurate. The inability to rather than laboratory animals, because it is likely that the
correctly identify certain cell types (especially basophils), biological variation would be higher in client-owned animals
abnormal cell morphology, or abnormal cell types can lead to that were not accustomed to the phlebotomy procedure, the
the misclassifications of cell types. For example, the Advia 120 individuals handling them, or the environment in which the
failed to identify any basophils in blood from a cat with 39% phlebotomy was done.
basophils or a dog with 14% basophils identified on manual Unfortunately, critical difference measurements have been
differential leukocyte counts. done for few analytes in veterinary medicine, and values will
vary depending on methods and instruments used and animal 3. Giguère S, Hernandez J, Gaskin J, et al. Evaluation of white blood cell concentration,
plasma fibrinogen concentration, and an agar gel immunodiffusion test for the early
populations evaluated. Nonetheless, clinicians develop knowl- identification of foal with Rhodococcus equi pneumonia. J Am Vet Med Assoc.
edge and intuition through study and experience that can help 2003;222:775-781.
them to make informed judgments concerning the impor- 4. Jensen AL, Iversen L, Petersen TK. Study on biologic variability of haematological
components in dogs. Comp Haematol Int. 1998;8:202-204.
tance of changes in laboratory data. 5. Lumsden JH. Reference values. In: Feldman BF, Zinkl JG, Jain NC, eds. Schalm’s
Veterinary Hematology. 5th ed. Philadelphia: Lippincott Williams & Wilkins;
2000:12-15.
R EF ER EN C E S 6. Marshall WJ. The interpretation of biochemical data. In: Marshall WJ, Bangert SK, eds.
Clinical Biochemistry. Metabolic and Clinical Aspects. 2nd ed. New York: Churchill
Livingstone Elsevier; 2008:17-27.
1. Braun JP, Concordet D, Lyazrhi M, et al. Overestimation of the predictive value of posi- 7. Sparkes AH, Gruffydd-Jones TJ, Harbour DA. An appraisal of the value of laboratory
tives by the usual calculations of the specificity of diagnostic tests. Vet Res Commun. tests in the diagnosis of feline infectious peritonitis. J Am Anim Hosp Assoc.
2000;24:17-24. 1994;30:345-350.
2. Farver TB. Concepts of normality in clinical biochemistry. In: Kaneko JJ, Harvey JW,
Bruss ML, eds. Clinical Biochemistry of Domestic Animals. 6th ed. San Diego: Academic
Press; 2008:1-25.
C H A P T E R
2
Hematology Procedures
CO M P O S I T I O N O F B L O O D C A L C U L AT I O N O F
B L O O D VO LU M E
Blood is composed of cells (erythrocytes, leukocytes, and
platelets) circulating within fluid called plasma (Fig. 2-1). Total blood volume accounts for about 10% to 11% of body
Erythrocytes or red blood cells are most numerous, with weight in hot-blooded horses; 8% to 9% in dogs; 6% to 7%
several million erythrocytes per microliter of blood in in cats, ruminants, laboratory rodents, and cold-blooded
mammals (Appendix Table 1). Depending on the species, (draft) horses; and 5% to 6% in pigs. The total blood volume
erythrocytes typically account for one-fourth to one-half of of young growing animals often exceeds 10% of body weight.33
the total blood volume as measured by determining the hema- It may be desirable to calculate the total blood volume of an
tocrit. Platelets or thrombocytes are the next most numerous animal in determining the size of a needed blood transfusion,
cell type in blood, with platelet counts as low as 100 × 103/µL or the amount of blood that can safely be removed for a series
in healthy horses to several hundred thousand per microliter of diagnostic tests, or when an animal is to be used as a blood
in other mammalian species. Total leukocyte or white blood donor. For example, the total blood volume of a 4-kg cat is
cell counts are much lower than erythrocyte and platelet 0.07 × 4 kg = 0.28 kg = 280 mL, assuming that 7% of body
counts, with total leukocyte counts ranging from about 5 × weight is blood in cats and the specific gravity of blood is 1.0
103/µL to about 20 × 103/µL in mammals. The proportion of (1 mL weighs 1 g). Since one can safely remove 20% of the
leukocyte types present varies by species, with neutrophils blood volume from an animal, the calculated amount that can
being the most numerous leukocyte type present in the blood be removed from the cat in this example is 280 × 0.2 = 56 mL.
of carnivores and lymphocytes being the most numerous leu-
kocyte type present in the blood of ruminants and rodents.
Plasma consists primarily of water that contains about 6 to S A M P LE CO LLE C T I O N
8 g/dL of plasma proteins and 1.5 to 2.0 g/dL of inorganic A N D H A N D LI N G
salts, lipids, carbohydrates, hormones, and vitamins.19 Plasma
is prepared in the laboratory by collecting blood with an In monogastric animals, an overnight fast avoids postprandial
anticoagulant, followed by centrifugation to remove the blood lipemia, which can interfere with plasma protein, fibrinogen,
cells. If blood is collected without anticoagulant and allowed and hemoglobin determinations. Ethylenediaminetetraacetic
to clot, the fluid that is obtained following centrifugation is acid (EDTA) is the preferred anticoagulant for complete
called serum. The protein concentration in serum is usually blood count (CBC) determination in most species, but blood
about 0.2 to 0.5 g/dL lower than that in plasma, primarily from some birds and reptiles hemolyzes when collected into
owing to the absence of fibrinogen—which is consumed EDTA. In those species, heparin is often used as the anti
during coagulation—in serum. Serum proteins may be sepa- coagulant. The disadvantage of heparin is that leukocytes
rated by electrophoresis into albumin, α-globulins, β-globulins, do not stain as well (presumably because heparin binds to
and γ-globulins (Fig. 2-2). Albumin is a single protein that leukocytes),24 and platelets generally clump more than
generally accounts for nearly half of the total plasma proteins they do in blood collected with EDTA. However, as discussed
present by weight. Each of the globulin classes is composed later, platelet aggregates and leukocyte aggregates may occur
of many different proteins.12 even in properly collected EDTA-anticoagulated blood
11
Whole
blood Formed Leukocytes
(volume) elements (differential)
(number per L)
Neutrophils
60-77%
Leukocytes
6,000-17,000
Formed
elements Erythrocytes
45% 5.5-8.5
million Eosinophils 2-10%
Platelets
Body 200,000-500,000 Basophils (rare)
weight Lymphocytes
12-30%
Blood 8%
Monocytes 3-10%
Proteins
(electrophoresis)
Plasma
Other weight
fluids
Proteins 7% Albumin
and Plasma
Serum proteins
tissues 55% 44%
92% Globulins
(electrophoresis)
Plasma proteins Alpha 14%

Water Beta 15%
91.5% Globulins
52% Gamma 23%
Fibrinogen 4%
Inorganic salts, lipids, hormones,

vitamins, carbohydrates 1.5%
FIGURE 2-1
Approximate composition of normal dog blood.
samples.10,29,41,49,53 In those cases, collection of blood using be made as soon as possible and rapidly dried to minimize
another anticoagulant (e.g., citrate) may prevent the forma- morphologic changes.
tion of cell aggregates. Cell aggregation tends to be more
pronounced as blood is cooled and stored; consequently
processing samples as rapidly as possible after collection G RO S S EXA M I N AT I O N O F
may minimize the formation of leukocyte and/or platelet B L O O D S A M P LE S
aggregates.
Collection of blood directly into a vacuum tube is preferred Samples are checked for clots and mixed well (gently inverted
to collection of blood by syringe and transfer to a vacuum tube. 20 times) immediately before removing aliquots for hematol-
This method reduces platelet clumping and clot formation in ogy procedures. Horse erythrocytes settle especially rapidly
samples for CBC determinations, as even small clots render because of rouleaux formation (adhesion of erythrocytes
a sample unusable. Platelet counts are markedly reduced, and together like a stack of coins). Blood should be examined
a significant reduction can sometimes occur in hematocrit grossly for color and evidence of erythrocyte agglutination.
(HCT) and leukocyte counts as well. Also, when the tube is The presence of marked lipemia may result in a blood sample
allowed to fill based on the vacuum within the tube, the proper with a milky red color resembling “tomato soup” when
sample-to-anticoagulant ratio will be present. Inadequate oxygenated.
sample size results in decreased HCT due to excessive EDTA
solution. Care should be taken to avoid iatrogenic hemolysis, Methemoglobinemia
which interferes with plasma protein, fibrinogen, and various Hemoglobin is a protein consisting of four polypeptide globin
erythrocyte measurements. Samples should be submitted to chains, each of which contains a heme prosthetic group within
the laboratory as rapidly as possible, and blood films should a hydrophobic pocket. Heme is composed of a tetrapyrrole
C ha p ter 2 n Hematology Procedures 13
with a central iron molecule that must be maintained in the

ferrous (+2) state to reversibly bind oxygen. Methemoglobin
differs from hemoglobin only in that the iron molecule of the
heme group has been oxidized to the ferric (+3) state and is no
longer able to bind oxygen.28 The presence of large amounts of
A deoxyhemoglobin accounts for the dark bluish color of normal
venous blood samples. Methemoglobinemia may not be recog-
Albumin nized in venous blood samples because the brownish color of
methemoglobin is not readily apparent when methemoglobin
2
is mixed with deoxyhemoglobin (Fig. 2-3, A). When deoxyhe-
moglobin binds oxygen to form oxyhemoglobin, it becomes
bright red; consequently the brownish coloration of methemo-
globin becomes more apparent in the oxygenated samples (Fig.
2-3, B). A simple spot test provides a rapid way to oxygenate a
venous blood sample and to determine whether clinically sig-

nificant levels of methemoglobin are present. One drop of
blood from the patient is placed on a piece of absorbent white
paper and a drop of normal control blood is placed next to it.
1 If the methemoglobin content is 10% or greater, the patient’s
blood will have a noticeably brown color compared with the
bright red color of control blood (Fig. 2-4).28 Accurate deter-
B mination of methemoglobin content requires that blood be
submitted to a laboratory that has this test available. Methe-
FIGURE 2-2 moglobinemia results from either increased production of
Serum protein electrophoresis from a dog with a polyclonal hyperglobu- methemoglobin by oxidants or decreased reduction of methe-
linemia, including increased α2-globulin and β-γ bridging. A, Agarose moglobin resulting from a hereditary deficiency in the eryth-
gel stained with Coomassie blue for protein. Albumin (band on the far
left) migrates more rapidly than other proteins. B, Densitometer tracing rocyte cytochrome-b5 reductase enzyme (see Chapter 4).27
of the agarose gel used to determine the contribution of each protein
type: total protein 7.7 g/dL, albumin 2.05 g/dL, α1 0.42 g/dL, α2 1.51 g/ Erythrocyte Agglutination
dL, β 1.86 g/dL, γ 1.86 g/dL. The appearance of red granules in a well-mixed blood sample
(Fig. 2-5) suggests the presence of erythrocyte autoagglutina-
tion. However, it is important to differentiate agglutination
A B
FIGURE 2-3
Gross appearance of mixtures of oxyhemoglobin, deoxyhemoglobin, and methemoglobin. A, Venous blood sample from a cat with 28% methemoglobin
(left sample) compared with blood from a normal cat with less than 1% methemoglobin (right sample). Both samples also contain a mixture of
oxyhemoglobin and deoxyhemoglobin. B, Oxygenated blood sample from a cat with 28% methemoglobin (left sample) compared with blood from a
normal cat with less than 1% methemoglobin (right sample). The sample on the left contains a mixture of oxyhemoglobin and methemoglobin; the
one on the right contains almost exclusively oxyhemoglobin.
FIGURE 2-4
Methemoglobin spot test. A drop of blood from a methemoglobin
reductase-deficient cat with 50% methemoglobin (left) is placed on
absorbent white paper next to a drop of blood from a normal cat with
less than 1% methemoglobin.
FIGURE 2 -6
Microscopic rouleaux in an unstained wet mount preparation of normal
cat blood.
FIGURE 2-5
Grossly visible agglutination in blood from a dog with immune-mediated
hemolytic anemia.
FIGURE 2 -7
(aggregation of erythrocytes together in clusters) from rou-
Microscopic agglutination in an unstained wet mount preparation of
leaux (adherence of erythrocytes together like a stack of coins), saline-washed erythrocytes from a foal with neonatal isoerythrolysis.
which can be seen in the blood from healthy horses and cats
(Fig. 2-6). Rouleaux formation is eliminated by washing
erythrocytes in physiologic saline, but agglutination is not. between negatively charged erythrocytes.67 In addition, there
This differentiation requires centrifugation of blood, removal are 10 antigen-binding sites per IgM molecule compared with
of plasma, and resuspension of erythrocytes in saline. A rapid 2 binding sites per IgG molecule (Fig. 2-8). A direct anti-
way to differentiate rouleaux from agglutination is to mix five globulin test is not needed to identify the presence of immu-
drops of physiologic saline with a drop of anticoagulated noglobulin bound to erythrocytes if agglutination is present
blood on a glass slide and examine it as a wet mount using a in saline solution-washed erythrocyte samples.
microscope. This dilution reduces rouleaux, but agglutination
is not affected. The microscopic appearance of agglutination
in a sample of washed erythrocytes is shown (Fig. 2-7). The M I C RO H EM AT O C R I T T U B E
presence of agglutination indicates that the erythrocytes have EVA LUAT I O N
increased surface-bound immunoglobulins. These immuno-
globulins are usually of the IgM type when agglutination is A microhematocrit tube is filled to about 90% of capacity with
present, because the greater distance between binding sites on well-mixed blood and sealed with clay at one end. The tube is
IgM molecules compared with IgG molecules makes it easier then placed in a microhematocrit centrifuge with the clay
for IgM molecules to overcome normal repulsive forces plug oriented to the periphery of the centrifuge head and
hyperbilirubinemia) in horses owing to reduced removal of

unconjugated bilirubin by the liver.13 Failure of the liver to
Erythrocytes
remove unconjugated bilirubin from blood has also been
reported to cause hyperbilirubinemia in one-fourth of the sick
(often anorectic) cattle in one study.40 In other species, yellow
plasma with a normal HCT suggests hyperbilirubinemia sec-
ondary to liver disease. Hyperbilirubinemia associated with a
marked decrease in the HCT suggests an increased destruc-
tion of erythrocytes; however, the concomitant occurrence of
liver disease and a nonhemolytic anemia could produce a
similar finding (Fig. 2-9, B).
Red discoloration of plasma indicates the presence of free
hemoglobin. This discoloration may represent either true
Anti-erythrocyte IgM hemoglobinemia, resulting from intravascular hemolysis, or
hemolysis occurring after sample collection due to such causes
as rough handling, fragile cells, lipemia, or prolonged storage
(Fig. 2-9, C,D). The HCT value may help differentiate between
FIGURE 2-8 these two possibilities, with red plasma and a normal HCT
Antierythrocyte IgM antibodies causing erythrocyte agglutination. suggesting in vitro hemolysis. The concomitant occurrence of
hemoglobinuria indicates that intravascular hemolysis has
occurred. Plasma will also appear red following treatment with
centrifuged for 5 minutes. After centrifugation, the blood cross-linked hemoglobin blood substitute solutions.
sample will be separated into three layers based on density, Lipemia is recognized as a white opaque appearance caused
with packed erythrocytes located at the bottom. A white by chylomicrons and very low density lipoproteins (VLDLs).
“buffy coat” is located above the erythrocyte layer, and the The presence of chylomicrons may also result in a white layer
acellular plasma is located above the buffy coat. at the top of the plasma column (Fig. 2-9, E). The presence
When blood is submitted for a CBC, most commercial of lipemia is frequently the result of a recent meal (postpran-
laboratories determine an electronic HCT rather than a dial lipemia), but diseases including diabetes mellitus, pancre-
packed cell volume (PCV) by centrifugation. This efficiency atitis, hypothyroidism, hyperadrenocorticism, protein-losing
negates the need to centrifuge a microhematocrit tube filled nephropathy, cholestasis, obesity, and starvation may also con-
with blood. Unfortunately useful information concerning the tribute to the development of lipemia in dogs and cats.23,72
appearance of plasma is missed unless a serum or plasma Transient lipemia and accompanying anemia has been
sample is also prepared for clinical chemistry tests. described as a syndrome in kittens, but a direct link between
hyperlipidemia and anemia has not been clearly documented.23
Packed Cells Hereditary causes of lipemia include lipoprotein lipase defi-
The PCV is measured after centrifugation by determining the ciency in cats and idiopathic hyperlipidemia in miniature
fraction of total blood volume in a microhematocrit tube that schnauzer dogs.20,73
is occupied by erythrocytes. Leukocytes and platelets are pri- Equids of any age and either sex may develop lipemia, but
marily located within the buffy coat, although certain leuko- obese ponies, miniature horses, and miniature donkeys are
cyte types may be present in the top portion of the packed most susceptible. Lipemia has been associated with a broad
erythrocyte column in some species (e.g., neutrophils in range of diseases, but it is more prevalent in association with
cattle). The width of the buffy coat generally correlates directly pregnancy, lactation, and/or anorexia.65 Lipemia has also been
with the total leukocyte count. A large buffy coat suggests reported in llamas and alpacas with severe systemic diseases.
leukocytosis (Fig. 2-9, A) or thrombocytosis, and a small buffy It is not associated with age, sex, or reproductive status in these
coat suggests that low numbers of these cells may be present. camelids.64 The pathogenesis of marked secondary hyperlip-
The buffy coat may appear reddish owing to the presence of idemia is not always clear. It often appears to result from a
a marked reticulocytosis. mobilization of unesterified fatty acids from adipose tissue
and the subsequent overproduction of VLDLs by the liver,
The Appearance of Plasma but it may also result from ineffective clearance of VLDLs
Plasma is normally clear in all species. It is nearly colorless in by tissues or a combination of both, as occurs in diabetes
small animals, pigs, and sheep but light yellow in horses mellitus.
because they naturally have higher bilirubin concentrations.
Plasma varies from colorless to light yellow (carotenoid pig- Plasma Protein Determination
ments) in cattle, depending on their diet.62 Increased yellow After the PCV is measured and the appearance of the plasma
coloration usually indicates increased bilirubin concentration. and buffy coat is noted, the microhematocrit capillary tube is
This increase often occurs secondarily to anorexia (fasting broken just above the buffy coat and the plasma is placed in
A B C D E
FIGURE 2 -9
Gross appearance of microhematocrit tubes demonstrating leukocytosis, icterus, hemolysis, and lipemia.
A, Microhematocrit tube from a cat with a large buffy coat resulting from a chronic lymphocytic leukemia.
B, Microhematocrit tube from an anemic cat, with icteric plasma secondary to hepatic lipidosis. C, Micro-
hematocrit tube with evidence of hemolysis in plasma from a cat with acetaminophen-induced Heinz body
hemolytic anemia. Less dense erythrocyte ghosts can be seen above the packed intact erythrocytes. D, Micro-
hematocrit tube with evidence of hemolysis in plasma from a horse with intravascular hemolysis induced by
the inadvertent intravenous and intraperitoneal administration of hypotonic fluid. Less dense erythrocyte
ghosts can be seen above the buffy coat. E, Microhematocrit tube with marked lipemia in plasma from a dog
with hypothyroidism that was also being treated with prednisone for an allergic dermatitis. It was unclear
from the medical record if this was a fasting blood sample.
A, Courtesy of Heather Wamsley.
a refractometer for plasma protein determination. Plasma B L O O D C ELL CO U N T I N G

protein concentrations in newborn animals (approximately AND SIZING
4.5 to 5.5 g/dL) are lower than adult values and increase to
within the adult range by 3 to 4 months of age.33 The presence Total leukocyte counts, erythrocyte counts, and platelet counts
of lipemia or hemolysis will falsely increase the measured in mammals may be determined using manual or automated
plasma protein value. Maximum information can be gained techniques.
by interpretation of the HCT and plasma protein concentra-
tions simultaneously (see Chapter 4). Manual Cell Counting
Manual erythrocyte counts are not accurate enough to be
Fibrinogen Determination useful. Manual total leukocyte counts and platelet counts can
Fibrinogen can be measured in a hematocrit tube because it be performed using commercially available reservoirs and
readily precipitates from plasma when heated to 56°C to 58oC pipettes to dilute the sample and lyse erythrocytes prior to the
for 3 minutes. The difference between the total protein of the microscopic counting of leukocytes and platelets using a
plasma and the total protein of the defibrinogenated (heated) hemacytometer chamber. Manual leukocyte counts and plate-
plasma gives an estimate of the fibrinogen concentration in let counts are done when errors are suspected in cell counts
the plasma.32 This method is useful in identifying high fibrin- generated by automated cell counters. Manual cell counts may
ogen concentrations, but it is not accurate in identifying low also be done in an emergency situation when automated cell
fibrinogen concentrations.4,11 counts are not available.
All blood cell types in birds and reptiles are nucleated, must usually be corrected for the presence of nucleated eryth-
making accurate separation and counting difficult or impos- rocytes when present.
sible with automated cell counters. Consequently manual
leukocyte counts are generally required in nonmammalian Errors in Blood Cell Counting and Sizing
species. Thrombocytes can be estimated based on the number The accuracy of blood cell counting depends on the quality
present in stained blood films.47 and characteristics of the blood sample as well as the accuracy
If properly maintained, automated blood cell counters are of the analytic methods used. Storage of blood samples for
more precise and accurate than manual techniques in mam- more than a few hours can result in sample deterioration and
malian species. Various technologies—including quantitative inaccurate cell counts. The presence of even small clots in the
buffy coat analysis, impedance measurements, laser flow blood tube invalidates all cell counts. Even without clot for-
cytometry, and cytochemistry—are utilized to generate these mation, platelet aggregation may occur if platelets become
cell counts. activated during blood collection, as can happen when speci-
mens are collected with a syringe and then transferred to an
Automated Cell Counting anticoagulant tube. Heparin is generally not used as an anti-
Quantitative buffy coat analysis (QBC VetAutoread Hema- coagulant because it does not prevent platelet clumping, and
tology System, IDEXX, Inc., Westbrook, ME) depends on leukocyte staining is poor on blood films. EDTA is the pre-
variations in cell density to separate cell types. Cells are not ferred anticoagulant for CBC determinations in most species,
actually counted; instead the widths of the various layers of but EDTA may induce platelet, leukocyte, and erythrocyte
cell types that form are measured and the cell “count” is aggregate formation in some individuals with antibodies
generated assuming a standard cell size for the cell type in bound to the surfaces of these respective cell types.10,29,54,71,75
question.35 When this occurs, collection of blood into citrate may prevent
Impedance counters such as the Heska CBC-Diff (Heska the problem. Cell aggregation tends to be more pronounced
Corporation, Loveland, CO) and Abaxis VetScan HMII as blood is cooled and stored; consequently processing samples
(Abaxis, Union City, CA) depend on the principle that cells as rapidly as possible after collection may minimize the for-
are poor electrical conductors. Blood is diluted in an electri- mation of leukocyte and/or platelet aggregates. EDTA can
cally conductive solution and a precise volume of this diluted cause marked hemolysis in some species of birds (ostriches,
suspension is drawn through a small aperture between two emus, and jays),25,35 reptiles (turtles and tortoises),25,42 and fish
electrodes. Each cell produces a change in electrical imped- (carp and brown trout).43,66 Heparin is used as an anticoagu-
ance, resulting in a change in voltage that is proportional to lant in these species. Both lysis and clumping can result in
the size of the cell counted. Several thousand cells per second reduced cell counts of the cell types involved.
can be counted and sized. Erythrocytes and platelets can be Platelet clumping can not only result in an erroneously
differentiated by size alone in many species using impedance decreased platelet count59 but also in a falsely increased mean
counters, but not in cats, because their platelets are large and platelet volume (MPV) and occasionally a falsely increased
their erythrocytes relatively small. Leukocytes are counted as total leukocyte count.35,75 Autoagglutination of erythrocytes
free nuclei following lysis of erythrocytes and platelets. in the blood sample can result in a spuriously increased mean
The development of laser flow cytometry for use in cell volume (MCV) and mean cell hemoglobin concentration
instruments such as the CELL-DYN 3500R (Abbott Labo- (MCHC) and a decreased HCT and total erythrocyte count.48
ratories, Abbott Park, IL), the Advia 120 (Siemens Healthcare As indicated earlier, with the use of most hematology analyz-
Diagnostics, Inc., Tarrytown, NY), the Sysmex XT-2000iV ers, the presence of nucleated erythrocytes can result in a
(Sysmex Corporation, Kobe, Japan), and the LaserCyte falsely increased total leukocyte count. Residual erythrocyte
(IDEXX, Inc., Westbrook, ME) has allowed for cells to be stroma from inadequate lysing of erythrocytes may also result
characterized in greater detail. Individual cells pass through a in spuriously increased total leukocyte counts.35,76
laser beam, absorbing and scattering light. Interruptions in The presence of severe lipemia can result in spuriously
light are used to count cells and light scatter is used to deter- increased hemoglobin and MCHC values and possibly even
mine size and internal complexity or density. The Advia 120 increased platelet and total leukocyte counts.75 The presence
also utilizes a peroxidase channel to aid in determining spe- of in vitro or in vivo hemolysis in the sample or prior treat-
cific leukocyte types. When properly calibrated, laser flow ment with a cross-linked hemoglobin solution can result in
analysis allows for a more complete and accurate differential an erroneously increased MCHC value.38 The presence of
leukocyte count than can be done using cell size alone. Laser numerous Heinz bodies in erythrocytes can also result in
flow cytometry also permits the development of automated spuriously increased hemoglobin and MCHC values, increased
reticulocyte counts and reticulated platelet counts36,44 as well automated reticulocyte counts,14,52 and sometimes increased
as the development of new erythrocyte and reticulocyte total leukocyte counts.69 The precipitation of an IgM parapro-
parameters based on the simultaneous measurement of size tein in blood from a dog by the lysing reagent used in the
and hemoglobin concentration within individual cells.17 CELL-DYN 3500 falsely increased the spectrophotometric
Nucleated erythrocytes are counted as leukocytes in most measurement of hemoglobin (Hb), which falsely increased the
electronic cell counters; consequently total leukocyte counts calculated MCHC.8
Laboratory errors may occur as a result of mistakes made testing programs provide external quality control. Samples are
by operators. These operator errors include a lack of knowl- periodically sent to participating laboratories for analysis.
edge or the skills for the test being done, improper labeling Results are sent back to the agency supplying the samples and
of samples, dilution errors, use of outdated reagents, inade- these values are compared with those from reference labora-
quate quality control measures, or improper calibration of tories and other participating laboratories using the same
equipment. The instruments being used must be optimized methods. Proficiency testing programs provide valuable peer
and validated for the species being tested. review of instruments and methods used, but the expense is
Erythrocytes and platelets vary considerably in volume beyond the means of most private practices.
among animal species, and instruments must be adjustable to Manual methods are also used as quality control measures
accurately count and size these blood cells for the species for automated hematology instruments. A PCV can be mea-
being assayed. Cats naturally have large platelets and moder- sured following centrifugation of a blood sample for compari-
ately small erythrocytes. The resultant overlap of erythrocyte son to the HCT determined electronically. If values do not
and platelet size makes separation of cat platelets and eryth- match within a couple of percentage points after adequate
rocytes unreliable with the use of impedance counters.77 Con- sample mixing, it suggests that the MCV or the RBC count
sequently cat platelet counts are spuriously decreased when is probably incorrect, because these two values are used to
measured with impedance counters. This inclusion of platelets determine the HCT.
in the erythrocyte measurements can result in increased Stained blood films should be examined as a part of each
erythrocyte counts and HCT values and reduced MCV CBC. The blood film is scanned to verify that the automated
and MCHC values, but the ratio of platelets to erythrocytes total leukocyte count and the platelet count appear to be
is usually not large enough to have appreciable effects on correct. If the automated platelet count is low, it is especially
these parameters (exceptions include cats with severe anemia important to examine for platelet clumps that could result in
and/or marked thrombocytosis).76 Leukocytes are generally a spuriously low count. If an electronic differential leukocyte
included in erythrocyte measurements, but the ratio of leuko- count is to be reported, one must review the blood film to
cytes to erythrocytes is usually not large enough to have verify that the percentages of each leukocyte type recorded
appreciable effects on erythrocyte parameters (exceptions appear to be correct and that there are no other cell types
include animals with severe anemia and marked leukocytosis). present that are not identified (e.g., basophils are not identi-
In these instances the inclusion of leukocytes in erythrocyte fied by most electronic cell counters). More details concerning
measurements can result in increased erythrocyte counts, the estimation of cell counts and the proper examination of
HCT and MCV values, and reduced MCHC values because stained blood films are given later in this chapter.
leukocytes are larger than erythrocytes.76 These errors result-
ing from difficulties in separating cell type by size alone should
be minimized in automated cell counters that separate cells B L O O D - F I L M P R EPA R AT I O N
not only by size but also by internal complexity.
With the advent of new laser flow cytometry techniques, Blood films should be prepared within a couple of hours of
it is now possible to perform automated differential leukocyte blood sample collection to avoid artifactual changes that will
counts; however, most instruments cannot accurately count distort the morphology of blood cells. Blood films are pre-
basophils. For most instruments, a high basophil count is an pared in various ways including the slide (wedge) method,
error,35 as has been noted in old blood samples.63 These flow coverslip method, and automated slide spinner method. It is
cytometers must be specifically calibrated for each species, and essential that a monolayer of intact cells is present on the slide
they work best when leukocyte morphology is normal. More so that accurate examination and differential leukocyte counts
reliable flags are needed to identify the presence of left shifts can be performed. If blood films are too thick, cells will be
and neoplastic cells.35 shrunken and may be difficult to identify. If blood films are
The examination of a stained blood film is essential as a too thin, erythrocytes will be flattened and lose their central
quality control measure regardless of the technology used to pallor and some leukocytes (especially lymphocytes and blast
count blood cells. In addition to verifying the accuracy of cells) will be ruptured.30
leukocyte and platelet counts, a number of other evaluations
are made. Examples include determining whether erythrocyte Glass-Slide Blood-Film Method
polychromasia, erythrocyte shape abnormalities, neutrophilic A clean glass slide is placed on a flat surface and a small drop
left shifts, neutrophil toxicity, reactive lymphocytes, blast cells, of well-mixed blood is placed on one end of the slide (Fig.
mast cells, and/or blood parasites are present. 2-10). This slide is held in place with one hand and a second
glass slide (spreader slide) is placed on the first slide and held
Quality Control between the thumb and forefinger with the other hand at
Commercial control samples (preferably at two levels) about a 30- to 45-degree angle in front of the drop of blood.
approved by the instrument’s manufacturer should be run each The spreader slide is then backed into the drop of blood, and
day; the values obtained should fall within the confidence as soon as the blood flows along the back side of the spreader
intervals supplied with the control sample.37 Proficiency slide, the spreader slide is rapidly pushed forward. The
A B
A B C
C D

FIGURE 2-10
Slide blood-film preparation. A, A glass slide is placed on a flat surface D E
and a small drop of well-mixed blood is placed on one end of the slide
using a microhematocrit tube. B, A second glass slide (spreader slide) is FIGURE 2 -11
placed on the first slide at about a 30-degree angle in front of the drop Coverslip blood film preparation. A, B, One clean coverslip is held
of blood. The spreader slide is then backed into the drop of blood. C, As between the thumb and index finger of one hand and a small drop of
soon as the blood flows along the back side of the spreader slide, the blood is placed in the middle of it using a microhematocrit tube. C, A
spreader slide is rapidly pushed forward. D, The blood film produced is second clean coverslip is dropped on top of the first in a crosswise posi-
thick at the back of the slide, where the drop of blood was placed, and tion. D, Blood spreads evenly between the two coverslips and a feathered
thin at the front (feathered) edge of the slide. edge forms at the periphery. E, The coverslips are rapidly separated by
grasping an exposed corner of the top coverslip with the other hand and
pulling apart in a smooth, horizontal manner.
thickness of the smear is influenced by the size of the blood
drop, the viscosity of the sample (HCT), the angle of the
spreader slide, and the speed of spreading. The greater the One coverslip is held between the thumb and index finger of
angle (a more upright spreader slide) and faster the speed of one hand and a small drop of blood is placed in the middle
spreading, the thicker and shorter the blood film will be. Thus of the coverslip using a microhematocrit tube. The drop of
the spreader slide should be held more upright in preparing blood should be as perfectly round as possible to produce even
smears from anemic patients so as to create a thicker blood spreading between coverslips. The second coverslip is dropped
film.30 on top of the first in a crosswise position. After the blood
If the drop of blood is the proper size, all of the blood will spreads evenly between the two coverslips and a feathered
remain on the slide and a smear will be prepared that is thick edge forms at the periphery, the coverslips are rapidly sepa-
at the back of the slide, where the drop of blood was placed, rated by grasping an exposed corner of the top coverslip with
and thin at the front (feathered) edge of the slide. If the drop the other hand and pulling apart in a smooth, horizontal
of blood is too large, some of the blood will be pushed off the manner. Coverslips are immediately dried as described above
end of the slide, causing potential problems. Often these blood and then identified by marking on the thick end of the smears
films will be too thick for accurate evaluation. Second, clumps with a graphite pencil or a pen containing ink that is not
of cells tend to be pushed off the slide, making them unavail- removed by alcohol fixation.
able for examination. If the drop of blood used is too large, a feathered edge will
Once prepared, the slide is immediately dried by waving it not form and the blood film will be too thick. Multiple cov-
in the air or holding it in front of a hair dryer set on a slightly erslip blood films may be stained in small slotted coplin jars
warm-air setting. Holding the slide close to a dryer set on a or in ceramic staining baskets that are placed in beakers of
hot-air setting can result in fragmentation of cells. Slow fixative and stain.
drying can cause cells to contract, making them difficult to
identify.30 Slides are identified by writing on the thick end of
B L O O D - F I L M S TA I N I N G
the smear or the frosted end of the slide with a graphite pencil
P RO C ED U R E S
or a pen containing ink that is not removed by alcohol
fixation. Romanowsky-Type Stains
Blood films are routinely stained with a Romanowsky-type
Coverslip Blood-Film Method stain (e.g., Wright or Wright-Giemsa) either manually or
Two 22-mm square No. 1 1 2 coverslips are required to make using an automatic slide stainer. Romanowsky-type stains are
coverslip blood films (Fig. 2-11). A camel’s hair brush is used composed of a mixture of eosin and oxidized methylene blue
to remove particles from the surfaces that will contact blood. (azure) dyes. The azure dyes stain acids, resulting in blue to
A B
FIGURE 2-12
Appearance of blood films from a normal cat stained with Wright-Giemsa. A, Blood film was rinsed in distilled water. Five neutrophils, a basophil
( far right), and a lymphocyte (round nucleus) are present. Erythrocytes exhibit rouleaux, a normal finding in cats. B, Blood film was rinsed in tap
water. A neutrophil (left), monocyte (bottom right), and lymphocyte (top right) are present. The blue color of the erythrocytes results from using water
with inappropriate pH.
purple colors, and eosin stains bases, resulting in red color-

ation (Fig. 2-12, A). These staining characteristics depend on
the pH of the stains and the rinse water as well as the nature
of the cells present (Fig. 2-12, B). Low pH, inadequate stain-
ing time, degraded stains, or excessive washing can result in
excessively pink-staining blood films. High pH, prolonged
staining, or insufficient washing can result in excessively blue-
staining blood films.30
Blood films should be fixed in methanol within 4 hours
(preferably within 1 hour) of preparation. If the methanol
contains more than 3% water, morphologic artifacts including
loss of cellular detail and vacuolation may be present.30 Blood
films may have an overall blue tint if stored unfixed for long
periods before staining or if the unfixed blood films are
exposed to formalin vapors, as occurs when blood films are
shipped to the laboratory in a package that also contains
formalin-fixed tissue. Blood films prepared from blood col- FIGURE 2 -13
lected with heparin as the anticoagulant have an overall Refractile inclusions in erythrocytes from a horse are artifacts resulting
magenta tint owing to the mucopolysaccharides present.30 from drying or fixation problems. Erythrocytes exhibit rouleaux, a normal
finding in horses. Wright-Giemsa stain.
Drying or fixation problems can result in variably shaped
refractile inclusions in the erythrocytes; these may be confused
with erythrocyte parasites (Fig. 2-13). The presence of stain obtained by longer staining procedures, but staining quality is
precipitation can make identification of leukocytes and blood improved by allowing the blood film to remain in the fixative
parasites difficult (Fig. 2-14). Precipitated stain may be present for several minutes. The Diff-Quik stain is classified as an
because the stain or stains needed to be filtered, the staining aqueous stain, even though the fixation is done in methanol,
procedure was too long, or the washing was not sufficient. because the component stains were prepared in water. A sig-
Carboxymethylcellulose has been infused into the peritoneal nificant limitation of the Diff-Quik stain and other aqueous
cavity of horses and cattle in an attempt to prevent abdominal stains, such as Hema 3 (Fisher Scientific, Pittsburgh, PA) and
adhesions after surgery. This material can appear as a precipi- the stain used in the automated stainer Aerospray 7120
tate between cells in blood that resembles stain precipitation (Westcore, Inc, Logan, UT), is that they do not stain basophil,
(Fig. 2-15).6 mast cell, or large granular leukocyte granules well (Fig.
Various rapid stains, such as Diff-Quik (Dade Behring 2-16).1 However, these aqueous stains are superior to conven-
Inc., Newark, DE), are available. The quality of Diff-Quik- tional methanolic Wright or Wright-Giemsa stains in the
stained blood films is generally somewhat lower than that staining of distemper inclusions in canine blood cells.1,26
A B
FIGURE 2 -16
Band basophils in the blood of a horse. A, Granules stain well with
Wright-Giemsa. B, Most granules are not stained with Diff-Quik stain.
FIGURE 2-14
Stain precipitation in blood from a dog. The two neutrophils present
might be mistaken for basophils because of the adherent precipitated
stain. Wright-Giemsa stain.
FIGURE 2 -17
Reticulocytes in dog blood. Four reticulocytes (with blue-staining mate-
rial) and three mature erythrocytes in blood from a dog with a regenera-
tive anemia. New methylene blue reticulocyte stain.
reticulocytes. The use of a Miller’s disc in one of the microscope

FIGURE 2-15
oculars saves time in performing the reticulocyte count.
Carboxymethylcellulose precipitation. The blue-to-purple precipitate
present between erythrocytes in this horse blood results from treatment The blue-staining aggregates or “reticulum” seen in reticu-
with carboxymethylcellulose. Wright-Giemsa stain. locytes (Fig. 2-17) does not occur as such in living cells but
results from the precipitation of ribosomal ribonucleic acid
Photograph of a stained blood film from a 1994 ASVCP slide review case (RNA; the same RNA that causes the bluish color seen in
submitted by M.J. Burkhard, M.A. Thrall, and G. Weiser. polychromatophilic erythrocytes) in immature erythrocytes
during the staining process.33 As a reticulocyte matures, the
number of ribosomes decreases until only small punctate (dot-
Reticulocyte Stains like) inclusions are observed in erythrocytes (punctate reticu-
Reticulocyte stains are commercially available (N. Brecher, locytes) stained with the reticulocyte stain (Fig. 2-18). To
Harleco, EMD Chemicals). Those wishing to prepare their reduce the chance that a staining artifact would result in
own stain can do so by dissolving 0.5 g of new methylene blue misclassifying a mature erythrocyte as a punctate reticulocyte
and 1.6 g of potassium oxalate in 100 mL of distilled water. when using a reticulocyte stain, the cell in question should
Following filtration, equal volumes of blood and stain are have two or more discrete blue granules that are visible without
mixed together in a test tube and incubated at room tempera- requiring fine focus adjustment. These inclusions should be
ture for 10 to 20 minutes. After incubation, blood films away from the cell margin to avoid confusion with hemotro-
are made and reticulocyte counts performed by examining phic mycoplasmas (formerly Haemobartonella organisms) or
1000 erythrocytes and determining the percentage that are small Heinz bodies.

Reticulocytes in cat blood. Three whole aggregate reticulocytes (contain- Reticulocytes in the blood of a dog. Five reticulocytes (with blue-staining
ing blue-staining aggregates of RNA) and half of an aggregate reticulo- material) in blood from a dog with a regenerative anemia in a new
cyte (far right) in blood from a cat with a markedly regenerative anemia. methylene blue-stained wet preparation. Note the difference in morphol-
A majority of the remaining cells are punctate reticulocytes containing ogy compared with reticulocytes stained with a standard reticulocyte
discrete dot-like inclusions. New methylene blue reticulocyte stain. stain (see Figure 2-17). New methylene blue-stained wet preparation.
In normal cats as well as in cats with a regenerative anemia,

the number of punctate reticulocytes is much greater than New Methylene Blue “Wet Mounts”
that seen in other species.2 This apparently occurs because A new methylene blue “wet mount” preparation can be used
the maturation (loss of ribosomes) of reticulocytes in cats is for rapid information concerning the number of reticulocytes,
slower than in other species. Consequently reticulocytes in platelets, and Heinz bodies present on a blood film. The stain
cats are classified as aggregate (if coarse clumping is observed) consists of 0.5% new methylene blue dissolved in 0.85% NaCl.
or punctate (if small individual inclusions are present). Per- One milliliter of formalin is added per 100 mL of stain as a
centages of both types should be reported. Based on compos- preservative. This stain is filtered after preparation and stored
ite results from several authors, normal cats generally have in dropper bottles. Alternatively, the stain may be stored in a
from 0% to 0.5% aggregate and 1% to 10% punctate reticu- plastic syringe with a 0.2-µm syringe filter attached so that
locytes as determined by manual means. Higher punctate the stain is filtered as it is used. Dry, unfixed blood films are
numbers of 2% to 17% have been reported using flow stained by placing a drop of stain between the coverslip and
cytometry.46 a glass slide. This preparation is not permanent and does not
The percentages of aggregate reticulocytes in cats directly stain mature erythrocytes or eosinophil granules. Punctate
correlate with the percentages of polychromatophilic erythro- reticulocytes are not demonstrated, but aggregate reticulocytes
cytes observed in blood films stained with Wright-Giemsa.2 appear as erythrocyte ghosts containing blue to purple granu-
Aggregate reticulocytes mature to punctate types in a day or lar material (Fig. 2-19). Platelets stain blue to purple, and
less. Several more days are required for maturation (total dis- Heinz bodies appear as refractile inclusions within erythro-
appearance of ribosomes) of punctate reticulocytes in cats.9,15 cyte ghosts. Although this staining method is not optimal for
In contrast to those of the cat, most reticulocytes in differential leukocyte counts, the number and type of leuko-
other species are of the aggregate type. Consequently no cytes present can be appreciated.
attempts are made to differentiate the stages of reticulocytes
in species other than the cat. In most species, the percentages Iron Stains
of reticulocytes directly correlate with the percentages of poly- An iron stain such as the Prussian blue stain is used to verify
chromatophilic erythrocytes observed on routinely stained the presence of iron-containing (siderotic) inclusions in blood
blood films. and bone marrow cells and to evaluate bone marrow iron
In examining a reticulocyte-stained blood film, one should stores. Smears may be sent to a commercial laboratory for this
also be careful not to confuse precipitated RNA with Heinz stain, or a stain kit can be purchased and applied in house
bodies. Heinz bodies are composed of denatured precipitated (Harleco Ferric Iron Histochemical Reaction Set, #6498693,
hemoglobin. They are spherical, stain pale blue with reticulo- EM Diagnostic Systems, Gibbstown, NJ). Iron-positive mate-
cyte stains, and are usually found at the periphery of the rial stains blue, in contrast to the pink color of the cells and
erythrocyte. the background when this stain is applied.
A B
FIGURE 2-20
Neutrophils containing hemosiderin (sideroleukocytes) in blood from a
dog with a hemolytic anemia. A, Wright-Giemsa stain. Hemosiderin
inclusions stain gray or brownish. B, Prussian blue stain. Hemosiderin
inclusions stain dark blue.
FIGURE 2 -21
The presence of focal areas of basophilic stippling within Autoagglutination of erythrocytes in blood from a dog with immune-
erythrocytes stained with Romanowsky-type blood stains sug- mediated hemolytic anemia and marked leukocytosis. Wright-Giemsa
stain.
gests that the stippling may contain iron. Iron-containing eryth-
rocytes are referred to as siderocytes. Neutrophils and monocytes
may also contain dark bluish-black or greenish iron-positive
particles within their cytoplasm when they are stained with
Romanowsky-type stains. Leukocytes containing iron-positive
inclusions have been called sideroleukocytes (Fig. 2-20).
Prussian blue stain applied to bone marrow aspirate smears
is a useful way to evaluate the amount of storage iron present
in the marrow. Minimal or no iron is expected in iron defi-
ciency anemia (although cats normally have no stainable iron
in the marrow), whereas normal or excess iron may be observed
in animals with hemolytic anemia and those with anemia
resulting from decreased erythrocyte production.
Cytochemical Stains
A variety of cytochemical stains—such as peroxidase,
chloroacetate esterase, alkaline phosphatase, and nonspecific
esterase—are utilized to classify cells in animals with acute
myeloid leukemias.22,33,50 Reactions vary not only by cell type
and stage of maturation but also by species.50 These stains are FIGURE 2 -22
done in a limited number of laboratories, and special training Leukocyte aggregate in blood from a dog. Leukocyte aggregates were
and/or experience is required to interpret the results. The present when EDTA was used as the anticoagulant but not when citrate
appearance of positive reactions also varies depending on the was used as the anticoagulant. Wright-Giemsa stain.
reagents used. Because of the complexities of the staining pro-
cedures and interpretation of results, minimal information on
cytochemical stains is presented in this book. As more antibod-
ies become available for immunophenotyping acute myeloid Blood films are generally examined following staining with
leukemias, the need for cytochemical stains will decrease.60 Romanowsky-type stains such as Wright or Wright-Giemsa.
These stains allow for the examination of erythrocyte, leuko-
cyte, and platelet morphology. Blood films should first be
EXA M I N AT I O N O F S TA I N ED scanned using a low power objective to estimate the total
BLOOD FILMS leukocyte count and to look for the presence of erythrocyte
autoagglutination (Fig. 2-21), leukocyte aggregates (Fig.
An overview and organized method of blood film examina- 2-22), platelet aggregates (Fig. 2-23), microfilaria (Fig. 2-24),
tion are presented here. Descriptions and photographs of and abnormal cells that might be missed during the differen-
normal and abnormal blood cell morphologies, inclusions, and tial leukocyte count. It is particularly important that the feath-
infectious agents are given in subsequent chapters. ered end of blood films made on glass slides be examined

Platelet aggregate in blood from a cow. Wright-Giemsa stain. Leukocytes concentrated in the feathered edge of a blood film from a
dog. The blood film was prepared using glass slides. Wright-Giemsa
stain.

Dirofilaria immitis microfilaria in blood from a cat with heartworm Platelet aggregate in the feathered edge of a blood film from a cat. The
disease. Wright-Giemsa stain. blood film was prepared using glass slides. Wright-Giemsa stain.
because leukocytes (Fig. 2-25) and platelet aggregates (Fig. Leukocyte Evaluation
2-26) may be concentrated in this area. Conversely, aggregates As a quality control measure, the number of leukocytes present
of cells tend to be in the center of blood films rather than at should be estimated to assure that the number present on the
the feathered edge when the coverslip blood-film method is slide is consistent with the total leukocyte count measured. If
used. 10× oculars and a 10× objective are used (100× magnification),
In examining a glass slide blood film, the blood film will the total leukocyte count in blood (cells per microliter) may
be too thick to evaluate blood cell morphology at the back of be estimated by determining the average number of leukocytes
the slide (Fig. 2-27, A) and too thin at the feathered edge, present per field and multiplying by 100 to 150. If a 20×
where cells are flattened (Fig. 2-27C). The optimal area for objective is used, the total leukocyte count may be estimated
evaluation is generally in the front half of the smear behind by multiplying the average number of leukocytes per field by
the feathered edge (Fig. 2-27, B). This area should appear as 400 to 600. The correction factor used may vary, depending
a well-stained monolayer (a field in which erythrocytes are on the microscope used. Consequently the appropriate cor-
close together with approximately half of the cells touching rection factors for the microscope being used should be deter-
each other). mined by performing estimates on a number of blood films
in which the total leukocyte counts have been accurately

determined. Manual differential leukocyte counts are inher-
ently imprecise, so it is also important to scan the blood film
to gain an appreciation for the relative distribution of leuko-
cyte types present prior to performing the differential leuko-
cyte count.34
Leukocytes tend to smudge in samples with high hemato-
crits, making it difficult to perform a differential leukocyte
count (Fig. 2-28, A). When this occurs, one can dilute a
portion of the blood sample with equal parts of plasma
or serum and make another blood film for examination
(Fig. 2-28, B). Obviously estimated cell counts will be pro-
portionally reduced in this dilute blood film.
A differential leukocyte count is done by identifying 200
A consecutive leukocytes using a 40× or 50× objective. Because
monocytes and other large leukocytes are pushed to the sides
and ends of wedge-prepared blood films, differential counts
are done by examining cells in a pattern that evaluates both
FIGURE 2-27
Selection of the appropriate area for microscopic examination of a slide
blood film prepared from a dog. A, Thick area in the back end of a blood
film. B, Optimal area for morphologic evaluation in the front half of the B
blood film. C, Thin area near the feathered edge of the blood film.
Erythrocytes are flattened to the extent that central pallor is not readily
apparent. Wright-Giemsa stain. FIGURE 2 -28
Blood films from a dog with a high hematocrit and toxic left shift prior
to (A) and after (B) dilution of the blood sample with serum. Leukocytes
are smudged and difficult to classify prior to blood sample dilution.
the edges and the center of the smear (Fig. 2-29).30 After the those from normal horses and cats may contain a low percentage
count is complete, the percentage of each leukocyte type of small, spherical nuclear remnants called Howell-Jolly bodies.
present is calculated and multiplied by the total leukocyte Rouleaux and the presence of Howell-Jolly bodies should be
count to get the absolute number of each cell type present per recorded on the hematology form when they appear in blood
microliter of blood. films from species in which these are not normal findings.
It is the absolute number of each leukocyte type that is Additional observations regarding erythrocyte morphol-
important. Relative values (percentages) can be misleading ogy, such as the degree of polychromasia (presence of poly-
when the total leukocyte count is abnormal. Let us consider chromatophilic erythrocytes), anisocytosis (variation in size),
two dogs, one with 7% lymphocytes and a total leukocyte and poikilocytosis (abnormal shape) should also be made.
count of 40,000/µL and the other with 70% lymphocytes and Polychromatophilic erythrocytes are reticulocytes that stain
a total leukocyte count of 4000/µL. The first would be said to bluish-red because of the combined presence of hemoglobin
have a “relative” lymphopenia and the second would be said (red-staining) and ribosomes (blue-staining). Abnormal
to have a “relative” lymphocytosis, but they would both have erythrocyte shapes should be classified as specifically as pos-
the same normal absolute lymphocyte count (2800/µL). sible, because specific shape abnormalities can help to deter-
Nucleated red blood cells (NRBCs) are counted along with mine the nature of a disorder that may be present. Examples
the leukocytes when leukocyte counts are done using manual of abnormal erythrocyte morphology include echinocytes,
methods or automated impedance cell counters. As encoun- acanthocytes, schistocytes, keratocytes, dacryocytes, ellipto-
tered during blood-film examination, the number of NRBCs cytes, eccentrocytes, and spherocytes. The number of abnormal
per 100 leukocytes should be tabulated and the total leukocyte cells should be reported in a semiquantitative fashion, such as
count corrected for the number of NRBCs present (see that shown in Table 2-1.68
formula below) prior to calculating the absolute cell counts
for each blood cell type. Platelets
Platelet numbers should be estimated as low, normal, or
Corrected leukocyte count = increased. Blood smears from most domestic animals nor-
( measured leukocyte count × 100 ) (1000 + NRBC ) mally average between 10 and 30 platelets per field examined
under 10× oculars and the 100× objective (1000× magnifica-
If leukocyte counts are performed with automated cell tion). As few as 6 platelets per 1000× field may be present in
counters using technology that can differentiate NRBCs from normal horse blood films.68 Platelet numbers may be esti-
leukocytes (e.g., morphologic characteristics of cells are deter- mated by multiplying the average number per field by 15,000
mined by using a laser beam), the above calculation will not to 20,000 to get the approximate number of platelets per
be necessary. However, the number of NRBCs present should microliter of blood.61,68 While special attention will be given
still be recorded. to the estimation of platelet numbers in animals with hemo-
The presence of abnormal leukocyte morphology—such as static diatheses, it is important to routinely estimate the
toxic cytoplasm in neutrophils or increased reactive lympho- platelet numbers on blood films, because many animals with
cytes (e.g., more than 5% reactive lymphocytes)—should be thrombocytopenia exhibit no evidence or past history of
recorded on the hematology report form. The frequency of
toxic neutrophils is reported as few (5% to 10%), moderate
(11% to 30%), or many (more than 30%) and the severity of
toxic change is recorded as 1+ to 3+ (Box 2-1).68
Box 2-1 Semiquantitative Evaluation of
Erythrocyte Morphology Toxic Changes in the Cytoplasm
Erythrocyte morphology should be examined and recorded as of Neutrophils
either normal or abnormal. Erythrocytes on blood films from Neutrophils with Toxic Change
normal horses, cats, and pigs often exhibit rouleaux formation; Few 5-10 (%)
Moderate 11-30 (%)
Many >30 (%)
Severity of Toxic Change in Cytoplasm
Döhle bodiesa 1+
Mildly basophilic 1+
633449
Date
Moderately basophilic with Döhle bodies 2+

Moderately basophilic and foamy 2+
Dark blue-gray and foamyb 3+
Basophilic with toxic granulesb 3+
FIGURE 2-29 a
One or two Döhle bodies are sometimes seen in a few neutrophils from cats
Patterns of slide blood film examination (marked in white) that may be that do not exhibit signs of illness.
used to improve the accuracy of differential leukocyte counts. b
May also contain Döhle bodies.
Table 2-1
Semiquantitative Evaluation of Erythrocyte Morphology Based on Average Number of Abnormal Cells per
1000× Microscopic Monolayer Fielda
1+ 2+ 3+ 4+
Anisocytosis
Dog 7-15 16-20 21-29 >30
Cat 5-8 9-15 16-20 >20
Cattle 10-20 21-30 31-40 >40
Horse 1-3 4-6 7-10 >10
Polychromasia
Dog 2-7 8-14 15-29 >30
Cat 1-2 3-8 9-15 >15
Cattle 2-5 6-10 11-20 >20
Horse Rarely observed — — —
Hypochromasia and Shapes

Hypochromasiaa 1-10 11-50 51-200 >200
Poikilocytosisa 3-10 11-50 51-200 >200
Codocytes (dogs) 3-5 6-15 16-30 >30
Spherocytesb 5-10 11-50 51-150 >150
Echinocytesb 5-10 11-100 101-250 >250
Other shapesc 1-2 3-8 9-20 >20
a
A monolayer field is defined as a field in which erythrocytes are close together with approximately half touching each other. In severely anemic animals, such
monolayers may not be present. When erythrocytes are generally not touching (e.g., tend to be separated by the distance of one cell diameter), the number of
erythrocytes with morphologic abnormalities are counted for two fields.
b
The same parameters are used for all species.
c
Parameters are used for acanthocytes, schistocytes, keratocytes, elliptocytes, dacrocytes, drepanocytes, and stomatocytes in all species.
From Weiss DJ. Uniform evaluation and semiquantitative reporting of hematologic data in veterinary medicine. Vet Clin Pathol. 184;13:27-31.
bleeding tendencies. If a thrombocytopenia is suspected, it blood storage.7 Nuclear swelling may be present in leukocytes,
should be confirmed with a platelet count. Dogs and cats have a process that can result in an increased percentage of bands
larger platelets than do horses and ruminants. Platelets contain on microscopic examination.35 Neutrophil cytoplasmic vacu-
magenta-staining granules, but these granules generally stain olation develops, which may be confused with toxic changes.21
poorly in horses. The presence of abnormal platelet morphol- Leukocytes that undergo programmed cell death (apopto-
ogy (large or hypogranular platelets) should also be recorded sis) exhibit pyknosis and karyorrhexis (Fig. 2-30). Pyknosis
on the hematology form. involves shrinkage or condensation of a cell with increased
nuclear compactness or density. Karyorrhexis refers to the
Degenerative Changes in Blood Samples subsequent nuclear fragmentation. It may not be possible to
Degenerative changes are apparent in blood samples within a determine the cell of origin. These abnormalities can occur in
few hours after collection; consequently blood films should be vivo but are more commonly associated with prolonged or
made and stained and blood cell counts performed as soon as inadequate sample storage before blood films are made.30
possible after collection. Blood samples should be refrigerated With excessive storage, all blood cell types will lyse, resulting
if tests cannot be performed within a couple of hours. CBCs in cytopenias.
are often performed on day-old blood that has been kept
refrigerated and submitted to a commercial laboratory, but Infectious Agents or Inclusions of Blood Cells
some changes will already be present. Various progressive Blood films are examined for the presence of infectious agents
changes may be observed depending on the animal species or intracellular inclusions using the 100× objective. Infectious
and the time and temperature of storage of blood samples.7,18,31 agents or inclusions that may be seen in blood cells include
Erythrocytes can swell, which results in increased MCV and Howell-Jolly bodies, Heinz bodies (unstained), basophilic
HCT and decreased MCHC within 12 hours. Platelets tend stippling, canine distemper inclusions, siderotic inclusions,
to aggregate and degranulate, resulting in lower automated Döhle bodies, Babesia species, Cytauxzoon felis, hemotrophic
platelet counts and higher MPV values.18,45 Electronically Mycoplasma (formerly Haemobartonella) species, Ehrlichia
determined differential counts become less accurate with species, Anaplasma species, Hepatozoon species, and Theileria
A B A B
FIGURE 2 -31
Mitotic cells in blood. A, Mitotic cell in anaphase in blood from a cat
with erythroleukemia (AML-M6). B, Mitotic cell (presumably lym-
phoid) in prophase in blood from a horse with equine infectious anemia.
Wright-Giemsa stain.
C D
FIGURE 2-30
Pyknotic and karyorrhexic cells in blood. A, Pyknotic cell with con-
densed chromatin in blood from a dog with a toxic left shift. B, Pyknosis
and karyorrhexis of a cell in blood from a dog with dirofilariasis.
C, Pyknosis and karyorrhexis of a cell in blood from a dog with
acute monocytic leukemia (AML-M5). D, Pyknosis and karyorrhexis of
a cell in blood from a cow with leukemic lymphoma. Wright-Giemsa
stain. A B
FIGURE 2 -32
Free nuclei in blood. A, Free nucleus in blood from a dog with chronic
species. The appearance of these agents and inclusions is dis- lymphocytic leukemia. B, Free nucleus with a distorted net-like structure
cussed in subsequent chapters. (“basket cell”) in blood from a cat. Wright-Giemsa stain.
Miscellaneous Cells and Parasites in Blood

Degenerative cells, mitotic cells, vascular lining cells, and
other cells not typically seen in blood may occasionally be a basket cell is not truly a cell but only the distorted nucleus
recognized during blood film examination. Parasites and bac- of a cell. Lymphocytes are the most likely blood cell type to
teria that are not associated with blood cells may also be seen lyse during blood-film preparation.
in blood. However, bacterial rods and cocci between cells are
usually the result of contaminated stain. Endothelial Cells
Spindle-shaped endothelial cells with elongated nuclei may
Mitotic Cells sometimes be seen in blood films (Fig. 2-33). Endothelial cells
Mitotic cells may be present in the blood of animals with line vessels and may become dislodged as the needle enters
malignant neoplasia (Fig. 2-31, A), but they may also occur in the vein during blood sample collection.
nonneoplastic disorders, such as lymphocytes undergoing
blast transformation (Fig. 2-31, B), nucleated erythroid pre- Megakaryocytes
cursors in regenerative anemia, and activated mononuclear Megakaryocytes are multilobulated, platelet-producing giant
phagocytes. cells that lie against the outside of vascular sinuses in bone
marrow (see Thrombopoiesis section in Chapter 3 for more
Free Nuclei details). Cytoplasmic processes of mature megakaryocytes
When cells are lysed during blood-film preparation, free extend into the sinus lumen, where they develop into pro-
nuclei (nuclei without cytoplasm) may be seen (Fig. 2-32, A). platelets and subsequently individual platelets. Sometimes
When a free nucleus is spread thin on the blood film, it whole megakaryocytes enter vascular sinuses, accounting for
appears as a net-like pinkish structure, which has been referred the rare recognition of these cells in blood films from animals
to as a “basket cell” (Fig. 2-32, B). This is a misnomer, because (Fig. 2-34).51 Megakaryocytes are more easily found by
FIGURE 2 -34
Mature megakaryocyte in blood from dog with an abscess and accom-
panying toxic left shift. Wright-Giemsa stain.
FIGURE 2-33
Courtesy of Heather Wamsley.
Two spindle-shaped endothelial cells with elongated nuclei in blood
from a cow. These cells were likely dislodged from the vessel wall during
blood sample collection. Wright-Giemsa stain.
A B C
FIGURE 2 -35
Dwarf megakaryocytes in blood from dogs with myeloid neoplasms. A, Dwarf megakaryocyte with single
nucleus in blood from a dog with chronic myeloid leukemia (CML). B, Dwarf megakaryocyte with two nuclei
in blood from a dog with CML. C, Dwarf megakaryocyte in blood from a dog with AML-M7. Wright-
Giemsa stain.
examination of blood buffy coat smears. Those reaching the Microfilaria

blood are quickly trapped in lung capillaries, where continued Microfilariae (nematode larvae) that might be observed
platelet production may occur. include Dirofilaria immitis (see Fig. 2-24) and Dirofilaria
Dwarf megakaryocytes are smaller than normal mature repens in dogs, cats, and wild canids, Dipetalonema reconditum
megakaryocytes and have decreased nuclear ploidy, but their in dogs, and Setaria species in cattle and horses.74
cytoplasm generally contains granules and appears similar to
that of blood platelets (Fig. 2-35). Dwarf megakaryocytes are Trypanosoma Species
common in the bone marrow of animals with myeloid neo- Various Trypanosoma species may be seen in blood. These
plasms but are only rarely seen in blood. elongated, flagellated protozoa cause important diseases of

Trypanosoma theileri in blood from a 3-day-old female Angus calf. Two Borrelia turicatae spirochetes in blood from a North Central Florida
Wright stain. dog. Wright-Giemsa stain.
Photograph of a stained blood film from a 1989 ASVCP slide review case
submitted by H. Bender, A. Zajak, G. Moore, and G. Saunders.
bacteria within neutrophils indicates that the bacteria are
likely of clinical significance. Spirochetes have been seen in
blood from dogs with Borrelia infections.55 Borrelia burgdorferi
does not usually result in microscopically detectable spiro-
chetemia.56 However, relapsing fever spirochetes, Borrelia
turicatae and Borrelia hermsii, have been readily identified
in stained blood films from dogs in the United States (Fig.
2-38).5,57,58,70
R EF ER EN C E S
A B
1. Allison RW, Velguth KE. Appearance of granulated cells in blood films stained by
automated aqueous versus methanolic Romanowsky methods. Vet Clin Pathol.
FIGURE 2-37 2010;39:99-104.
A, B, Trypanosoma cruzi organisms in blood from a dog. Wright stain. 2. Alsaker RD, Laber J, Stevens JB, et al. A comparison of polychromasia and reticulocyte
counts in assessing erythrocyte regenerative response in the cat. J Am Vet Med Assoc.
Photographs of a stained blood film from a 2006 ASVCP slide review case 1977;170:39-41.
3. Barr SC. American trypanosomiasis. In: Greene CE, ed. Infectious Diseases of the Dog
submitted by P.K. Patten and J.M. Meinkoth.
and Cat. 3rd ed. St. Louis, MO: Saunders Elsevier; 2006:676-681.
4. Blaisdell FS, Dodds WJ. Evaluation of two microhematocrit methods for quantitating
plasma fibrinogen. J Am Vet Med Assoc. 1977;171:340-342.
livestock outside the United States,74 but the species seen in 5. Breitschwerdt EB, Nicholson WL, Kiehl AR, et al. Natural infections with Borrelia
spirochetes in two dogs in Florida. J Clin Microbiol. 1994;32:352-357.
cattle (T. theileri) in the United States is usually nonpatho- 6. Burkhard MJ, Baxter G, Thrall MA. Blood precipitate associated with intra-abdominal
genic (Fig. 2-36).16,39 Many dogs are infected with T. cruzi in carboxymethylcellulose administration. Vet Clin Pathol. 1996;25:114-117.
the United States, but organisms are rarely seen in blood and 7. Clark P, Mogg TD, Tvedten HW, et al. Artifactual changes in equine blood following
storage, detected using the Advia 120 hematology analyzer. Vet Clin Pathol.
most cases are subclinical (Fig. 2-37). When present, clinical 2002;31:90-94.
forms of disease have principally involved heart or neural 8. Corn SC, Wellman ML, Burkhard MJ, et al. IgM paraprotein interference with hemo-
dysfunction.3 globin measurement using the CELL-DYN 3500. Vet Clin Pathol. 2008;37:61-65.
9. Cramer DV, Lewis RM. Reticulocyte response in the cat. J Am Vet Med Assoc.
1972;160:61-67.
Bacteria 10. Deol I, Hernandez AM, Pierre RV. Ethylenediamine tetraacetic acid-associated leuko
Various bacterial species may be present in blood films. It is agglutination. Am J Clin Pathol. 1995;103:338-340.
11. Dintenfass L, Kammer S. Re-evaluation of heat precipitation method for plasma fibrino-
important to verify that these are not contaminants, especially gen estimation: effect of abnormal proteins and plasma viscosity. J Clin Pathol.
during the staining procedure. The presence of phagocytized 1976;29:130-134.
12. Eckersall PD. Proteins, proteomics, and the dysproteinemias. In: Kaneko JJ, Harvey JW, 43. Orlov SN, Aksentsev SL, Kotelevtsev SV. Extracellular calcium is required for the
Bruss ML, eds. Clinical Biochemistry of Domestic Animals. 6th ed. San Diego: Academic maintenance of plasma membrane integrity in nucleated cells. Cell Calcium.
Press; 2008:117-155. 2005;38:53-57.
13. Engelking LR. Evaluation of equine bilirubin and bile acid metabolism. Comp Cont Ed 44. Pankraz A, Bauer N, Moritz A. Comparison of flow cytometry with the Sysmex XT2000iV
Pract Vet. 1989;11:328-336. automated analyzer for the detection of reticulated platelets in dogs. Vet Clin Pathol.
14. Espanol I, Pedro C, Remacha AF. Heinz bodies interfere with automated reticulocyte 2009;38:30-38.
counts. Haematologica. 1999;84:373-374. 45. Pastor J, Cuenca R, Velarde R, et al. Evaluation of a hematology analyser with canine
15. Fan LC, Dorner JL, Hoffman WE. Reticulocyte response and maturation in experimental and feline blood. Vet Clin Pathol. 1997;26:138-147.
acute blood loss anemia in the cat. J Am Anim Hosp Assoc. 1978;14:219-224. 46. Perkins PC, Grindem CB, Cullins LD. Flow cytometric analysis of punctate and
16. Farrar RG, Klei TR. Prevalence of Trypanosoma theileri in Louisiana cattle. J Parasitol. aggregate reticulocyte responses in phlebotomized cats. Am J Vet Res. 1995;56:
1990;76:734-736. 1564-1569.
17. Fry MM, Kirk CA. Reticulocyte indices in a canine model of nutritional iron deficiency. 47. Pierson FW. Laboratory techniques for avian hematology. In: Feldman BF, Zinkl JG, Jain
Vet Clin Pathol. 2006;35:172-181. NC, eds. Schalm’s Veterinary Hematology. 5th ed. Philadelphia: Lippincott Williams &
18. Furlanello T, Tasca S, Caldin M, et al. Artifactual changes in canine blood following Wilkins; 2000:1145-1147.
storage, detected using the ADVIA 120 hematology analyzer. Vet Clin Pathol. 48. Porter RE, Weiser MG. Effect of immune-mediated erythrocyte agglutination on analy-
2006;35:42-46. sis of canine blood using a multichannel blood cell counting system. Vet Clin Pathol.
19. George JW. The usefulness and limitations of hand-held refractometers in veterinary 1990;19:45-50.
laboratory medicine: an historical and technical review. Vet Clin Pathol. 2001; 49. Ragan HA. Platelet agglutination induced by ethylenediaminetetraacetic acid in blood
30:201-210. samples from a miniature pig. Am J Vet Res. 1972;33:2601-2603.
20. Ginzinger DG, Clee SM, Dallongeville J, et al. Lipid and lipoprotein analysis of cats 50. Raskin RE. Cytochemical staining. In: Weiss DJ, Wardrop KJ, eds. Schalm’s Veterinary
with lipoprotein lipase deficiency. Eur J Clin Invest. 1999;29:17-26. Hematology. 6th ed. Ames, IA: Wiley-Blackwell; 2010:1141-1156.
21. Gossett KA, Carakostas MC. Effect of EDTA on morphology of neutrophils of healthy 51. Roszel J, Prier JE, Koprowska I. The occurrence of megakaryocytes in the peripheral
dogs and dogs with inflammation. Vet Clin Pathol. 1984;13:22-25. blood of dogs. J Am Vet Med Assoc. 1965;147:133-137.
22. Grindem CB. Blood cell markers. Veterinary Clinics of North America, Small Animal 52. Sato S, Hirayama K, Koyama A, et al. Pseudoreticulocytosis in a patient with hemo-
Practice. 1996;26:1043-1064. globin Koln due to autofluorescent erythrocytes enumerated as reticulocytes by the
23. Gunn-Moore DA, Watson TD, Dodkin SJ, et al. Transient hyperlipidaemia and anaemia Cell-Dyn 4000. Lab Hematol. 2004;10:65-67.
in kittens. Vet Rec. 1997;140:355-359. 53. Savage RA. Pseudoleukocytosis due to EDTA-induced platelet clumping. Am J Clin
24. Harenberg J, Malsch R, Piazolo L, et al. Preferential binding of heparin to granulocytes Pathol. 1984;81:317-322.
of various species. Am J Vet Res. 1996;57:1016-1020. 54. Schaefer DM, Priest H, Stokol T, et al. Anticoagulant-dependent in vitro hemagglutina-
25. Harr KE, Raskin RE, Heard DJ. Temporal effects of 3 commonly used anticoagulants tion in a cat. Vet Clin Pathol. 2009.
on hematologic and biochemical variables in blood samples from macaws and Burmese 55. Schalm OW. Uncommon hematologic disorders: spirochetosis, trypanosomiasis, leish-
pythons. Vet Clin Pathol. 2005;34:383-388. maniasis, and Pelger-Huet anomaly. Can Pract. 1979;6:46-49.
26. Harvey JW. Hematology tip—stains for distemper inclusions. Vet Clin Pathol. 56. Schwan TG, Burgdorfer W, Rosa PA. Borrelia. In: Murray PR, Baron EJ, Pfaller MA, et
1982;11:12. al, eds. Manual of Clinical Microbiology. 7th ed. Washington, DC: ASM Press;
27. Harvey JW. Pathogenesis, laboratory diagnosis, and clinical implications of 1999:746-758.
erythrocyte enzyme deficiencies in dogs, cats, and horses. Vet Clin Pathol. 2006;35: 57. Schwan TG, Raffel SJ, Schrumpf ME, et al. Phylogenetic analysis of the spirochetes
144-156. Borrelia parkeri and Borrelia turicatae and the potential for tick-borne relapsing fever
28. Harvey JW. The erythrocyte: physiology, metabolism and biochemical disorders. in Florida. J Clin Microbiol. 2005;43:3851-3859.
In: Kaneko JJ, Harvey JW, Bruss ML, eds. Clinical Biochemistry of Domestic Animals. 58. Stevenson C, Schwan T. Borrelia hermsii spirochetemia in a dog (abstract). Proc XIVth
6th ed. San Diego: Academic Press; 2008:173-240. Cong Int Soc Anim Clin Pathol. 2010;18.
29. Hinchcliff KW, Kociba GJ, Mitten LA. Diagnosis of EDTA-dependent pseudothrombo- 59. Stokol T, Erb HN. A comparison of platelet parameters in EDTA- and citrate-
cytopenia in a horse. J Am Vet Med Assoc. 1993;203:1715-1716. anticoagulated blood in dogs. Vet Clin Pathol. 2007;36:148-154.
30. Houwen B. Blood film preparation and staining procedures. Lab Hematol. 2000; 60. Tasca S, Carli E, Caldin M, et al. Hematologic abnormalities and flow cytometric
6:1-7. immunophenotyping results in dogs with hematopoietic neoplasia: 210 cases (2002-
31. Ihedioha JI, Onwubuche RC. Artifactual changes in PCV, hemoglobin concentration, 2006). Vet Clin Pathol. 2009;38:2-12.
and cell counts in bovine, caprine, and porcine blood stored at room and refrigerator 61. Tasker S, Cripps PJ, Macklin AJ. Estimation of platelet counts on feline blood smears.
temperatures. Vet Clin Pathol. 2007;36:60-63. Vet Clin Pathol. 1999;28:42-45.
32. Jain NC. Schalm’s Veterinary Hematology. 4th ed. Philadelphia: Lea & Febiger; 62. Tennant BC, Center SA. Hepatic function. In: Kaneko JJ, Harvey JW, Bruss ML, eds.
1986. Clinical Biochemistry of Domestic Animals. 6th ed. San Diego, CA: Academic Press;
33. Jain NC. Essentials of Veterinary Hematology. Philadelphia: Lea & Febiger; 1993. 2008:379-412.
34. Kjelgaard-Hansen M, Jensen AL. Is the inherent imprecision of manual leukocyte 63. Tvedten H. Advanced hematology analyzers. Interpretation of results. Vet Clin Pathol.
differential counts acceptable for quantitative purposes? Vet Clin Pathol. 2006;35: 1993;22:72-80.
268-270. 64. Waitt LH, Cebra CK. Characterization of hypertriglyceridemia and response to treat-
35. Knoll JS. Clinical automated hematology systems. In: Feldman BF, Zinkl JG, Jain NC, ment with insulin in llamas and alpacas: 31 cases (1995-2005). J Am Vet Med Assoc.
eds. Schalm’s Veterinary Hematology. 5th ed. Philadelphia: Lippincott Williams & 2008;232:1362-1367.
Wilkins; 2000:3-11. 65. Waitt LH, Cebra CK. Characterization of hypertriglyceridemia and response to treat-
36. Lilliehook I, Tvedten H. Validation of the Sysmex XT-2000iV hematology system for ment with insulin in horses, ponies, and donkeys: 44 cases (1995-2005). J Am Vet Med
dogs, cats, and horses. I. Erythrocytes, platelets, and total leukocyte counts. Vet Clin Assoc. 2009;234:915-919.
Pathol. 2009;38:163-174. 66. Walencik J, Witeska M. The effects of anticoagulants on hematological indices and
37. Lumsden JH. Quality control. In: Feldman BF, Zinkl JG, Jain NC, eds. Schalm’s blood cell morphology of common carp (Cyprinus carpio L.). Comp Biochem Physiol C
Veterinary Hematology. 5th ed. Philadelphia: Lippincott Williams & Wilkins; 2000: Toxicol Pharmacol. 2007;146:331-335.
16-19. 67. Wardrop KJ. The Coombs’ test in veterinary medicine: past, present, future. Vet Clin
38. March H, Barger A, McCullough S, et al. Use of the ADVIA 120 for differentiating Pathol. 2005;34:325-334.
extracellular and intracellular hemoglobin. Vet Clin Pathol. 2005;34:106-109. 68. Weiss DJ. Uniform evaluation and semiquantitative reporting of hematologic data in
39. Matthews DM, Kingston N, Maki L, et al. Trypanosoma theileri Laveran, 1902, in veterinary laboratories. Vet Clin Pathol. 1984;13:27-31.
Wyoming cattle. Am J Vet Res. 1979;40:623-629. 69. Werner LL, Christopher MM, Snipes J. Spurious leukocytosis and abnormal WBC
40. McSherry BJ, Lumsden JH, Baird JD. Hyperbilirubinemia in sick cattle. Can Vet J. histograms associated with Heinz bodies (abstract). Vet Clin Pathol. 1997;26:
1984;48:237-240. 20.
41. Moraglio D, Banfi G, Arnelli A. Association of pseudothrombocytopenia and pseudo- 70. Whitney MS, Schwan TG, Sultemeier KB, et al. Spirochetemia caused by Borrelia
leukopenia: evidence for different pathogenic mechanisms. Scand J Clin Lab Invest. turicatae infection in 3 dogs in Texas. Vet Clin Pathol. 2007;36:212-216.
1994;54:257-265. 71. Wills TB, Wardrop KJ. Pseudothrombocytopenia secondary to the effects of EDTA in
42. Muro J, Cuenca R, Pastor J, et al. Effects of lithium heparin and tripotassium EDTA on a dog. J Am Anim Hosp Assoc. 2008;44:95-97.
hematologic values of Hermann’s tortoises (Testudo hermanni). J Zoo Wildl Med. 72. Xenoulis PG, Steiner JM. Lipid metabolism and hyperlipidemia in dogs. Vet J.
1998;29:40-44. 2009.
73. Xenoulis PG, Suchodolski JS, Levinski MD, et al. Serum liver enzyme activities in 76. Zandecki M, Genevieve F, Gerard J, et al. Spurious counts and spurious results
healthy Miniature Schnauzers with and without hypertriglyceridemia. J Am Vet Med on haematology analysers: a review. Part II: white blood cells, red blood cells,
Assoc. 2008;232:63-67. haemoglobin, red cell indices and reticulocytes. Int J Lab Hematol. 2007;29:
74. Zajac A, Conboy GA, Sloss MW. Veterinary Clinical Parasitology. 7th ed. Hoboken, NJ: 21-41.
Wiley-Blackwell; 2006. 77. Zelmanovic D, Hetherington EJ. Automated analysis of feline platelets in whole blood,
75. Zandecki M, Genevieve F, Gerard J, et al. Spurious counts and spurious results on including platelet count, mean platelet volume, and activation state. Vet Clin Pathol.
haematology analysers: a review. Part I: platelets. Int J Lab Hematol. 2007;29:4-20. 1998;27:2-9.
C H A P T E R
3
Hematopoiesis
OV ERV I EW Organization of Bone Marrow

Sites of Blood Cell Production Bone marrow develops in mammals during the second trimes-
In mammals, primitive hematopoiesis begins outside the ter.21 Rudimentary fetal bone is initially filled with cartilage.
body of the embryo in the yolk sac and shortly thereafter Chondrocytes hypertrophy and promote mineralization of the
within the aorta-gonad-mesonephros (AGM) region of the cartilage matrix in the center of the rudimentary bone. This is
embryo.42,130,139 Small clusters of hematopoietic stem cells followed by the entry of progenitor cells, which develop into
(HSCs) have been identified attached to the endothelium of chondroclasts that partially degrade the mineralized cartilage
the yolk sac and the dorsal aorta. These HSCs and the associ- and form bone marrow spaces colonized by incoming blood
ated endothelial cells are produced by common embryonic vessels.23,131 Osteoblast progenitors enter the space created,
stem cells known as hemangioblasts.34,74,150 In addition to adhere to remaining cartilage, develop into mature osteoblasts,
HSCs, committed erythroid and megakaryocytic progenitor and begin the formation of bony trabeculae. Vascular sinuses
cells, primitive erythrocytes (large nucleated cells containing and extravascular mesenchymal cells subsequently form a con-
embryonal hemoglobin), large primitive reticulated platelets, nective tissue meshwork within which HSCs originating from
and rare primitive macrophages are also produced in the yolk the liver (and probably the spleen) bind, proliferate, and dif-
sacs of rodents and humans.98,136 Notably, these primitive ferentiate, ultimately producing circulating blood cells.131
macrophages appear to develop directly from progenitor cells When these structures are fully developed, blood is supplied
in the yolk sac without passing through a monocyte stage.14 to the bone marrow by nutrient arteries and periosteal capil-
Definitive erythropoiesis, prominent megakaryocytopoiesis, laries (Fig. 3-2).7
and limited leukocyte production also occur in the yolk sacs The stroma of the marrow is a connective tissue consisting
of cats, with hematopoiesis persisting longer during gestation of stromal cells (fibroblast-like cells, also called reticular cells),
than it does in rodents and humans.134 The AGM region adipocytes, vascular elements (endothelial cells and myocytes),
transiently supports the development of HSCs and some neural elements, and extracellular matrix (ECM), with the
committed hematopoietic progenitor cells (HPCs), but rec- arrangement creating both intravascular and extravascular
ognizable blood cells are not produced in the AGM.93 spaces (Figs. 3-3, 3-4).146,147 In postnatal mammals, blood cells
Sites of blood cell production shift during embryonic and are continuously produced within the extravascular spaces of
fetal development as optimal microenvironments are produced bone marrow. Leukocytes are also produced within the extra-
in various tissues (Fig. 3-1).102 The liver and, to a lesser extent, vascular spaces of bone marrow in birds, but erythrocytes and
the spleen become the major hematopoietic organs by midges- thrombocytes are produced within the vascular spaces of the
tation in the fetus.129,134 Current evidence suggests that the avian marrow.20
AGM is more important than the yolk sack in providing HSCs This specialized arrangement of the marrow vasculature is
to seed the liver and spleen, but the relative importance of each important in the organization of intramedullary hematopoi-
area in embryonic and fetal hematopoiesis remains to be clari- etic microenvironments, as marrow endothelial cells are
fied.102 Blood cell production begins in bone marrow and lym- actively involved in the regulation of transendothelial (not
phoid organs during midgestation in mammals, with nearly all interendothelial) movement of hematopoietic cells and blood
blood cells being produced in these organs at the time of cells between the extravascular hematopoietic space and
birth.134 Blood cells are produced in the bone marrow of peripheral blood.92,114 Together, endothelial cells and stromal
adult birds20; the bone marrow and sometimes the spleen cells produce the ECM, which consists of collagen fibers,
of adult reptiles30; the kidney, liver, spleen, and/or bone marrow various macromolecules capable of binding cells, and basal
of amphibians5,42; and the kidney and/or spleen of fish.42,48 laminae of the sinuses.100,109 The marrow stromal cells have
33
100%
80%
Yolk sac
60% Bone
Liver marrow
40%
20% Spleen
0%
10 15 20 25 30 35 40 45 50 55 60 65
Gestation days
FIGURE 3-1
Sites of definitive hematopoiesis during prenatal development in cats.
Percentages represent relative contributions of sites to definitive blood
cell production. Lymphoid populations also develop in lymph nodes and
thymus beginning in midgestation (not shown).
Redrawn from Tiedemann K, van Ooyen B. Prenatal hematopoiesis and blood FIGURE 3 -2
characteristics of the cat. Anat Embryol (Berl). 1978;153:243-267. Anatomy and circulation of the bone marrow. Periosteum (p), cortical
bone (cb), nutrient foramen (nf ), nutrient artery (na), nutrient vein (nv),
central longitudinal artery (cla), central longitudinal vein (clv), periosteal
capillaries (pc), arteriole (a), sinuses (s), hematopoietic compartment (h),
extensively branched cytoplasmic processes and, along with anastomosis of the nutrient capillaries and sinuses (1), anastomosis of
the fibers that they produce, provide structural support for the the nutrient artery capillaries and periosteal capillaries (2), anastomosis
marrow (Fig. 3-5).147 These stromal cells have generally been of the periosteal capillaries and sinuses (3).
considered to be fibroblast-like, but they also display smooth
muscle characteristics in culture and have been classified as From Alsaker RD. The formation, emergence, and maturation of the reticulo-
cyte: a review. Vet Clin Pathol. 1977;6(3):7-12.
myofibroblasts by some investigators.122 The particular stromal
cells that support the endothelium of the venous sinuses are
termed adventitial stromal cells (Fig. 3-6).147 Granulopoiesis
also occurs primarily on the surface of stromal cells.125 Adi- Hematopoietic Stem Cells and Progenitor Cells
pocytes develop from mesenchymal stem cells and may share Beginning in midgestation and continuing throughout post-
common hematopoietic functions with stromal cells.47 Auto- natal life, mammalian blood cells are produced continuously
nomic nerves occur in bone marrow. Their function is not from HSCs within the extravascular spaces of the bone
clear, but direct and indirect effects of the sympathetic nervous marrow. HSCs are capable of proliferation; they exhibit long-
system on hematopoietic stem cell and hematopoietic pro- term self-renewal and differentiation. HSCs replicate only
genitor cell proliferation and motility have been described.61 once every 8 to 10 weeks.2 The term hematopoietic progenitor
In addition to hematopoietic cells and developing blood cell (HPC) refers to cells that form colonies in bone marrow
cells, a number of accessory cells involved in regulating hema- culture like HSCs but do not have long-term self-renewal
topoiesis reside within the extravascular space of mammalian capacities. HSCs and HPCs are mononuclear cells that cannot
bone marrow. These accessory cells include macrophages, be distinguished morphologically from lymphocytes. The
mature lymphocytes, and natural killer (NK) cells.12,31,33,85 presence of a transmembrane glycoprotein termed cluster of
Erythrocyte development occurs in close association with differentiation antigen 34 (CD34) has been used to identify
marrow macrophages.26 HSCs and early HPCs, but some HSCs (possibly inactive
In contrast to other organs such as skin and intestine, ones) lack CD34.43 In addition, CD34 is also present on the
where continuous new cell production occurs throughout life, surface of nonhematopoietic stem cells and vascular endo
hematopoietic cells and their progeny in bone marrow are not thelial cells.72,149 CD34 is believed to play a role in cell
arranged in stratified layers of progressively more differenti- adhesion.43
ated cells. Although some segregation of cell types may be The most primitive HSC has the capacity to differentiate
visualized by microscopic examination of stained bone marrow into HPCs of all blood cell lineages and several cell types in
sections, the overall impression is that bone marrow contains tissue. The frequency of HSCs in the marrow is estimated to
an unstructured mixture of cells of different lineages and be less than 0.01% of nucleated marrow cells in adult mice
stages of maturation. Nonetheless, hematopoietic cells develop and less than 0.0001% of nucleated marrow cells in adult cats.2
in specialized microenvironmental niches within the bone HSCs produce HPCs that can give rise to one or more blood
marrow. cell types. Thus, HPCs are much more numerous in marrow
C ha p ter 3 n Hematopoiesis 35
end
SINUS HEMATOPOIETIC
ARTER
CAPIL COMPARTMENTS
fat
cell end
meg
SINUS
emp
adv
end
SINUS
SINUS
ARTERY SINUS
CENTRAL adv
LONGITUDINAL end
VEIN
eryth islet
FIGURE 3-3
Schematic view of a cross section of bone marrow near the central longitudinal vein. Hematopoietic cells lie
in the hematopoietic compartment between the vascular sinuses that drain into the central vein. The sinus
wall consists of endothelial cells (end), a basement membrane, and, in some areas, adventitial stromal cells
(adv). Megakaryocytes (meg) lie against the outside of the vascular sinus wall and discharge proplatelets
directly into the vascular lumen through apertures in the sinus wall. Erythroid cells are shown developing in
an erythroid islet (eryth islet) around a central macrophage. Emperipolesis (emp), the entry of megakaryocyte
cytoplasm by other cells, is occasionally observed.
From Weiss L. The Blood Cells and Hematopoietic Tissues. New York: Elsevier; 1984.
than are HSCs. Less than 2% of nucleated bone marrow cells The CMP (also called a colony-forming unit-granulocyte-
in adult dogs are CD34+, but up to 18% CD34+ cells have erythrocyte-monocyte-megakaryocyte [CFU-GEMM])
been reported in neonatal pups.38,128 gives rise to the megakaryocyte-erythrocyte progenitor
The HSC produces a common lymphoid progenitor (CLP) (MkEP) and the granulocyte-monocyte progenitor (GMP).
and a common myeloid progenitor (CMP), as shown in The MkEP produces megakaryocyte progenitors (MkP) and
Figure 3-7. The CLP is believed to give rise to B lymphocytes, erythrocyte progenitors (EP). The GMP produces the granu-
T lymphocytes, and NK cells.16 The CMP is believed to give locyte progenitor (GP), the monocyte-dendritic cell progeni-
rise to all nonlymphoid blood cells (see Fig. 3-7) as well as tor (MDP), the basophil-mast cell progenitor (BMaP), and
macrophages, dendritic cells, osteoclasts, and mast cells.66,89 the eosinophil progenitor (EoP) in mice (see Fig. 3-7).
HPCs proliferate with higher frequency than do HSCs, but However, in humans, the EoP may develop from the CMP,
the self-renewal capabilities of HPCs decrease as progressive rather than the GMP.89
differentiation and cell lineage restrictions occur. When mea-
sured in an in vitro cell culture assay, HPCs are referred to as Mesenchymal Stem Cells
colony-forming units (CFUs). HPCs that rapidly proliferate, Mesenchymal stem cells (MSCs) are estimated to occur in
retain their ability to migrate, and form multiple subcolonies bone marrow at a frequency of 0.001% to 0.0002% of nucle-
around a larger central colony in culture are called burst- ated marrow cells.84 Evidence suggests that the MSC lineage
forming units (BFUs). differentiation pathways are less strictly delineated (exhibit
FIGURE 3 -6
Structure of the bone marrow sinus wall. Sinus lumen (s), endothelial
cell (e), basement membrane (bm), hematopoietic compartment (h),
adventitial stromal cell with processes (ac).
FIGURE 3-4
A scanning electron micrograph of the cut surface of bone marrow From Alsaker RD. The formation, emergence, and maturation of the reticulo-
showing a system of vascular sinuses originating at the periphery of the cyte: a review. Vet Clin Pathol. 1977;6(3):7-12.
marrow (right side of field) and draining into a large vein (upper left corner).
The large vein has several apertures in its wall, representing tributary
venous sinuses. Hematopoietic tissue lies between the vascular sinuses.
From Weiss L. The hematopoietic microenvironment of the bone marrow: an

ultrastructural study of the stroma of rats. Anat Rec. 1976;186:161-184. NKP
T/NKP NK cell
TLP
Pro-T lymphocyte
CLP BLP
B lymphocyte
MkP
HSC MkEP Platelets
EP
Erythrocytes
CMP
MaP
BMaP
Basophil
EoP
GMP Eosinophil
GP
Neutrophil
MDP
Monocyte
CDP
FIGURE 3 -7
FIGURE 3-5 Simplified working model of hematopoiesis. HSC, hematopoietic stem
A scanning electron micrograph from the extravascular space in rat bone cell; CLP, common lymphoid progenitor; CMP, common myeloid pro-
marrow. Spherical hematopoietic cells are shown developing in close genitor; T/NKP, T lymphocyte-natural killer cell progenitor; MkEP,
association with marrow stromal cells and their cytoplasmic processes. megakaryocyte-erythroid progenitor; GMP, granulocyte-monocyte pro-
genitor; NKP, natural killer cell progenitor; TLP, T lymphocyte progeni-
tor; BLP, B lymphocyte progenitor; MkP, megakaryocyte progenitor; EP,
Courtesy of Ahmed Deldar.
erythroid progenitor; BMaP, basophil-mast cell progenitor; EoP, eosino-
phil progenitor, GP, granulocyte progenitor; MDP, monocyte-dendritic
cell progenitor; NK, natural killer; MaP, mast cell progenitor; CDP,
common dendritic progenitor.
greater plasticity) than the HSC pathways.115 MSCs have the injection of growth factors is one approach used to collect
ability to differentiate into multiple lineages, including marrow increased numbers of stem cells from blood for human bone
stromal cells, adipocytes, osteoblasts, chondrocytes, fibroblasts, marrow transplantation.46
and myoblasts.95 Progenitor cells for a variety of peripheral
tissue cell types are also present in bone marrow. Some studies Hematopoietic Microenvironment
suggest that MSCs may also produce epithelial cells, hepato- Blood cell production occurs throughout life in the bone
cytes, and neuronal cells.81,115 Endothelial progenitor cells are marrow of adult animals because of the unique microenviron-
present in bone marrow and blood; however, their origin ment present there. The hematopoietic microenvironment is
remains to be clarified. Evidence has been presented suggest- a complex meshwork composed of stromal cells, endothelial
ing an association with both MSCs and HSCs.24,118 cells, adipocytes, macrophages, subsets of lymphocytes, NK
cells, osteoblasts, ECM components, and glycoprotein growth
Homing of Hematopoietic Stem Cells and Progenitor factors that profoundly affect HSC and HPC engraftment,
Cells to the Marrow survival, proliferation, and differentiation.1
Homing is the process by which circulating HSCs and HPCs Stromal cells and endothelial cells produce components of
bind to the luminal surface of bone marrow endothelial cells, the ECM, including collagen fibers, basement membranes of
migrate through the endothelial cells, bind selectively to sites vessels and vascular sinuses, proteoglycans, and glycoproteins.
in the extravascular space, and begin the process of prolifera- In addition to providing structural support, the ECM is
tion and differentiation. Homing of HSCs/HPCs is mediated important in the binding of hematopoietic cells and soluble
by chemoattractants produced by endothelial cells and other growth factors to stromal cells and other cells in the micro-
cells in the microenvironment and by adhesion molecules environment so that optimal proliferation and differentiation
expressed on the surfaces of HSCs/HPCs that bind to pro- can occur by virtue of these cell-cell interactions (Fig. 3-8).1,106
teoglycans and glycoproteins on the surfaces of various marrow Collagen fibers produced by stromal cells may not have
cells and the extracellular matrix.27 direct stimulatory effects on hematopoiesis but rather are
The chemokine (chemoattractant cytokine) CXCL12, also permissive, promoting hematopoiesis by forming a scaffolding
called stromal cell-derived factor-1 (SDF-1), is especially around which the other elements of the microenvironment are
important in the homing of HSCs/HPCs, but other chemoat- organized. Hematopoietic cells can adhere to collagen types I
tractants are also involved in this process. SDF-1 is produced and VI.22
by both marrow endothelial cells and stromal cells, and migra- Adhesion molecules (most importantly β1-integrins) on
tion of HSCs/HPCs from blood to bone marrow occurs the surface of hematopoietic cells bind to ECM glycoproteins
toward an SDF-1 gradient by virtue of an SDF-1 receptor such as VCAM-1, hemonectin, fibronectin, laminin, vitronec-
CXCR4 expressed on these migrating cells. SDF-1 promotes tin, and thrombospondin. The spectrum of the expression of
the expression of CXCR4 and other adhesion molecules on adhesion molecules on hematopoietic cells that differentially
the surface of HSCs/HPCs and induces transendothelial bind to ECM glycoproteins varies with the type, maturity, and
migration.27 activation state of the hematopoietic cells. In addition to
HSCs/HPCs must be activated by locally produced factors anchoring cells to a given microenvironmental niche, the
(including SDF-1) for optimal transendothelial migration to binding of adhesion molecules on hematopoietic cells also
occur. P- and E-selectin molecules (membrane-spanning,
sugar-binding glycoproteins), expressed on bone marrow
endothelial cells, bind to glycosylated ligands on HSCs/HPCs
to promote an initial loose, rolling-type adhesion between
HSCs/HPCs and endothelial cells in blood. Tight adhesion
and migration through endothelial cells is dependent on
integrin molecules—particularly the α4β1-integrin (very late
antigen-4, VLA-4) on the surfaces of migrating cells—
binding to their counterreceptors, especially vascular cell
adhesion molecule-1 (VCAM-1), on endothelial cells.27
The first successful bone marrow transplants were done
experimentally in dogs in the late 1950s.6 Because of the
homing properties of HSCs, bone marrow transplants are
performed by injecting bone marrow cells into the blood. In
addition, HSCs/HPCs naturally circulate in blood. The physi-
ologic mechanisms involved in the release of these hemato-
poietic cells from the bone marrow are not well defined, but FIGURE 3 -8
HSC and HPC numbers can be increased markedly in blood Interactions between a progenitor cell and a stromal cell in the extravas-
following injection of growth factors such as granulocyte cular microenvironment of the bone marrow. VCAM-1, vascular cell
colony-stimulating factor (G-CSF).113 In fact, intravenous adhesion molecule-1.
plays a role in cell regulation directly by activating signal cells, and T lymphocytes to produce HGFs. Different combi-
pathways for cell growth, survival, and differentiation or indi- nations of HGFs regulate the growth of different types of
rectly by modulating the responses to hematopoietic growth HSCs and/or HPCs.66
factors.22 Early-acting HGFs are involved with triggering dormant
A proteoglycan consists of a protein core with repeating (GO) primitive HSCs to begin cycling. Stem cell factor (SCF),
carbohydrate glycosaminoglycans (GAGs) attached. Major fms-like tyrosine kinase 3 ligand (Flt3L), and TPO are
proteoglycans in the marrow include heparan sulfate, chon- important early factors that act in combination with one or
droitin sulfate, hyaluronic acid, and dermatan sulfate. Proteo- more other cytokines such as IL-3, IL-6, IL-11, and G-CSF.
glycans enhance hematopoiesis by trapping soluble growth Intermediate-acting HGFs have broad specificity. IL-3
factors in the vicinity of hematopoietic cells and by strength- (multi-CSF), granulocyte-macrophage-CSF (GM-CSF), and
ening the binding of hematopoietic cells to the stroma.28 IL-4 support proliferation of multipotent HPCs. These factors
Hematopoietic cells develop in specific niches within the also interact with late-acting factors to stimulate the prolifera-
marrow. During steady-state conditions, quiescent HSCs are tion of a wide variety of committed progenitor cells. Late-
concentrated near endosteal and trabecular bone, where osteo- acting HGFs have restricted specificity. Macrophage-CSF
blasts help to regulate their numbers.152 HSCs and HPCs are (M-CSF), G-CSF, EPO, TPO, and IL-5 are more restrictive
also located near vascular sinuses, where they appear more in their actions. They have their most potent effects on com-
active. HSCs and HPCs in this vascular niche likely have mitted progenitor cells and on later stages of development
homeostatic roles during steady-state conditions.94 Erythroid when cell lines can be recognized morphologically.67 TPO
cells develop around macrophages and megakaryocytes form appears to be an exception. In addition to stimulating platelet
adjacent to sinusoidal endothelial cells; granulocyte develop- production, it is important in maintaining a population of
ment is associated with stromal cells located away from the HSCs in their osteoblastic niche.8
vascular sinuses.1,63,66
Hematopoietic Growth Factors ERY T H RO P O I E S I S

Proliferation of HSCs and HPCs cannot occur spontaneously Primitive Erythropoiesis
but requires the presence of specific hematopoietic growth Primitive erythropoiesis begins and predominates in the yolk
factors (HGFs); these may be produced locally in the bone sac but also occurs later in the liver. Primitive erythrocytes are
marrow (paracrine or autocrine) or more remotely by periph- large (more than 400 fL in humans), generally nucleated cells
eral tissues and transported to the marrow through the blood with high nuclear:cytoplasmic ratios. Their nuclei have open
(endocrine). All cells in the hematopoietic microenvironment, (noncondensed) chromatin and their cytoplasm contains pre-
including the hematopoietic cells themselves, produce HGFs dominantly embryonal hemoglobin (Hb) with a high oxygen
and/or inhibitors of hematopoiesis.69 Some HGFs have been affinity.117,133,138 In mammals as in nonmammalian species,
called poietins (erythropoietin [EPO] and thrombopoietin primitive RBCs enter the blood as nucleated cells, but in
[TPO]). Other growth factors have been classified as colony- contrast to nonmammalian species, enucleation can eventually
stimulating factors (CSFs) based on in vitro culture studies. occur in the circulation.70 These extruded nuclei circulate for
Finally, some HGFs have been described as interleukins a short time in the blood. They are surrounded by a small
(ILs).67 amount of cytoplasm and have been called pyrenocytes.97
Hematopoietic cells express receptors for more than one A switch to definitive erythropoiesis occurs during fetal
HGF on their surfaces. The number of each receptor type development. This results in the production of smaller cells
present depends on the stage of cell activation and differentia- that generally extrude their nuclei before entering the blood,
tion. Binding of an HGF to its receptor results in a series of produce fetal Hb (in some species) and adult Hb, and are
enzymatic reactions that generate transcription factors; these highly dependent on EPO for proliferation.138
promote the synthesis of molecules that inhibit apoptosis, the
formation of cell-cycle regulators (cyclins), and the synthesis Hematopoietic Progenitor Cells and the Bone
of additional HGFs and their receptors.22,67 The pathways Marrow Microenvironment
involved in generating lineage-restricted transcription factors The CMP gives rise to the MkEP, which can differentiate into
is complex and beyond the scope of this text.22 megakaryocyte progenitors (MkPs) or erythroid progenitors
HGFs vary in the type(s) of HSCs and/or HPCs that they (EPs). The production of EPs is stimulated by SCF, IL-3,
can stimulate to proliferate. Factors are often synergistic in GM-CSF, and TPO.67,78 The earliest EP is the burst-forming-
their effects on hematopoietic cells. In some instances, an unit erythrocyte (BFU-E), which differentiates into the
HGF may not directly stimulate the proliferation of a given colony-forming-unit erythrocyte (CFU-E). EPO is the
cell type, but may potentiate its proliferation by inducing the primary growth factor involved in the proliferation and dif-
expression of membrane receptors for HGFs that do directly ferentiation of CFU-Es into rubriblasts, the first morphologi-
stimulate proliferation. Some glycoproteins, such as IL-1 and cally recognizable erythroid cells. CFU-Es are more responsive
tumor necrosis factor-α (TNF-α), can modulate hematopoi- to EPO than BFU-E cells because CFU-Es exhibit greater
esis indirectly by stimulating marrow stromal cells, endothelial numbers of surface receptors for EPO.116
Marrow macrophages are important components of the

Rubriblast
hematopoietic microenvironment involved with erythropoie-
sis. Both early and late stages of erythroid development occur
with intimate membrane apposition to central macrophages Prorubricytes
in “erythroid islands.” Several adhesion molecules on ery-
throid cells and macrophages, and extracellular matrix glyco- Rubricytes
proteins are important in forming these erythroid islands.26
Direct contact with these macrophages enhances the prolif- Rubricytes
eration of erythroid precursors under basal conditions. Central
macrophages may promote basal erythrocyte production by Rubricytes
producing positive growth factors, including EPO; however,
they may inhibit erythropoiesis by producing negative factors
such as IL-1, TNF-α, transforming growth factor-β (TGF- Metarubricytes
β), and interferons (IFNs)-α, -β, and -γ in inflammatory
conditions.25,145,154 The finding that EPO can also be produced Reticulocytes
Marrow
by erythroid progenitors suggests that these cells may support release
erythropoiesis by autocrine stimulation.126 Although some Reticulocytes
degree of basal regulation of erythropoiesis occurs within the
marrow microenvironment, humoral regulation is also impor-
Erythrocytes
tant, with EPO production occurring primarily within peri-
tubular interstitial cells of the kidney and various inhibitory
FIGURE 3 -9
cytokines being produced at sites of inflammation throughout
Diagram of erythropoiesis showing the release of reticulocytes into blood
the body. as it normally occurs in dogs.
Nutrients Needed for Erythropoiesis

In addition to amino acids and essential fatty acids, several
metals and vitamins are required for normal erythropoiesis. progressively accumulates, imparting a red coloration to the
Iron is needed for the synthesis of heme, an essential compo- cytoplasm (Fig. 3-10). Cells with both red and blue coloration
nent of Hb and certain enzymes. Copper, in the form of are described as having polychromatophilic cytoplasm. An
ceruloplasmin, is important in the release of iron from tissue immature erythrocyte, termed a reticulocyte, is formed follow-
to plasma for transport to developing erythroid cells. Vitamin ing extrusion of the metarubricyte nucleus. This generally
B6 (pyridoxine) is needed as a cofactor in the first enzymatic occurs while cells are still bound to central macrophages.26
step in heme synthesis. Extruded nuclei are bound and phagocytosed by a novel
Tetrahydrofolic acid, the active form of folic acid (a B receptor on the surface of bone marrow macrophages.107
vitamin), is needed for the transfer of single carbon-containing However, nuclei can be extruded in blood when metarubri-
molecules in DNA and RNA synthesis. The physiologic cytes are released from the bone marrow (Fig. 3-11).119
mechanism of B12 involvement in erythrocyte production is Early reticulocytes have polylobulated surfaces. Their cyto-
not well understood, but it is related to folate metabolism. plasm contains ribosomes, polyribosomes, and mitochondria
Cobalt is essential for the synthesis of B12 by ruminants.51 necessary for the completion of Hb synthesis.15 Reticulocytes
derive their name from a network or reticulum that appears
Maturation of Erythroid Cells when they are stained with basic dyes such as new methylene
Rubriblasts are continuously generated from progenitor cells blue and brilliant cresyl green. That network is not preexisting
in the extravascular space of the bone marrow. The production but rather an artifact formed by the precipitation of ribosomal
of a rubriblast initiates a series of approximately four divisions ribonucleic acids and proteins secondary to staining.57 As
over a period of 3 or 4 days to produce about 16 metarubri- reticulocytes mature, the amount of ribosomal material
cytes that are no longer capable of division (Fig. 3-9).36 These decreases until only a few basophilic specks can be visualized
divisions are called maturational divisions because there is a with reticulocyte staining procedures. These mature reticulo-
progressive maturation of the nucleus and cytoplasm con- cytes have been referred to as type IV reticulocytes53 or punc-
comitant with each division. tate reticulocytes.7,101
When they are stained with Romanowsky-type blood The development of a reticulocyte into a mature erythro-
stains, early precursors have intensely blue cytoplasm owing cyte is a gradual process that requires a variable number of
to the presence of many basophilic ribosomes and polyribo- days depending on the species involved. Consequently the
somes that are actively synthesizing globin chains and smaller morphologic and physiologic properties of reticulocytes vary
amounts of other proteins. As these cells divide and mature, with the stage of maturation. The cell surface undergoes
overall cell size decreases, nuclear chromatin condensation extensive remodeling, with loss of membrane material and
increases, cytoplasmic basophilia decreases, and Hb ultimately the formation of the biconcave shape of mature
Basophilic Polychromatophilic
Rubriblast Prorubricyte Metarubricyte Reticulocyte
rubricyte rubricyte
Eosinophilic Eosinophilic Eosinophilic

Eosinophil
myelocyte metamyelocyte band
Neutrophilic Neutrophilic Neutrophilic

Myeloblast Promyelocyte Neutrophil
Basophilic Basophilic Basophilic

Basophil
FIGURE 3-10
Maturation of canine erythroid and granulocytic cells as they appear in Wright-Giemsa-stained bone marrow aspirate smears.
Drawing by Perry Bain.
erythrocytes.15 The loss of membrane protein and lipid com- Reticulocytes become progressively more deformable as they
ponents requires ATP and involves the formation of intracel- mature, a characteristic that also facilitates their release from
lular multivesicular endosomes that fuse with the plasma the marrow.144 To exit the extravascular space of the marrow,
membrane, releasing vesicles (exosomes) extracellularly.45,140 reticulocytes press against the abluminal surfaces of endothe-
This is a highly selective process in which some proteins (e.g., lial cells making up the sinus wall. Cytoplasm thins and small
transferrin receptor 1 and fibronectin receptor) are lost and pores develop in endothelial cells, which allow reticulocytes
cytoskeletal proteins (e.g., spectrin) and firmly bound trans- to be pushed through by a small pressure gradient across the
membrane proteins (e.g., the anion transporter and glycopho- sinus wall.79,143 These pores apparently close after cell passage.
rin A) are retained and concentrated.45,103 Relatively immature aggregate-type reticulocytes are
The mitochondria undergo degenerative changes in a pro- released from canine bone marrow; consequently most of
grammed death phenomenon (mitoptosis)45 and are either these cells appear polychromatophilic when they are viewed
digested or extruded following entrapment in structures following routine blood-film staining procedures.73 Reticulo-
resembling autophagic vacuoles (Fig. 3-12).90,120 The poly- cytes are generally not released from bone marrow of non
somes separate into monosomes, decrease in number, and anemic cats until they mature to punctate-type reticulocytes
disappear as reticulocytes mature into erythrocytes. The deg- (Fig. 3-13); consequently few or no aggregate reticulocytes
radation of ribosomes appears to be energy-dependent and (less than 0.4%) but up to 10% punctate reticulocytes are
presumably involves proteases and RNAases.112 found in blood from normal adult cats.29 The high percentage
Reticulocyte maturation begins in the bone marrow and is of punctate reticulocytes results from a long maturation time
completed in the peripheral blood and spleen in dogs, cats, with delayed degradation of RNA.39 Reticulocytes are gener-
and pigs.56 As reticulocytes mature, they lose the surface ally absent in the peripheral blood of healthy adult cattle and
receptors needed to adhere to the fibronectin and thrombo goats, but a small number of punctate types (0.5%) may occur
spondin components of the extracellular matrix, presumably in adult sheep.56 Based on microscopic examination of blood
facilitating their release from the bone marrow.132 films stained with new methylene blue, equine reticulocytes
A B
FIGURE 3 -12
Transmission electron microscopy of mitochondrial extrusion from
canine reticulocytes. A, A series of vesicles appear to be surrounding
three mitochondria. B, Fusion of a vacuole containing mitochondria with
A the reticulocyte outer membrane, thereby promoting mitochondrial
extrusion.
From Simpson CF, King JM. The mechanism of mitochondrial extrusion from
phenylhydrazine-induced reticulocytes in the circulating blood. J Cell Biol.
1968;36:103-109.
Wright-Giemsa Reticulocyte
Metarubricytes
Aggregate
reticulocytes
Punctate
B Marrow reticulocytes
FIGURE 3-11 Blood

Punctate
Nuclear extrusion of metarubricytes to form canine reticulocytes. A, reticulocytes
Blood film from a dog with a hemolytic anemia secondary to hemangio-
sarcoma. Frictional forces during smear preparation may have contrib-
uted to the nuclear extrusion. B, Transmission electron microscopy of
Erythrocytes
nuclear extrusion of a metarubricyte.
B, From Simpson CF, King JM. The mechanism of denucleation in circulating FIGURE 3 -13
erythroblasts. J Cell Biol. 1967;35:237-245. Cat erythroid cells demonstrating reticulocyte release into blood as it
occurs in most normal cats. Note that punctate reticulocytes do not
appear polychromatophilic when stained with Wright-Giemsa.
are normally absent from blood and are rarely released in
response to anemia.
values in CFU-E cells, decline in rubriblasts, and continue to
Control of Erythropoiesis decrease in the later stages of erythroid development.104,105
Early- and intermediate-acting growth factors—including Because of their EPO receptor density, CFU-E cells readily
SCF, IL-3, GM-CSF, and TPO—are utilized to produce EPs. respond to EPO, promoting their proliferation, differentia-
EPO is the principal growth factor promoting the viability, tion, and transformation into rubriblasts, the first morpho-
proliferation, and differentiation of EPs (BFU-E and CFU-E) logically recognizable erythroid cell type. High concentrations
that express specific cell-surface EPO receptors. The main of EPO may accelerate rubriblast entry into the first mitotic
mechanism used to achieve these effects is inhibition of apop- division, thus shortening the marrow transit time and result-
tosis.124 Early BFU-E cells do not express EPO receptors, but ing in the early release of stress reticulocytes.105
more mature BFU-E cells do and are thus responsive to EPO. In the presence of EPO, other hormones—including
EPO receptor numbers on cell surfaces increase to maximum androgens, glucocorticoid hormones, growth hormone, insulin,
and insulin-like growth factors (IGFs)—can enhance the of vascular tone, and exerting cardioprotective and neuropro-
growth of erythroid progenitor cells in vitro.76,88 The thyroid tective effects.59
hormone 3,5,3′-triiodothyronine (T3) promotes the differen-
tiation and maturation of erythroid cells.76 Thyroid hormones
LE U KO P O I E S I S
may also promote the synthesis of EPO in the kidney.82 EPO
production in adult mammals occurs primarily within peritu- Neutrophil Production
bular interstitial cells located within the inner cortex and outer Neutrophilic cells within the bone marrow can be included in
medulla of the kidney. The liver is an extrarenal source of EPO two pools (Fig. 3-15). The proliferation and maturation pool
in adults and the major site of EPO production in the mam- (mitotic pool) includes myeloblasts, promyelocytes, and
malian fetus.59 Bone marrow macrophages and erythroid promyelocytes. Approximately four or five divisions occur over
genitor cells themselves can produce EPO, suggesting the several days (Fig. 3-16). During this time primary (reddish
possibility of short-range regulation of erythropoiesis.126,141 purple) cytoplasmic granules are produced in late myeloblasts
Hematopoietic cells die not only as a consequence of lack or early promyelocytes and secondary (specific) granules are
of HGFs but also in response to the presence of molecules synthesized within myelocytes (see Fig. 3-10). Once nuclear
that induce apoptosis. Inhibitors of erythropoiesis include indentation and condensation become apparent, precursor
TGF-β, TNF-α, IFN-γ, IL-6, and TNF-related apoptosis-
inducing ligand (TRAIL).26
The ability to deliver oxygen to the tissues depends on
cardiovascular integrity, oxygen content in arterial blood, and Bone marrow Blood
Hb oxygen affinity. Low oxygen content in the blood can Pro
result from a low partial pressure of oxygen (PO2) in arterial Blast Myelo Meta Band Segmented
blood, as occurs at high altitudes or with congenital heart CNP
defects in which some of the blood flow bypasses the pulmo-
nary circulation. A low oxygen content in blood can also be
present when arterial PO2 is normal, as occurs with anemia MNP
and methemoglobinemia. An increased oxygen affinity of Hb
within erythrocytes results in a decreased tendency to release Mitosis Maturation and storage
oxygen to the tissues.86 Regardless of the cause, EPO produc- FIGURE 3 -15
tion is stimulated by tissue hypoxia (Fig. 3-14), which is Approximate sizes of mitotic and postmitotic neutrophilic compartments
mediated by hypoxia-inducible factors that control the tran- within bone marrow.
scription of the EPO gene in EPO-secreting cells.50,59
Other tissues also exhibit EPO receptors, and EPO stimu-
lates nonhematopoietic actions, including promoting the
proliferation and migration of endothelial cells, enhancing
neovascularization, stimulating the production of modulators Myeloblast
Promyelocytes
Marrow
BFU-E CFU-E NRBC Myelocytes
Myelocytes
Myelocytes
EPO RBC
Metamyelocytes
Kidney
Oxygen sensor Bands

Marrow
Blood flow O2 consumption release
Blood flow
O2 supply Oxygen content Neutrophils
Oxygen affinity
FIGURE 3-14 Neutrophils

Central role of erythropoietin (EPO) in the control of erythropoiesis.
BFU-E, burst-forming unit-erythroid; CFU-E, colony-forming unit- FIGURE 3 -16
erythroid; NRBC, nucleated red blood cell. A diagram of granulopoiesis.
cells are no longer capable of division. The maturation and neutrophilic colony formation. If this also happens in vivo, it
storage pool (postmitotic pool) includes metamyelocytes, might provide negative feedback for neutrophil production in
bands, and segmented neutrophils. Cells within this pool nor- the extravascular space of bone marrow.71 One possible mech-
mally undergo maturation and storage for several more days anism is the release of serine proteases, such as elastase, from
prior to the migration of mature neutrophils through the neutrophils. Elastase appears to inhibit granulopoiesis by
vascular endothelium and into the circulation.123 The number inactivating G-CSF.35 Second, increased neutrophil numbers
of mature neutrophils stored in marrow is more than seven in blood are associated with the increased clearance of circu-
times the number present in the circulation of the dog.32 The lating G-CSF following binding to surface receptors on neu-
marrow transit time from myeloblast to release of mature trophils, thereby decreasing the primary stimulus for their
neutrophils into the blood varies by species but is generally production.65,75 Third, mature neutrophils indirectly inhibit
between 6 and 9 days. This time can be shortened considerably granulopoiesis by the removal (through phagocytosis) of
when inflammation is present.108,123 invading microorganisms that would otherwise stimulate the
A variety of cytokines with overlapping specificities are production of HGFs by tissue cells. Activated T lymphocytes
important in neutrophil production (also called granulopoie- may also inhibit neutrophil production by secreting the soluble
sis). IL-3, GM-CSF, and G-CSF are of primary importance molecule Fas-ligand and the cytokine IFN-γ.99
in the production of neutrophils. These cytokines act on
various stages of development from CMPs to GMPs to GPs, Eosinophil, Basophil, and Mast Cell Production
depending on the array of growth factor receptors displayed The eosinophil progenitor (EoP) or colony-forming unit
on their surfaces. GPs are stimulated to proliferate and dif- eosinophil (CFU-Eo) is reported to develop from the CMP
ferentiate into myeloblasts by G-CSF. This cytokine appears in humans but downstream from the GMP in mice.89 Eosino-
to play a role in the basal regulation of granulopoiesis as well phil production in the marrow parallels that of neutrophils.
as to function as a primary regulator of the neutrophil response Eosinophil precursors become recognizable at the myelocyte
to inflammatory stimuli. G-CSF increases the number of cell stage, when their characteristic secondary granules appear (see
divisions and reduces the time for granulocytic progenitors to Fig. 3-10). The marrow transit time is 1 week or less, with a
develop into neutrophils. It also promotes the release of neu- significant storage pool of mature eosinophils.142 As in the
trophils from bone marrow into blood.108,127 case of neutrophils, growth factors, including IL-3 and
As neutrophils mature, there is a progressive downregula- GM-CSF, are needed for the proliferation of early progeni-
tion of certain surface receptors, including CXCR4 and the tors. In addition, activated TH2 lymphocytes produce IL-5,
α4β1 integrin, that adhere neutrophils to glycoproteins within which promotes the terminal maturation of eosinophils. IL-3,
the extravascular space. CXCR4 binds to CXCL12/SDF-1 GM-CSF, and IL-5 also inhibit eosinophil apoptosis,62 while
produced by stromal cells, and the α4β1 integrin binds to inhibitors of eosinophil production include IL-12 and
VCAM-1 on endothelial cells. Experimental neutralization of IFN-γ.110
CXCR4 and VCAM-1 results in an increased release of neu- The GMP reportedly produces the bipotential basophil-
trophils into blood. G-CSF promotes neutrophil release from mast cell progenitor (BMaP), which gives rise to the basophil
bone marrow at least in part by decreasing CXCL12/SDF-1 progenitor and the mast cell progenitor (MaP).121 Like eosin-
production and decreasing CXCR4 expression on the surface ophils, basophil precursors become recognizable at the myelo-
of neutrophils.127 cyte stage, when their characteristic secondary granules appear
Activated helper T lymphocytes produce various growth (see Fig. 3-10). A specific growth factor for the production of
factors including IL-3 and GM-CSF. Mononuclear phago- basophils has not been identified. IL-3 appears to be the
cytes, fibroblasts, and endothelial cells can also produce major growth and differentiation factor for basophils, but
GM-CSF and G-CSF when appropriately stimulated. Mono- other growth factors—including GM-CSF, IL-5, TGF-β,
nuclear phagocytes can not only synthesize HGFs when they and nerve growth factor—also promote the production of
contact bacterial products but can also stimulate other cells to basophils.37
produce them. The cytokines, IL-1 and TNF-α, produced by In contrast to basophils, which mature in the bone marrow,
monocytes and macrophages stimulate the production of maturation of mast cell progenitors into mast cells occurs in
HGFs by other cell types. These monokines are important in the tissues.40 SCF appears to be the major growth and dif-
the inflammatory response to foreign organisms and neoplas- ferentiation factor for mast cells. Additional cytokines—
tic cells, but their role in resting granulopoiesis is unclear.66,123 including IL-3, IL-4, IL-9, IL-10, and IL-13—also stimulate
IL-6 is a multifunctional cytokine that regulates inflammation mast cell production.54 Some local proliferation of mast cells
(including the hepatic acute-phase response), the immune can occur in tissues if they are appropriately stimulated.41
response, and hematopoiesis. In this latter role, it promotes
granulopoiesis and thrombopoiesis during inflammation.111 Production of Monocytes, Macrophages, Dendritic
Inhibition of neutrophil production is not well understood, Cells, and Osteoclasts
but mature neutrophils may provide negative feedback inhibi- Bone marrow MDPs give rise to monocytes and common
tion on their own production in three ways. First, the addition dendritic cell progenitors (CDPs).153 Monocytes are produced
of mature neutrophils to bone marrow culture inhibits through the combined effects of IL-3, GM-CSF, M-CSF, and
IL-34 on the proliferation and differentiation of bone marrow required for pre-B lymphocytes to develop into mature, naive
progenitor cells.44 Less time is required to produce monocytes B lymphocytes in the marrow and enter the circulation. Less
than granulocytes, and there is little marrow reserve of these than 20% of B lymphocytes produced in the marrow become
cells. part of the peripheral mature B lymphocyte pool, with most
Monocytes have long been viewed primarily as precursors of the cells being culled in the bone marrow or after
that develop into tissue macrophages and dendritic cells. their entry into blood.83 B lymphocytes also proliferate in
However, it is now recognized that many macrophage and peripheral lymphoid tissues in adults. As with other blood
dendritic cells in tissues do not originate from monocytes cells, the microenvironment of the marrow and lymphoid
under steady-state conditions because these cells are capable organs is important for lymphopoiesis. The production of
of self-replication. In addition, neither microglia (macro- antigen-sensitive, surface-immunoglobulin-positive B lym-
phages in the central nervous system) nor Langerhans cells phocytes is marked by successive rearrangements of the
(epidermal dendritic cells) depend on cells from the bone immunoglobulin gene loci and selective expression of surface
marrow for their renewal under steady-state conditions and proteins. Although a number of cytokines—including SCF,
possibly also during inflammation.10 In fact, Langerhans cells Flt3L, SDF-1, and IGF—are involved in B lymphocyte pro-
appear to develop from embryonic progenitor cells that enter duction in marrow, IL-7 appears to be an especially important
the epidermis before birth.44 positive growth factor.19,77 B lymphocyte lymphopoiesis is
Monocytes are important effector cells during inflamma- inhibited by several factors, including TGF-β, IFN-α, IFN-β,
tory conditions. They exit the blood, respond to the tissue and IFN-γ.77
environment, and differentiate into subsets of macrophages Recirculating B lymphocytes are activated by antigenic
and inflammatory dendritic cells. Exposure to M-CSF pro- stimulation in the T lymphocyte region of secondary lym-
motes the development of monocytes into macrophages. The phoid organs, followed by migration to the cortex in lymph
addition of IFN-γ to M-CSF promotes the formation of nodes and to follicles in jejunal Peyer’s patches and the spleen
M1-like macrophages, while the addition of IL-4 to M-CSF in mammals.135 B lymphocyte activation and differentiation
induces the differentiation of M2-like macrophages. The into plasmablasts is induced by combinations of microbial
exposure of monocytes to GM-CSF, IL-4, and TNF-α pro- products, cytokines, and molecules bound to the surfaces of T
motes their development into inflammatory dendritic cells lymphocytes and dendritic cells. Plasmablasts can develop
or TNF-α and inducible nitric oxide synthase (iNOS)- into plasma cells in the lymphoid organs where they are pro-
producing (TiP)-dendritic cells.10,44,153 duced or can migrate through blood and develop into plasma
CDPs can give rise to preclassic dendritic cells (Pre-cDC) cells in peripheral tissues or bone marrow. SDF-1 attracts
and plasmacytoid dendritic cells (pDCs) in bone marrow.96 circulating plasmablasts to the bone marrow, and factors
Both cell types are released into blood and enter the tissues, including SDF-1 and IL-6 promote plasma cell development
where the Pre-cDCs develop into classic dendritic cells (cDC) by preventing apoptosis.91
in lymphoid organs and mucosal dendritic cells.44,153 T lymphocyte progenitors leave the marrow and migrate
Cytokines—including Flt3L, GM-CSF, and lymphotoxin to the thymus. Homing of these cells to the thymus depends
α1β2—appear to be important for the development of cDCs on their interaction with various adhesion molecules on
and pDCs.10,44 thymic endothelial cells and the production of specific che-
Osteoclasts develop when monocyte progenitors are cul- motactic factors by thymic stromal cells. T lymphocyte pro-
tured with M-CSF and a soluble form of receptor activator genitors develop into T lymphocytes under the influence of
of nuclear factor-κB ligand (RANKL).9 IL-3 and GM-CSF the thymic microenvironment and growth factors (including
inhibit osteoclast formation. The relative amounts of these Flt3L and IL-7) produced in the thymus.151 After maturation
growth factors and presumably others present in the micro- in the thymus, T lymphocytes accumulate within paracortical
environment of a monocyte progenitor apparently determine areas of lymph nodes, periarteriolar lymphoid sheaths of the
whether macrophages, dendritic cells, or osteoclasts are spleen, and the interfollicular areas of jejunal Peyer’s patches
formed. in mammals.135
Most NK cells are produced from progenitor cells in the
Lymphocyte and NK Cell Production bone marrow, where they undergo expansion and maturation
The CLP is believed to give rise to B lymphocytes, T lym- for a week or more before their release into the blood.155
phocytes, and NK cells.16 The development of B lymphocyte Growth factors controlling their production need further
and T lymphocyte progenitors in bone marrow is antigen- characterization, but SCF, IL-2, IL-7, and IL-15 can stimu-
independent. Both SCF and Flt3L appear to be involved late NK cell development from progenitor cells in vitro.3
in the production of early lymphoid progenitor cells in Subsets of NK cells also develop in the thymus and possibly
mice.17 other organs, such as lymph nodes, liver, and spleen. These
B lymphocyte progenitors produce mature, naive B lym- sites may depend on the trafficking of bone marrow–derived
phocytes in the marrow in most mammals, in specialized ileal progenitor cells and/or immature NK cells into these organs
Peyer’s patches in dogs, pigs, and ruminants, and in the bursa from the blood, where they mature under the influence of
of Fabricius in birds.83,135 Approximately 2 to 3 days are microenvironmental factors.49,55
individual platelets within the sinuses and general circula-

T H RO M B O P O I E S I S tion.60 Megakaryocytes may rarely migrate through the vas-
cular endothelium into the sinuses, enter the general venous
Blood platelets in mammals are produced from multinucle- circulation (see Fig. 2-34), and become lodged in pulmonary
ated giant cells in the bone marrow called megakaryocytes. capillaries.11 It is estimated that 1000 to 3000 platelets are
The CMP gives rise to the MkEP, which can differentiate into produced from each megakaryocyte, depending on mega-
megakaryocyte progenitors (MkPs) or erythroid progenitors karyocyte size.68 Megakaryocytes are not present in nonmam-
(EPs). The earliest MkP is the burst-forming-unit mega- malian species. Like erythrocytes and leukocytes, the nucleated
karyocyte (BFU-Mk). When appropriately stimulated, this thrombocytes of nonmammalian species are produced by
progenitor cell divides and differentiates into colony-forming- mitosis of precursor cells.
unit megakaryocyte (CFU-Mk) progenitor cells, which divide A number of cytokines can stimulate or enhance the pro-
and differentiate into megakaryoblasts (Fig. 3-17).68 Mitosis liferation and expansion of megakaryocyte progenitor cells.
stops at this stage and endomitosis (nuclear reduplication Factors that may be involved include SCF, Flt3L, IL-3,
without cell division) begins. Generally 2 to 5 nuclear redu- GM-CSF, IL-11, and EPO.13,18,68 TPO is the key stimulator
plications occur resulting in 8 to 64 sets of chromosomes of platelet production by stimulating megakaryocyte prolifera-
(8 N-64 N) in mature megakaryocytes, compared to two sets tion, survival, and size (ploidy).60,68 TPO also transiently
of chromosomes (2 N) in most cells in the body. Individual enhances the aggregatory response of platelets to agonists.4
nuclei can be observed following the first two reduplications Although various cells in the body can produce TPO,
(promegakaryocytes), but a large polylobulated nucleus is seen including cells in the kidney and bone marrow stromal
when mature megakaryocytes are formed. The mean ploidy of cells,80,87 the major sites of TPO production appear to be the
human and mouse megakaryocytes (16 N) is lower than mean endothelial cells of the liver.58,148 The amount of TPO pro-
values (32 N to 64 N) reported for megakaryocytes in dogs, duced in the body appears to be relatively constant. TPO
cats, and cattle.18 The cytoplasm in promegakaryocytes is receptors (c-Mpl receptors) on blood platelets and maturing
intensely basophilic. There is a progressive decrease in baso- megakaryocytes can bind, internalize, and degrade TPO,
philia and increase in granularity as megakaryocytes mature. providing negative feedback on platelet production.68 Conse-
Cell volume increases with each reduplication; consequently, quently blood TPO concentration is remarkably high in the
megakaryocytes are much larger than all other marrow cells case of thrombocytopenia resulting from megakaryocytic
except osteoclasts. In contrast to mature megakaryocytes, hypoplasia. In contrast, blood TPO concentrations are much
osteoclasts have multiple discrete nuclei. lower with ongoing immune-mediated thrombocytopenia,
Mature megakaryocytes develop just outside vascular because megakaryocytes are generally increased in the marrow
sinuses. SDF-1 and fibroblast growth factor-4 promote the and rapid platelet turnover is occurring, resulting in increased
localization and binding of megakaryocyte progenitors in this binding and removal of TPO from blood.52
vascular niche (via adhesion molecules VCAM-1 and the α4β1 However, the number of maturing megakaryocytes and
integrin), which promotes survival, maturation, and platelet blood platelets present may not be the only determinants of
production.11 Protrusions of cytoplasm (proplatelets) from blood TPO concentrations. IL-6 stimulates thrombopoiesis
megakaryocytes form and extend into sinuses where they can by increasing the production of TPO by the liver, which con-
be sheared off by the force of flowing blood (see Fig. 3-3). tributes to the thrombocytosis seen in some inflammatory
These beaded-appearing proplatelets eventually fragment into conditions.64 Conversely, platelet factor 4 (PF4), TGF-β,
IL-4, and TNF-α appear to be inhibitors of megakaryocyte
production.13,137
Mitosis Endomitosis
R EF ER EN C E S
1. Abboud CN, Lichtman MA. Structure of the marrow and the hematopoietic microen-
Promegakaryocyte vironment. In: Lichtman MA, Beutler E, Kipps TJ, et al, eds. Williams Hematology.
7th ed. New York: McGraw-Hill; 2006:35-72.
MkEP 2. Abkowitz JL, Catlin SN, McCallie MT, et al. Evidence that the number of hematopoi-
etic stem cells per animal is conserved in mammals. Blood. 2002;100:2665-2667.
Megakaryoblast 3. Aiba Y, Hirayama F, Ogawa M. Clonal proliferation and cytokine requirement of
Megakaryocyte murine progenitors for natural killer cells. Blood. 1997;89:4005-4012.
BFU-Mk
4. Akkerman JW. Thrombopoietin and platelet function. Semin Thromb Hemost.
CFU-Mk 2006;32:295-304.
5. Allender MC, Fry MM. Amphibian hematology. Vet Clin North Am Exot Anim Pract.
FIGURE 3-17 2008;11:463-480.
6. Alpen EL, Baum SJ. Modification of x-radiation lethality by autologous marrow infu-
Stages of megakaryocyte development. MkEP, Megakaryocyte-erthroid sion in dogs. Blood. 1958;13:1168-1175.
progenitor; BFU-Mega, burst-forming unit megakaryocyte; CFU- 7. Alsaker RD. The formation, emergence, and maturation of the reticulocyte: a review.
Mega, colony-forming unit megakaryocyte. Vet Clin Pathol. 1977;6(3):7-12.
8. Arai F, Yoshihara H, Hosokawa K, et al. Niche regulation of hematopoietic stem cells 42. Galloway JL, Zon LI. Ontogeny of hematopoiesis: examining the emergence of
in the endosteum. Ann N Y Acad Sci. 2009;1176:36-46. hematopoietic cells in the vertebrate embryo. Curr Top Dev Biol. 2003;53:
9. Asagiri M, Takayanagi H. The molecular understanding of osteoclast differentiation. 139-158.
Bone. 2007;40:251-264. 43. Gangenahalli GU, Singh VK, Verma YK, et al. Hematopoietic stem cell antigen CD34:
10. Auffray C, Sieweke MH, Geissmann F. Blood monocytes: development, heterogeneity, role in adhesion or homing. Stem Cells Dev. 2006;15:305-313.
and relationship with dendritic cells. Annu Rev Immunol. 2009;27:669-692. 44. Geissmann F, Manz MG, Jung S, et al. Development of monocytes, macrophages, and
11. Avecilla ST, Hattori K, Heissig B, et al. Chemokine-mediated interaction of hemato- dendritic cells. Science. 2010;327:656-661.
poietic progenitors with the bone marrow vascular niche is required for thrombopoi- 45. Geminard C, de Gassart A, Vidal M. Reticulocyte maturation: mitoptosis and exosome
esis. Nat Med. 2004;10:64-71. release. Biocell. 2002;26:205-215.
12. Barao I, Ascensao JL. Human natural killer cells. Arch Immunol Ther Exp (Warsz). 46. Gertz MA. Current status of stem cell mobilization. Br J Haematol. 2010;150:
1998;46:213-229. 647-662.
13. Battinelli EM, Hartwig JH, Italiano JE, Jr. Delivering new insight into the biology of 47. Gimble JM, Robinson CE, Wu X, et al. The function of adipocytes in the bone marrow
megakaryopoiesis and thrombopoiesis. Curr Opin Hematol. 2007;14:419-426. stroma: an update. Bone. 1996;19:421-428.
14. Bertrand JY, Jalil A, Klaine M, et al. Three pathways to mature macrophages in the 48. Glomski CA, Tamburlin J, Chainani M. The phylogenetic odyssey of the erythrocyte.
early mouse yolk sac. Blood. 2005;106:3004-3011. III. Fish, the lower vertebrate experience. Histol Histopathol. 1992;7:501-528.
15. Bessis M. Living Blood Cells and Their Ultrastructure. New York: Springer-Verlag; 49. Gregoire C, Chasson L, Luci C, et al. The trafficking of natural killer cells. Immunol
1973. Rev. 2007;220:169-182.
16. Blom B, Spits H. Development of human lymphoid cells. Annu Rev Immunol. 50. Gruber M, Hu CJ, Johnson RS, et al. Acute postnatal ablation of Hif-2α results in
2006;24:287-320. anemia. Proc Natl Acad Sci U S A. 2007;104:2301-2306.
17. Borge OJ, Adolfsson J, Jacobsen AM. Lymphoid-restricted development from multi- 51. Harvey JW. The erythrocyte: physiology, metabolism and biochemical disorders. In:
potent candidate murine stem cells: distinct and complimentary functions of the c-kit Kaneko JJ, Harvey JW, Bruss ML, eds. Clinical Biochemistry of Domestic Animals.
and flt3-ligands. Blood. 1999;94:3781-3790. 6th ed. San Diego: Academic Press; 2008:173-240.
18. Boudreaux MK. Thrombopoiesis. In: Weiss DJ, Wardrop KJ, eds. Schalm’s Veterinary 52. Hou M, Andersson PO, Stockelberg D, et al. Plasma thrombopoietin levels in throm-
Hematology. 6th ed. Ames, IA: Wiley-Blackwell; 2010:56-60. bocytopenic states: implication for a regulatory role of bone marrow megakaryocytes.
19. Burkhard MJ. Lymphopoiesis. In: Weiss DJ, Wardrop KJ, eds. Schalm’s Veterinary Br J Haematol. 1998;101:420-424.
Hematology. 6th ed. Ames, IA: Wiley-Blackwell; 2010:61-64. 53. Houwen B. Reticulocyte maturation. Blood Cells. 1992;18:167-186.
20. Campbell F. Fine structure of the bone marrow of the chicken and pigeon. J Morphol. 54. Hu ZQ, Zhao WH, Shimamura T. Regulation of mast cell development by inflammatory
1967;123:405-439. factors. Curr Med Chem. 2007;14:3044-3050.
21. Carbonell F, Calvo W, Fliedner TM. Cellular composition of human fetal bone marrow. 55. Huntington ND, Vosshenrich CA, Di Santo JP. Developmental pathways that
Histologic study in methacrylate sections. Acta Anat (Basel). 1982;113:371-375. generate natural-killer-cell diversity in mice and humans. Nat Rev Immunol.
22. Carr BD. The hematopoietic system. In: Weiss DJ, Wardrop KJ, eds. Schalm’s Veteri- 2007;7:703-714.
nary Hematology. 6th ed. Ames, IA: Wiley-Blackwell; 2010:27-35. 56. Jain NC. Schalm’s Veterinary Hematology. 4th ed. Philadelphia: Lea & Febiger; 1986.
23. Cecchini MG, Hofstetter W, Halasy J, et al. Role of CSF-1 in bone and bone marrow 57. Jain NC. Essentials of Veterinary Hematology. Philadelphia: Lea & Febiger; 1993.
development. Mol Reprod Dev. 1997;46:75-83. 58. Jelkmann W. The role of the liver in the production of thrombopoietin compared with
24. Chao H, Hirschi KK. Hemato-vascular origins of endothelial progenitor cells? Micro- erythropoietin. Eur J Gastroenterol Hepatol. 2001;13:791-801.
vasc Res. 2010;79:169-173. 59. Jelkmann W. Erythropoietin after a century of research: younger than ever. Eur J
25. Chasis JA. Erythroblastic islands: specialized microenvironmental niches for erythro- Haematol. 2007;78:183-205.
poiesis. Curr Opin Hematol. 2006;13:137-141. 60. Junt T, Schulze H, Chen Z, et al. Dynamic visualization of thrombopoiesis within bone
26. Chasis JA, Mohandas N. Erythroblastic islands: niches for erythropoiesis. Blood. marrow. Science. 2007;317:1767-1770.
2008;112:470-478. 61. Kalinkovich A, Spiegel A, Shivtiel S, et al. Blood-forming stem cells are nervous: direct
27. Chute JP. Stem cell homing. Curr Opin Hematol. 2006;13:399-406. and indirect regulation of immature human CD34+ cells by the nervous system. Brain
28. Coombe DR. Biological implications of glycosaminoglycan interactions with haemo- Behav Immun. 2009;23:1059-1065.
poietic cytokines. Immunol Cell Biol. 2008;86:598-607. 62. Kankaanranta H, Moilanen E, Zhang X. Pharmacological regulation of human eosino-
29. Cramer DV, Lewis RM. Reticulocyte response in the cat. J Am Vet Med Assoc. phil apoptosis. Curr Drug Targets Inflamm Allergy. 2005;4:433-445.
1972;160:61-67. 63. Kaplan RN, Psaila B, Lyden D. Niche-to-niche migration of bone-marrow-derived cells.
30. Dabrowski Z, Sano MI, Tabarowski Z, et al. Haematopoiesis in snakes (Ophidia) in Trends Mol Med. 2007;13:72-81.
early postnatal development. Cell Tissue Res. 2007;328:291-299. 64. Kaser A, Brandacher G, Steurer W, et al. Interleukin-6 stimulates thrombopoiesis
31. Dent AL, Kaplan MH. T cell regulation of hematopoiesis. Front Biosci. 2008;13: through thrombopoietin: role in inflammatory thrombocytosis. Blood. 2001;98:
6229-6236. 2720-2725.
32. Deubelbeiss KA, Dancey JT, Harker LA, et al. Neutrophil kinetics in the dog. J Clin 65. Kastner M, Maurer HR. Pure bovine granulocytes as a source of granulopoiesis inhibi-
Invest. 1975;55:833-839. tor (chalone). Hoppe Seylers Z Physiol Chem. 1980;361:197-200.
33. Di Rosa F. T-lymphocyte interaction with stromal, bone and hematopoietic cells in the 66. Kaushansky K. Hematopoietic stem cells, progenitors, and cytokines. In: Lichtman
bone marrow. Immunol Cell Biol. 2009;87:20-29. MA, Beutler E, Kipps TJ, et al, eds. Williams Hematology. 7th ed. New York: McGraw-
34. Durand C, Dzierzak E. Embryonic beginnings of adult hematopoietic stem cells. Hae- Hill; 2006:201-220.
matologica. 2005;90:100-108. 67. Kaushansky K. Lineage-specific hematopoietic growth factors. N Engl J Med.
35. El Ouriaghli F, Fujiwara H, Melenhorst JJ, et al. Neutrophil elastase enzymatically 2006;354:2034-2045.
antagonizes the in vitro action of G-CSF: implications for the regulation of granulo- 68. Kaushansky K. Historical review: megakaryopoiesis and thrombopoiesis. Blood.
poiesis. Blood. 2003;101:1752-1758. 2008;111:981-986.
36. Erslev AJ, Beutler E. Production and destruction of erythrocytes. In: Beutler E, Licht- 69. Kim DH, Yoo KH, Choi KS, et al. Gene expression profile of cytokine and growth factor
man MA, Coller BS, et al, eds. Williams Hematology. 5th ed. New York: McGraw-Hill; during differentiation of bone marrow-derived mesenchymal stem cell. Cytokine.
1995:425-441. 2005;31:119-126.
37. Falcone FH, Haas H, Gibbs BF. The human basophil: a new appreciation of its role in 70. Kingsley PD, Malik J, Fantauzzo KA, et al. Yolk sac-derived primitive erythroblasts
immune responses. Blood. 2000;96:4028-4038. enucleate during mammalian embryogenesis. Blood. 2004;104:19-25.
38. Faldyna M, Sinkora J, Knotigova P, et al. Flow cytometric analysis of bone marrow 71. Kovacs P, Brunch C, Fliedner TM. Colony formation by canine hemopoietic cells in
leukocytes in neonatal dogs. Vet Immunol Immunopathol. 2003;95:165-176. vitro. Inhibition by polymorphonuclear leukocytes. Acta Haematol. 1976;56:107-115.
39. Fan LC, Dorner JL, Hoffman WE. Reticulocyte response and maturation in 72. Kucia M, Reca R, Jala VR, et al. Bone marrow as a home of heterogenous populations
experimental acute blood loss anemia in the cat. J Am Anim Hosp Assoc. of nonhematopoietic stem cells. Leukemia. 2005;19:1118-1127.
1978;14:219-224. 73. Laber J, Perman V, Stevens JB. Polychromasia or reticulocytes—an assessment of
40. Födinger M, Fritsch G, Winkler K, et al. Origin of human mast cells: development from the dog. J Am Anim Hosp Assoc. 1974;10:399-406.
transplanted hematopoietic stem cells after allogeneic bone marrow transplantation. 74. Lancrin C, Sroczynska P, Serrano AG, et al. Blood cell generation from the hemangio-
Blood. 1994;84:2954-2959. blast. J Mol Med. 2010;88:167-172.
41. Galli SJ, Metcalfe DD, Arber DA, et al. Basophils and mast cells and their disorders. 75. Layton JE, Hockman H, Sheridan WP, et al. Evidence for a novel control mechanism
In: Lichtman MA, Beutler E, Kipps TJ, et al, eds. Williams Hematology. 7th ed. New of granulopoiesis: mature cell related control of a regulatory growth factor. Blood.
York: McGraw-Hill; 2006:879-897. 1989;74:1303-1307.
76. Leberbauer C, Boulme F, Unfried G, et al. Different steroids co-regulate long-term 107. Qui LB, Dickson H, Hajibagheri N, et al. Extruded erythroblast nuclei are bound and
expansion versus terminal differentiation in primary human erythroid progenitors. phagocytosed by a novel macrophage receptor. Blood. 1995;85:1630-1639.
Blood. 2005;105:85-94. 108. Radin MJ, Wellman ML. Granulopoiesis. In: Weiss DJ, Wardrop KJ, eds. Schalm’s
77. Lebien TW. Lymphopoiesis. In: Lichtman MA, Beutler E, Kipps TJ, et al, eds. Williams Veterinary Hematology. 6th ed. Ames, IA: Wiley-Blackwell; 2010:43-49.
Hematology. 7th ed. New York: McGraw-Hill; 2006:1039-1049. 109. Rafii S, Mohle R, Shapiro F, et al. Regulation of hematopoiesis by microvascular
78. Liang DC, Shih LY, Chai IJ, et al. The synergistic effect of thrombopoietin in erythro- endothelium. Leuk Lymphoma. 1997;27:375-386.
poiesis with erythropoietin and/or IL-3 and myelopoiesis with G-CSF or IL-3 from 110. Rais M, Wild JS, Choudhury BK, et al. Interleukin-12 inhibits eosinophil differentiation
umbilical cord blood cells of premature neonates. Pediatr Hematol Oncol. from bone marrow stem cells in an interferon-gamma-dependent manner in a mouse
2002;19:399-405. model of asthma. Clin Exp Allergy. 2002;32:627-632.
79. Lichtman MA, Santillo P. Red cell egress from the marrow. Blood Cells. 111. Raj DS. Role of interleukin-6 in the anemia of chronic disease. Semin Arthritis Rheum.
1986;12:11-23. 2009;38:382-388.
80. Linthorst GE, Folman CC, van Olden RW, et al. Plasma thrombopoietin levels in 112. Rapoport SM. The Reticulocyte. Boca Raton, FL: CRC Press, Inc.; 1986.
patients with chronic renal failure. Hematol J. 2002;3:38-42. 113. Robinson SN, Seina SM, Gohr JC, et al. Hematopoietic progenitor cell mobilization
81. Liu ZJ, Zhuge Y, Velazquez OC. Trafficking and differentiation of mesenchymal stem by granulocyte colony-stimulating factor and erythropoietin in the absence of matrix
cells. J Cell Biochem. 2009;106:984-991. metalloproteinase-9. Stem Cells Dev. 2005;14:317-328.
82. Ma Y, Freitag P, Zhou J, et al. Thyroid hormone induces erythropoietin gene expres- 114. Sage PT, Carman CV. Settings and mechanisms for trans-cellular diapedesis. Front
sion through augmented accumulation of hypoxia-inducible factor-1. Am J Physiol Biosci. 2009;14:5066-5083.
Regul Integr Comp Physiol. 2004;287:R600-R607. 115. Satija NK, Singh VK, Verma YK, et al. Mesenchymal stem cell-based therapy: a new
83. Macallan DC, Wallace DL, Zhang Y, et al. B-cell kinetics in humans: rapid turnover paradigm in regenerative medicine. J Cell Mol Med. 2009;13:4385-4402.
of peripheral blood memory cells. Blood. 2005;105:3633-3640. 116. Sawada K, Krantz SB, Dai CH, et al. Purification of human blood burst-forming units-
84. Martin DR, Cox NR, Hathcock TL, et al. Isolation and characterization of multipotential erythroid and demonstration of the evolution of erythropoietin receptors. J Cell
mesenchymal stem cells from feline bone marrow. Exp Hematol. 2002;30: Physiol. 1990;142:219-230.
879-886. 117. Segel GB, Palis J. Hematology of the newborn. In: Lichtman MA, Beutler E, Kipps TJ,
85. Mayani H, Guilbert LJ, Janowska-Wieczorek A. Biology of the hemopoietic micro et al, eds. Williams Hematology. 7th ed. New York: McGraw-Hill; 2006:81-99.
environment. Eur J Haematol. 1992;49:225-233. 118. Silva GV, Litovsky S, Assad JA, et al. Mesenchymal stem cells differentiate into an
86. McCully K, Chance B, Giger U. In vivo determination of altered hemoglobin saturation endothelial phenotype, enhance vascular density, and improve heart function in a
in dogs with M-type phosphofructokinase deficiency. Muscle Nerve. 1999;22: canine chronic ischemia model. Circulation. 2005;111:150-156.
621-627. 119. Simpson CF, Kling JM. The mechanism of denucleation in circulating erythroblasts.
87. McIntosh B, Kaushansky K. Transcriptional regulation of bone marrow thrombopoietin J Cell Biol. 1967;35:237-245.
by platelet proteins. Exp Hematol. 2008;36:799-806. 120. Simpson CF, Kling JM. The mechanism of mitochondrial extrusion from
88. Miyagawa S, Kobayashi M, Konishi N, et al. Insulin and insulin-like growth factor I phenylhydrazine-induced reticulocytes in the circulating blood. J Cell Biol.
support the proliferation of erythroid progenitor cells in bone marrow through the 1968;36:103-109.
sharing of receptors. Br J Haematol. 2000;109:555-562. 121. Siracusa MC, Perrigoue JG, Comeau MR, et al. New paradigms in basophil develop-
89. Mori Y, Iwasaki H, Kohno K, et al. Identification of the human eosinophil lineage- ment, regulation and function. Immunol Cell Biol. 2010;88:275-284.
committed progenitor: revision of phenotypic definition of the human common myeloid 122. Sitnicka E, Wang QR, Tsai S, et al. Support versus inhibition of hematopoiesis by two
progenitor. J Exp Med. 2009;206:183-193. characterized stromal cell types. Stem Cells. 1995;13:655-665.
90. Mortensen M, Ferguson DJ, Simon AK. Mitochondrial clearance by autophagy in 123. Smith CW. Production, distribution, and fate of neutrophils. In: Lichtman MA, Beutler
developing erythrocytes: Clearly important, but just how much so? Cell Cycle. E, Kipps TJ, et al, eds. Williams Hematology. 7th ed. New York: McGraw-Hill;
2010;15:1901-1906. 2006:855-861.
91. Moser K, Tokoyoda K, Radbruch A, et al. Stromal niches, plasma cell differentiation 124. Socolovsky M. Molecular insights into stress erythropoiesis. Curr Opin Hematol.
and survival. Curr Opin Immunol. 2006;18:265-270. 2007;14:215-224.
92. Nilsson SK, Simmons PJ. Transplantable stem cells: home to specific niches. Curr 125. Sorrell JM, Weiss L. Cell interactions between hematopoietic and stromal cells in
Opin Hematol. 2004;11:102-106. the embryonic chick bone marrow. Anat Rec. 1980;197:1-19.
93. Nobuhisa I, Takizawa M, Takaki S, et al. Regulation of hematopoietic development 126. Stopka T, Zivny JH, Stopkova P, et al. Human hematopoietic progenitors express
in the aorta-gonad-mesonephros region mediated by Lnk adaptor protein. Mol Cell erythropoietin. Blood. 1998;91:3766-3772.
Biol. 2003;23:8486-8494. 127. Summers C, Rankin SM, Condliffe AM, et al. Neutrophil kinetics in health and disease.
94. Oh IH, Kwon KR. Concise review: multiple niches for hematopoietic stem cell regula- Trends Immunol. 2010;31:318-324.
tions. Stem Cells. 2010;28:1243-1249. 128. Suter SE, Gouthro TA, McSweeney PA, et al. Isolation and characterization of
95. Ohishi M, Schipani E. Bone marrow mesenchymal stem cells. J Cell Biochem. pediatric canine bone marrow CD34+ cells. Vet Immunol Immunopathol.
2010;109:277-282. 2004;101:31-47.
96. Onai N, Manz MG, Schmid MA. Isolation of common dendritic cell progenitors (CDP) 129. Tavassoli M. Embryonic and fetal hemopoiesis: an overview. Blood Cells.
from mouse bone marrow. Methods Mol Biol. 2010;595:195-203. 1991;17:269-281.
97. Palis J. Ontogeny of erythropoiesis. Curr Opin Hematol. 2008;15:155-161. 130. Tavian M, Biasch K, Sinka L, et al. Embryonic origin of human hematopoiesis. Int J
98. Palis J, Malik J, McGrath KE, et al. Primitive erythropoiesis in the mammalian embryo. Dev Biol. 2010;54:1061-1065.
Int J Dev Biol. 2010;54:1011-1018. 131. Tavian M, Peault B. The changing cellular environments of hematopoiesis in human
99. Papadaki HA, Stamatopoulos K, Damianaki A, et al. Activated T-lymphocytes with development in utero. Exp Hematol. 2005;33:1062-1069.
myelosuppressive properties in patients with chronic idiopathic neutropenia. Br J 132. Telen MJ. Red blood cell surface adhesion molecules: their possible roles in normal
Haematol. 2005;128:863-876. human physiology and disease. Semin Hematol. 2000;37:130-142.
100. Penn PE, Jiang DZ, Fei RG, et al. Dissecting the hematopoietic microenvironment. 133. Tiedemann K. On the yolk sac of the cat. II. Erythropoietic phases, ultrastructure of
IX. Further characterization of murine bone marrow stromal cells. Blood. 1993;81: aging primitive erythroblasts, and blood vessels. Cell Tissue Res. 1977;183:71-89.
1205-1213. 134. Tiedemann K, van Ooyen B. Prenatal hematopoiesis and blood characteristics of the
101. Perkins PC, Grindem CB, Cullins LD. Flow cytometric analysis of punctate and aggre- cat. Anat Embryol (Berl). 1978;153:243-267.
gate reticulocyte responses in phlebotomized cats. Am J Vet Res. 1995; 135. Tizard IR. Veterinary Immunology. An Introduction. 8th ed. Philadelphia: Saunders
56:1564-1569. Elsevier; 2009.
102. Pietila I, Vainio S. The embryonic aorta-gonad-mesonephros region as a generator of 136. Tober J, Koniski A, McGrath KE, et al. The megakaryocyte lineage originates from
haematopoietic stem cells. APMIS. 2005;113:804-812. hemangioblast precursors and is an integral component both of primitive and of
103. Ponka P, Beaumont C, Richardson DR. Function and regulation of transferrin and definitive hematopoiesis. Blood. 2007;109:1433-1441.
ferritin. Semin Hematol. 1998;35:35-54. 137. Tornquist SJ, Crawford TB. Suppression of megakaryocyte colony growth by plasma
104. Porter DL, Goldberg MA. Regulation of erythropoietin production. Exp Hematol. from foals infected with equine infectious anemia virus. Blood. 1997;90:2357-2363.
1993;21:399-404. 138. Tsuji-Takayama K, Otani T, Inoue T, et al. Erythropoietin induces sustained phosphory-
105. Prchal JT. Production of erythrocytes. In: Lichtman MA, Beutler E, Kipps TJ, et al, eds. lation of STAT5 in primitive but not definitive erythrocytes generated from mouse
Williams Hematology. 7th ed. New York: McGraw-Hill; 2006:393-403. embryonic stem cells. Exp Hematol. 2006;34:1323-1332.
106. Prosper F, Verfaillie CM. Regulation of hematopoiesis through adhesion receptors. 139. Ueno H, Weissman IL. The origin and fate of yolk sac hematopoiesis: application of
J Leukoc Biol. 2001;69:307-316. chimera analyses to developmental studies. Int J Dev Biol. 2010;54:1019-1031.
140. Vidal M. Exosomes in erythropoiesis. Transfus Clin Biol. 2010;17:131-137. 149. Wu H, Riha GM, Yang H, et al. Differentiation and proliferation of endothelial pro
141. Vogt C, Pentz S, Rich IN. The role for the macrophage in normal hematopoiesis: III. genitor cells from canine peripheral blood mononuclear cells. J Surg Res.
In: vitro and in vivo erythropoietin gene expression in macrophages detected by in 2005;126:193-198.
situ hybridization. Exp Hematol. 1989;17:391-397. 150. Yao H, Liu B, Wang X, et al. Identification of high proliferative potential precursors
142. Wardlaw A. Eosinophils and their disorders. In: Lichtman MA, Beutler E, Kipps TJ, et with hemangioblastic activity in the mouse aorta-gonad-mesonephros region. Stem
al, eds. Williams Hematology. 7th ed. New York: McGraw-Hill; 2006:863-878. Cells. 2007;25:1423-1430.
143. Waugh RE. Reticulocyte rigidity and passage through endothelial-like pores. Blood. 151. Ye M, Graf T. Early decisions in lymphoid development. Curr Opin Immunol.
1991;78:3037-3042. 2007;19:123-128.
144. Waugh RE, Mantalaris A, Bauserman RG, et al. Membrane instability in late-stage 152. Yin T, Li L. The stem cell niches in bone. J Clin Invest. 2006;116:1195-1201.
erythropoiesis. Blood. 2001;97:1869-1875. 153. Yona S, Jung S. Monocytes: subsets, origins, fates and functions. Curr Opin Hematol.
145. Weiss G, Goodnough LT. Anemia of chronic disease. N Engl J Med. 2010;17:53-59.
2005;352:1011-1023. 154. Zermati Y, Fichelson S, Valensi F, et al. Transforming growth factor inhibits erythro-
146. Weiss L. The hematopoietic microenvironment of the bone marrow: an ultrastructural poiesis by blocking proliferation and accelerating differentiation of erythroid progeni-
study of the stroma in rats. Anat Rec. 1976;186:161-184. tors. Exp Hematol. 2000;28:885-894.
147. Weiss L. The Blood Cells and Hematopoietic Tissues. New York: Elsevier; 1984. 155. Zhang Y, Wallace DL, de Lara CM, et al. In vivo kinetics of human natural killer
148. Wolber EM, Jelkmann W. Thrombopoietin: the novel hepatic hormone. News Physiol cells: the effects of ageing and acute and chronic viral infection. Immunology. 2007;
Sci. 2002;17:6-10. 121:258-265.
C H A P T E R
4
Evaluation of Erythrocytes
Each hemoglobin tetramer is capable of binding four mol-

N O R M A L ERY T H RO C Y T E S
ecules of O2 when fully oxygenated. The initial binding of a
Erythrocyte Morphology molecule of O2 to a monomer of tetrameric deoxygenated
Erythrocytes from all mammals are anucleated, and most are hemoglobin facilitates further binding of O2 to the hemoglo-
in the shape of biconcave discs called discocytes (Figs. 4-1, bin molecule. The changing O2 affinity of hemoglobin with
4-2).205 The biconcave shape results in the central pallor of oxygenation results in a sigmoid O2 dissociation curve (Fig.
erythrocytes observed in stained blood films. Among common 4-9), when the percent saturation of hemoglobin with O2
domestic animals, biconcavity and central pallor are most pro- is graphed against the PO2. The steepness of the middle
nounced in dogs (Figs. 4-3, 4-4), which also have the largest portion of the curve is of great physiologic significance
erythrocytes. Other species do not consistently exhibit central because it covers the range of O2 tensions present in tissues.
pallor in erythrocytes on stained blood films. The apparent Consequently a relatively small decrease in O2 tension in
benefit of the biconcave shape is that it gives erythrocytes high tissues results in substantial O2 release from hemoglobin.
surface area : volume ratios and allows for deformations that The overall affinity of hemoglobin for O2 is decreased by
must take place as they circulate. Erythrocytes from goats increasing H+, CO2, temperature, and, in most mammals,
generally have a flat surface with little surface depression; 2,3-diphosphoglycerate (2,3DPG). There is a direct correla-
a variety of irregularly shaped erythrocytes (poikilocytes) may tion between body weight and the O2 affinity of hemoglobin
be present in clinically normal goats (Fig. 4-5). Erythrocytes in whole blood (lower body weight, lower O2 affinity) when
from animals in the Camelidae family (camels, llamas, vicuñas, various species of mammals are compared.205
and alpacas) are anucleated, thin, elliptical cells termed ellip- The O2 affinity of fetal blood is greater than that of mater-
tocytes or ovalocytes (Fig. 4-6). They are not biconcave in nal blood except in the cat. Differences in fetal versus maternal
shape and are minimally deformable.437 Erythrocytes from O2 affinity may potentiate O2 transport from the mother to
birds (Fig. 4-7), reptiles, and amphibians are also elliptical in the fetus. However, the fetus is subjected to low arterial O2
shape, but they contain nuclei and are larger than mammalian tensions, and the increased O2 affinity of fetal blood is likely
erythrocytes. Blood cells in salamanders are the largest needed to more fully saturate hemoglobin with O2.205
recognized (Fig. 4-8). The ability of plasma to carry CO2 is small, but the car-
bonic anhydrase reaction in erythrocytes increases the CO2-
Erythrocyte Functions carrying capacity of blood 17-fold by rapidly converting CO2
Mammalian erythrocytes normally circulate for several to carbonic acid (H2CO3). The H2CO3 spontaneously ionizes
months in blood despite limited synthetic capacities and to H+ and bicarbonate (HCO3−). The HCO3− diffuses out of
repeated exposures to mechanical and metabolic insults. the cell down a concentration gradient and chloride (Cl−)
Erythrocytes have three functions: transport of oxygen (O2) moves in (chloride shift) to maintain electrical neutrality.
to tissue, transport of carbon dioxide (CO2) to the lungs, and These processes are reversed at the lungs. Some CO2 is also
buffering of hydrogen ions (H+). In nonanemic animals, the transported bound to hemoglobin as carbamino groups.
presence of hemoglobin within erythrocytes increases the Deoxyhemoglobin binds about twice the CO2 that oxyhemo-
O2-carrying capacity of blood more than 50 times that of globin does.205
plasma without erythrocytes. The O2 content of blood depends Hemoglobin is the major protein buffer in blood. Deoxy-
on the blood hemoglobin content, the partial pressure of dis- hemoglobin is a weaker acid than oxyhemoglobin. Conse-
solved oxygen (PO2) in blood, and the affinity of hemoglobin quently, when oxyhemoglobin releases its O2 in the tissues,
for O2. the formation of deoxyhemoglobin results in increased binding
49

Scanning electron photomicrograph of a normal horse erythrocyte called Scanning electron photomicrograph of dog erythrocytes (discocytes).
a discocyte.
Courtesy of K. S. Keeton and N. C. Jain.
From Stockham SL, Harvey JW, Kinden DA. Equine glucose-6-phosphate
dehydrogenase deficiency. Vet Pathol. 94;31:518-527.
FIGURE 4 -4
FIGURE 4-2
Blood from a dog with acute blood-loss anemia and normal erythrocyte
Blood film from a horse. Most erythrocytes are adhered together like morphology. Erythrocytes exhibit prominent central pallor. Two mature
stacks of coins (rouleaux), a normal finding in this species. Individual neutrophils and a platelet (bottom right corner) are also present. Wright-
nonadherent erythrocytes exhibit central pallor as a result of their bicon- Giemsa stain.
cave shape. A basophil with purple granules is present in the bottom
right of the image. Wright-Giemsa stain.
the outer layer of the membrane, with carbohydrate groups

of H+. Hemoglobin buffers the effects of H2CO3 and allows extending outward. Some blood group antigens are glycolip-
for the isohydric transport of CO2. Hemoglobin also buffers ids, with their specificity residing in the carbohydrate moieties
organic acids produced during metabolism.205 (see Chapter 6 for a discussion of clinically significant blood
groups).205
Erythrocyte Biochemistry Membrane proteins consist of integral membrane proteins
Membrane Structure and Function that penetrate the lipid portion, often spanning the bilayer,
The erythrocyte membrane contains a phospholipid bilayer and skeletal proteins that form or attach to the internal surface
with molecules of unesterified cholesterol intercalated between of the lipid bilayer. Glycoproteins associated with the mem-
fatty acid chains. Phospholipids can move in various ways and brane are integral membrane proteins with the carbohydrate
contribute to membrane fluidity. Glycolipids are located on residues extending from the outside surface of the cell
C ha p ter 4 n Evaluation of Erythrocytes 51

Poikilocytes in blood from a normal goat. Note the small size of the Blood film from a macaw. Nucleated erythrocytes and three heterophils
erythrocytes compared with the neutrophil in the left part of the image. are present.

Elliptocytes in blood from a normal llama. Wright-Giemsa stain. Blood film prepared by mixing equal parts of blood from a salamander
(Amphiuma means) with that of a domestic cat to demonstrate the large
size of the nucleated erythrocytes and a neutrophil in the salamander.
membrane. They carry erythrocyte antigens and function as Most of the cat erythrocytes are echinocytes, a shape artifact of sample
handling in this instance. Wright-Giemsa stain.
receptors or transport proteins (e.g., band 3 is an anion trans-
porter). The membrane skeleton is composed of various pro- Courtesy of H. L. Wamsley.
teins located in a lattice-like arrangement on the inner surface
of the erythrocyte membrane. This meshwork is attached to
the membrane by binding to transmembrane proteins. The dog with protein 4.1 deficiency and another with mutant
membrane skeleton is a major determining factor of mem- β-spectrin.122,205 Neither dog with elliptocytosis was anemic,
brane shape, deformability, and durability. It is in a condensed but the protein 4.1-deficient dog had a reticulocytosis, indi-
configuration in intact cells and can be stretched considerably cating a shortened erythrocyte life span. Hereditary stomato-
without rupturing.205 cytosis occurs in multiple dog breeds, but the specific
Inherited membrane defects can result in abnormally membrane defects have not been reported.205
shaped erythrocytes with shortened erythrocyte life spans and
variable degrees of anemia. Band 3 deficiency in cattle results ATP Generation
in marked spherocytosis with membrane instability and severe Mammalian erythrocytes lack nuclei; therefore they cannot
anemia.232 Hereditary elliptocytosis has been reported in one synthesize DNA or RNA. They also lack ribosomes,
100 is stimulated by increased blood inorganic phosphate (Pi) con-

Decr: 2,3DPG centration and increased blood pH, both of which stimulate
Temp anaerobic glycolysis. 2,3DPG is the most abundant organic
CO2 phosphate in the erythrocytes of most species but its concen-
75
H
tration is low in erythrocytes of Felidae (including domestic
cats), Bovidae (cattle, sheep, and goats), and Cervidae (deer).415
Animals with high erythrocyte 2,3DPG concentrations,
% Saturation
50 including dogs and horses, have the potential to alter their

Incr: 2,3DPG hemoglobin O2 affinity to meet their metabolic needs. The
Temp significance (and, in some cases, the appropriateness) of
CO2 alterations in 2,3DPG in disease states is not always clear.
25 H
Erythrocyte 2,3DPG concentration increases in some anemic
animals, and the resultant decrease in hemoglobin O2 affinity
would seem to be beneficial.205 Increased 2,3DPG has also
been reported in erythrocytes from horses with hypoxic con-
0 ditions.173 In the case of severe hypoxic hypoxemia, the
20 40 60 80 100
response might be detrimental because hemoglobin cannot be
pO2 (mmHg) fully saturated.241 High-altitude camelids (including llamas,
alpacas, guanacos, and vicuñas) have erythrocytes with high
FIGURE 4-9
hemoglobin oxygen affinity, even though their erythrocytes
Hemoglobin-oxygen dissociation curve and factors influencing the posi-
tion of the curve.
have relatively high 2,3DPG concentrations, because their
hemoglobin exhibits low reactivity toward 2,3DPG.377 The P50
From Harvey JW. The erythrocyte: physiology, metabolism, and biochemical for greyhound erythrocytes in whole blood is lower than that
disorders. In: Kaneko JJ, Harvey JW, Bruss ML, eds. Clinical Biochemistry for mongrel dogs, yet the groups have similar 2,3DPG con-
of Domestic Animals, 6th ed. San Diego, CA: Academic Press; 2008: centrations.458 The cause of this difference remains to be deter-
173-240. mined, but it could reflect a low reactivity to 2,3DPG. It is
suggested that the higher hematocrit found in greyhound
mitochondria, and endoplasmic reticula; consequently they dogs may represent a compensatory response to a higher
have no Krebs cycle or electron transport system and are oxygen affinity of hemoglobin in this species.
unable to synthesize proteins or lipids de novo. Glucose is the PFK deficiency inhibits glycolysis above the DPG pathway,
primary substrate for the energy needs of erythrocytes from resulting in markedly decreased 2,3DPG concentrations,
all species except the pig, where inosine appears to be the which makes dog erythrocytes alkaline-fragile. Episodes of
major substrate. Mature erythrocytes depend on anaerobic intravascular hemolysis occur when PFK-deficient dogs
glycolysis for energy (Fig. 4-10). The ATP generated by gly- develop alkalemia secondary to hyperventilation.203
colysis is needed for the maintenance of erythrocyte ionic
composition, shape, and deformability and for limited syn- Oxidant Injury
thetic activities such as glutathione synthesis. Hypophospha- Animals are exposed to low levels of oxidants in their environ-
temia results in decreased erythrocyte glycolytic rates and ments and from normal metabolic processes in the body.
decreased ATP generation. Hemolytic anemia resulting from Reactive oxygen species and reactive nitrogen species are
hypophosphatemia has been reported in diabetic cats and in formed as products of normal cellular metabolism. When they
a diabetic dog following insulin therapy, in a cat with hepatic are generated at higher concentrations in disease states, these
lipidosis, and in postparturient cattle and buffaloes.109,205 Defi- free radicals (and the even more potent oxidative metabolites
ciencies of rate-controlling enzymes in glycolysis also result that they produce) can overwhelm protective systems within
in insufficient ATP generation and shortened erythrocyte the body, producing cellular injury and/or destruction.492
survival. Pyruvate kinase (PK)-deficient dogs and cats have Oxidative reactions can damage hemoglobin, enzymes
mild to severe regenerative hemolytic anemia. Phosphofruc- (sulfhydryl groups especially), and the membrane lipids of
tokinase (PFK)-deficient dogs have compensated hemolytic erythrocytes. Methemoglobin forms when hemoglobin iron is
anemia plus sporadic episodes of intravascular hemolysis and oxidized from the +2 to the +3 state. Heinz bodies are inclu-
hemoglobinuria.203 sions that form within erythrocytes following the oxidative
denaturation of the globin portion of hemoglobin. Membrane
2,3DPG Pathway damage can result in intravascular hemolysis or erythropha-
2,3DPG binds to hemoglobin and reduces the affinity of gocytosis and shortened erythrocyte life spans.205
hemoglobin for oxygen in erythrocytes from most mammals. Protection against Oxidant Injury. NADPH generated
It is produced from a side pathway of the anaerobic glycolysis in the pentose phosphate pathway (PPP) provides electrons
pathway. No net ATP is generated when molecules traverse for protection against oxidants. It is needed to maintain glu-
this DPG pathway. The formation of 2,3DPG in erythrocytes tathione and thioredoxin in their reduced states, and it is
O2 O2 2H
SOD
O2
H2O2 2H2O
GPx
GSH GSSG
GR
NADP NADPH
HK
Glucose G6P 6PG
G6PD
NADP LMB
ATP ADP GPI
6PGD NADPH-D
F6P
ATP NADPH MB
PFK CO2
ADP TK TA
Pentose PO4
FDP
TPI
DHAP G3P
NAD Pi
GAPD
NADH
1,3 DPG
DPGM ADP
2,3DPG PGK
ATP
Pi 3PG
ADP ATP NADH NAD
2PG PEP Pyruvate Lactate

PK LDH
FIGURE 4-10
Metabolic pathways of the mature erythrocyte. HK, hexokinase; GPI, glucose phosphate isomerase; PFK,
phosphofructokinase; TPI, triosephosphate isomerase; GAPD, glyceraldehyde-3-phosphate dehydrogenase;
PGK, phosphoglycerate kinase; MPGM, monophosphoglycerate mutase; DPGM, diphosphoglycerate mutase;
PK, pyruvate kinase; G6PD, glucose-6-phosphate dehydrogenase; 6PGD, 6-phosphogluconate dehydroge-
nase; LDH, lactate dehydrogenase; GR, glutathione reductase; GPx, glutathione peroxidase; TK, transketolase;
TA, transaldolase; GSSG, oxidized glutathione; G6P, glucose 6-phosphate; F6P, fructose 6-phosphate;
FDP, fructose 1,6-diphosphate; DHAP, dihydroxyacetone phosphate; GAP, glyceraldehyde 3-phosphate;
1,3DPG, 1,3-diphosphoglycerate; 2,3DPG, 2,3-diphosphoglycerate; 3PG, 3-phosphoglycerate; 2PG,
2-phosphoglycerate; PEP, phosphoenolpyruvate; ADP, adenosine diphosphate; ATP, adenosine triphosphate;
NAD, nicotinamide adenine dinucleotide; NADH, reduced nicotinamide adenine dinucleotide; NADP, nico-
tinamide adenine dinucleotide phosphate; NADPH, reduced nicotinamide adenine dinucleotide phosphate;
GSH, reduced glutathione; Pi, inorganic phosphate; SOD, superoxide dismutase.
From Harvey JW. The erythrocyte: physiology, metabolism, and biochemical disorders. In: Kaneko JJ, Harvey JW, Bruss
ML, eds. Clinical Biochemistry of Domestic Animals, 6th ed. San Diego, CA: Academic Press; 2008:173-240.
important in maintaining catalase in a functional form. eccentrocytosis has been described in an American saddlebred
Defects in the PPP can render erythrocytes susceptible to colt with less than 1% of normal G6PD activity.452
endogenous and exogenous oxidant injury. Glucose-6- Reduced glutathione (GSH) is a tripeptide containing a
phosphate dehydrogenase (G6PD) is the rate-controlling highly reactive sulfhydryl group that may act nonenzymati-
enzyme in the PPP. A persistent hemolytic anemia with cally as a free radical acceptor to counteract oxidant damage.
GSH also functions as an electron donor in various reductive Circulating

enzyme reactions including glutathione peroxidase (GPx), erythrocytes
phospholipid hydroperoxide glutathione peroxidase, glutathi-
one S-transferase, and glutaredoxin. Following oxidation, glu-
tathione forms a disulfide (GSSG) that can be reduced back Marrow erythrocyte
to GSH by the flavin adenine dinucleotide (FAD)-dependent precursors Macrophages
glutathione reductase (GR) enzyme, using NADPH as the
Hb
source of electrons (see Fig. 4-10). Horses with erythrocyte Iron
FAD deficiency have markedly reduced GR activity, decreased cycle
Fe Ferritin/Hemosiderin
GSH concentration, and prominent eccentrocytosis.203,212
Selenium acts as an antioxidant when incorporated as
selenocysteine at the active site of a wide range of selenopro- Transferrin-Fe
teins, including GPx, phospholipid hydroperoxide glutathione Liver complex
Fe GI
peroxidase, and thioredoxin reductase in erythrocytes.73 Heinz Transferrin
body hemolytic anemia has been reported in selenium- Plasma
deficient cattle grazing on St. Augustine grass.327 Catalase is
an enzyme that can catalyze the conversion of H2O2 to water FIGURE 4 -11
and O2 without using energy. Iron cycle. Iron (Fe) is highly conserved in the body. Iron in plasma
is bound to transferrin, a transport protein that is synthesized in the
Recent studies suggest that peroxiredoxins may be more
liver. Iron is transported to all tissues, but most iron is utilized to syn-
important in protecting against H2O2 than GPx or catalase.285 thesize hemoglobin in developing erythroid cells. Aged blood erythro-
Oxidized peroxiredoxins are regenerated using reduced thio- cytes are phagocytized by macrophages and hemoglobin is degraded.
redoxin, and oxidized thioredoxin is reduced by NADPH Released iron is either returned to plasma or stored in macrophages as
using thioredoxin reductase.277 Ascorbate functions as an anti- ferritin and hemosiderin. Nearly all of the iron in plasma under normal
conditions comes from the release of iron by macrophages that have
oxidant by donating one or two electrons to a variety of oxi-
phagocytized and degraded erythrocytes. Only about 3% of the iron in
dants, including oxygen free radicals and peroxides. Vitamin plasma results from gastrointestinal (GI) enterocyte absorption in
E is lipid-soluble and acts as a free radical scavenger in the normal individuals.
membrane.205
From Harvey JW. Iron metabolisms and its disorders. In: Kaneko JJ,
Methemoglobin Formation and Reduction Harvey JW, Bruss ML, eds. Clinical Biochemistry of Domestic Animals,
6th ed. San Diego, CA: Academic Press; 2008:259-285.
About 3% of hemoglobin (Fe+2) is oxidized to methemoglobin
(Fe+3) each day. Methemoglobin is unable to bind O2. To
prevent hypoxemia, which would result from the accumulation
of a high level of methemoglobin, the methemoglobin formed
is reduced back to functional hemoglobin in a reaction that lysed, hemoglobin is degraded, and iron is released. Most iron
requires the cytochrome-b5 reductase (Cb5R) enzyme and released from degraded hemoglobin is quickly released back
NADH generated by anaerobic glycolysis.205 An inherited into plasma, but a small amount may be stored as ferritin or
deficiency in Cb5R in dogs and cats results in persistent met- hemosiderin within macrophages, which is released more
hemoglobinemia with minimal or no clinical signs.203 Methe- slowly into plasma. The vast majority of iron entering plasma
moglobinemia also occurs in horses that have decreased Cb5R each day comes from macrophage release.204
activity secondary to erythrocyte FAD deficiency.212 About 60% to 70% of total body iron is present in hemo-
globin (3.4 mg iron per gram of hemoglobin). About a third
Iron Metabolism of total body iron is stored as ferritin and hemosiderin (pri-
Iron metabolism is presented in this chapter because more marily within macrophages), 3% to 7% is present in myoglo-
iron is needed for the production of erythrocytes than for all bin (with the higher values occurring in dogs and horses), 1%
other cells in the body combined. Iron is absorbed from the is present in hemoprotein and flavoprotein enzymes, and only
diet in the small intestine and transferred to plasma, where it 0.1% is bound to transferrin in plasma.204
is bound to transferrin for transport to cells within the body. Iron Absorption. The absorption of iron from the diet
Once inside the body, iron cycles in a nearly closed system depends upon age, species, iron stores, rate of erythropoiesis,
(Fig. 4-11) because little iron is lost in domestic animals unless inflammation, and pregnancy, as well as the amount and
hemorrhage occurs. About 75% of the iron present in plasma chemical form of iron ingested. A low percentage of
will be transported to the bone marrow for incorporation into dietary iron is absorbed in normal adult animals. Iron absorp-
hemoglobin in developing erythroid cells.436 The remaining tion occurs through enterocytes of the duodenum and proxi-
plasma iron is taken up by nonerythroid tissues, primarily the mal jejunum. Iron can be taken in by enterocytes as free ions
liver.266 Erythrocytes containing hemoglobin normally circu- or as heme by different pathways (Fig. 4-12). The relative
late for several months before being phagocytized by macro- importance of these pathways varies depending on species and
phages when senescent. After phagocytosis, erythrocytes are diet.204
Intestinal Plasma
lumen Fe3 Ferritin
Fe3 mTf Fe3
DcytB Fe3 Fe3
Fe3 Fe3
Fe2 Fe2
Fe3 aTf
DMT1
Hemoglobin Fe2 Fe3
and myoglobin
Hephaestin
HO
Fe Fe Fe2 Fe2 Fe2
Heme HCP1 Ferroportin
Enterocyte
Hepcidin
Globin
FIGURE 4-12
Mechanisms of iron absorption. Ferrous iron (Fe+2) ions are transported into enterocytes in the duodenum by
the divalent transporter-1 (DMT1) after reduction of ferric iron (Fe+3) ions using a duodenal cytochrome b
(DcytB). Heme is transported into enterocytes using heme carrier protein-1 (HCP1). Once inside, inorganic
iron is released from heme by the action of the heme oxygenase (HO) reaction. Fe+2 ions are exported from
enterocytes using ferroportin, oxidized to Fe+3 using hephaestin, and bound by apotransferrin (aTf ) to form
monoferric transferrin (mTf ) and diferric transferrin (not shown). Hepcidin in plasma inhibits iron export
to plasma by interacting directly with ferroportin, leading to ferroportin’s internalization and lysosomal deg-
radation. Fe+2 not transported to plasma is stored as ferritin following oxidation to Fe+3. Iron stored as ferritin
is returned to the lumen of the small intestine when enterocytes are sloughed at the tip of the villus.
From Harvey JW. Iron metabolism and its disorders. In: Kaneko JJ, Harvey JW, Bruss ML, eds. Clinical Bio
chemistry of Domestic Animals. 6th ed. San Diego, CA: Academic Press; 2008:259-285.
Most inorganic iron in the diet is in the ferric (Fe+3) state. fates, export or storage, depending on the body’s iron needs.
+3
Fe iron is solubilized from food by hydrochloric acid in the If iron is required by the body, molecules will be transported
stomach and binds to mucins and various small molecules in from enterocytes to transferrin in plasma. This transportation
the stomach, which keep the iron soluble and available for is mediated by ferroportin, an iron transport protein located
absorption in the more alkaline environment of the small on the basolateral surface of mature enterocytes. In addition
intestine. The most important pathway for nonheme iron to ferroportin, the efflux of iron from enterocytes requires a
uptake utilizes the divalent metal transporter-1 (DMT1). Fe+3 copper-containing protein called hephaestin, which is also
ions must be reduced to ferrous (Fe+2) ions before they can be located on the basolateral membranes of mature enterocytes.
transported into the enterocyte via the DMT1. Although Hephaestin is a membrane-bound ferroxidase that has signifi-
some Fe+3 ion reduction may occur by direct interaction with cant homology to the plasma protein ceruloplasmin. Hepha-
dietary ascorbic acid, most reduction appears to rely on duo- estin’s function may relate to its ability to oxidize Fe+2 ions to
denal cytochrome b (DcytB) and possibly other brush border Fe+3 ions for binding to transferrin in plasma.20
ferrireductase enzymes. Although humans absorb Fe+2 salts If body iron requirements are low, enterocyte cytoplasmic
more readily from the intestine than Fe+3 salts, dogs are iron accumulates. Free iron is toxic; consequently the mucosal
reported to absorb both valence forms equally well.204 cell protects itself by increasing apoferritin synthesis and
Heme is released from dietary myoglobin and hemoglobin incorporating the excess iron into ferritin. Each ferritin mol-
by the action of digestive enzymes. Dietary heme iron is gen- ecule is composed of a protein shell of 24 apoferritin subunits
erally more bioavailable than is nonheme iron and is an surrounding a central core of up to 4500 iron atoms as ferric
important nutritional source of iron in carnivores and omni- oxyhydroxide. Ferritin is a storage protein that prevents free
vores. Heme enters duodenal enterocytes as an intact metal- iron from catalyzing oxidative reactions, which would injure
loporphyrin, possibly using a brush border transporter named the cell. Ferritin within mucosal cells is returned to the small
heme carrier protein 1 (HCP1). However, this protein trans- intestine lumen when enterocytes are sloughed at the tip of
ports folate more efficiently than heme; consequently, its the villus after 1 to 2 days.445
physiologic role in intestinal heme uptake remains to be Iron absorption is increased when total body iron content
clearly defined. After heme absorption, iron is released from is low or erythropoiesis is increased. Iron absorption is
heme by the action of the heme oxygenase reaction.20 decreased when total body iron content is high or inflamma-
Once within the enterocyte, intracellular iron molecules are tion is present.204 Components of brush-border iron uptake,
likely bound to one or more chaperone molecules. A potential including DMT1 and DcytB, are strongly influenced by the
chaperone named poly (rC)-binding protein 1 (PCBP1) has iron concentration within enterocytes, with increased compo-
been described.422 Iron taken up by enterocytes has one of two nents expressed when intracellular iron content is low and
decreased components expressed when iron content is high. form heme. A direct interorganelle transfer of iron occurs
These locally responsive changes in brush-border transport between endosomes and mitochondria in developing ery-
components help buffer the body against the absorption of throid cells.419 Some iron is presumably released from endo-
excessive iron, but it is the control of the basolateral transport somes into a cytoplasmic labile iron pool, with excess
of iron from enterocytes to plasma that represents the primary cytoplasmic iron stored as ferritin. TfR and apoferritin syn-
site at which iron absorption is regulated.445 thesis are regulated by the amount of intracellular iron present.
Systemic Control of Iron Metabolism. Hepcidin, a High iron content stimulates apoferritin synthesis and inhib-
peptide secreted by hepatocytes into the circulation, is an its TfR expression to minimize the potential of iron toxicity
important systemic regulator of iron metabolism.153 Its to the cell. Low iron content results in decreased apoferritin
production is modulated by body iron requirements, which synthesis and increased TfR expression on cell surfaces to
are largely influenced by the magnitude of erythropoiesis maximize iron uptake and use for heme synthesis. Free heme
present.354 Hepcidin inhibits iron export from enterocytes, concentration within erythroid cells controls hemoglobin
macrophages, and hepatocytes by interacting directly with synthesis. An increase in free heme promotes the synthesis
ferroportin, leading to the internalization and lysosomal deg- of globin chains and inhibits the uptake of iron from
radation of this iron export protein.20 Hepcidin production is transferrin.204
decreased in iron deficiency and disorders resulting in increased Macrophage Iron Metabolism. Little iron enters macro-
erythropoiesis, which increase the demand for iron.152 As a phages, in contrast to other cell types in the body, via plasma
result, ferroportin receptors are abundant on cell surfaces and transferrin. Rather, nearly all iron enters macrophages by the
dietary iron absorption is increased, as is the export of iron phagocytosis of aged or prematurely damaged erythrocytes
from macrophages and hepatocytes. Conversely, hepcidin pro- (Fig. 4-13).373 Following phagocytosis, erythrocytes are lysed
duction is increased and ferroportin transporter expression on and hemoglobin is degraded to heme and globin. The micro-
cell surfaces is decreased when iron overload is present. Hep- somal heme oxygenase reaction degrades heme and releases
cidin production is also increased during inflammation by a iron. Most of the iron released from degraded heme is quickly
pathway not dependent on body iron requirements.445 exported from the macrophage and bound to plasma trans
Plasma Iron. Nearly all of the iron in plasma is bound to ferrin for transport to other cells (especially erythrocyte
the protein apotransferrin to form transferrin. The binding of
iron to apotransferrin keeps iron molecules soluble and pre-
vents iron-catalyzed oxidative reactions. Apotransferrin is a
β-globulin with two binding sites for Fe+3. Normally, 25% to
50% of the iron-binding sites are saturated with iron. Plasma
iron turns over rapidly in 3 hours or less.436 Nearly all of the
iron in plasma under normal conditions comes from the
Hemosiderin
release of iron by macrophages that have phagocytized and
degraded erythrocytes. In contrast, in normal individuals, only
about 3% of the iron in plasma results from enterocyte
absorption.372 Cp
mTf
Iron Uptake by Erythroid Cells. The delivery of iron
LIP Ferritin
from plasma to developing erythroid cells and other cell types aTf Ferroportin
except macrophages is dependent on transferrin.266 Trans
ferrin (especially diferric transferrin) molecules bind to Hepcidin
transferrin receptor 1 (TfR1) on the surface of cells, and these
FIGURE 4 -13
transferrin-TfR1 complexes invaginate to initiate endocytosis.
Iron metabolism in macrophages. Nearly all iron enters macrophages by
After the transferrin-TfR1 complexes are internalized as the phagocytosis of aged or prematurely damaged erythrocytes. Follow-
endosomes, a proton pump decreases the pH in the endosome, ing phagocytosis, erythrocytes are lysed, and hemoglobin is degraded to
resulting in conformational changes in the proteins and sub- heme and globin. The microsomal heme oxygenase reaction within mac-
sequent release of iron ions from transferrin. The released iron rophages degrades heme and releases iron to the labile iron pool (LIP).
is exported from the endosome using DMT1.349 The resultant Most of the released iron is exported from the macrophage by ferroportin
as ferrous iron, oxidized to ferric iron by ceruloplasmin (Cp) in plasma,
apotransferrin-TfR1 complex is recycled to the cell mem- and bound to apotransferrin (aTf ) to form monoferric transferrin (mTf )
brane, where apotransferrin is released from the cell, and the or diferric transferrin (not shown). Hepcidin in plasma inhibits iron
TfR1 is again available for binding additional iron-containing export by interacting directly with ferroportin, leading to ferroportin’s
transferrin molecules. Erythroid precursor cells in the bone internalization and lysosomal degradation. Iron not rapidly released to
marrow and reticulocytes that synthesize hemoglobin have plasma is stored within macrophages as ferritin, which may be degraded
to hemosiderin within lysosomes.
TfR1 on their surfaces for iron uptake, but reticulocytes lose
their TfR1 as they develop into mature erythrocytes.372 From Harvey JW. Iron metabolism and its disorders. In: Kaneko JJ, Harvey
After its release from transferrin, iron is transported to the JW, Bruss ML, eds. Clinical Biochemistry of Domestic Animals. 6th ed.
mitochondria, where it is incorporated into protoporphyrin to San Diego, CA: Academic Press; 2008:259-285.
precursors in the bone marrow).204 The export of iron from antigenic remains to be clarified, but oxidative mechanisms
macrophages is mediated by ferroportin and controlled by are probably involved. A natural antibody against the senes-
hepcidin, as has been discussed for enterocytes.495 The copper- cent cell antigen is present in human plasma. This antibody
containing plasma protein ceruloplasmin oxidizes Fe+2 ions to binds to senescent cell antigens on the surface of aged cells
Fe+3 ions for binding to transferrin in plasma.341 and, together with bound complement, promotes the phago-
The mononuclear phagocyte system accounts for much of cytosis of aged erythrocytes by macrophages that exhibit Fc
the total body iron stores. Iron not rapidly released to plasma and C3b surface receptors. Senescent dog erythrocytes accu-
is stored within macrophages as ferritin and hemosiderin. Free mulate surface-associated immunoglobulin, which is believed
cytoplasmic ferritin molecules are visible by electron micros- to promote their removal by macrophages.395 The relative
copy but not by light microscopy. Hemosiderin is composed importance of the immune- and nonimmune-mediated
of aggregates of protein and iron within lysosomes. It is insol- phagocytosis of senescent erythrocytes remains to be
uble in water and thought to result from the degradation of clarified.
ferritin. Hemosiderin is visible by light microscopy when it is Erythrocytes lose volume by shedding microvesicles as they
stained with an iron stain (Prussian blue stain). Iron in the age. The composition of the resultant microvesicles varies, but
storage pool turns over slowly unless there is an increased need they typically contain hemoglobin. Other components that
for iron for hemoglobin synthesis.45 may be present include glycophorin A, breakdown products
of band 3 (senescent antigen), IgG, and exposed phosphati-
dylserine. They do not contain the skeletal proteins spectrin
ERY T H RO C Y T E D E S T R U C T I O N and ankyrin.186,536 Microvesicles are rapidly cleared from the
Normal Removal of Aged Erythrocytes circulation by macrophages, using receptors discussed above
Most erythrocytes circulate in blood for a finite time period for aged erythrocytes.186 It may be that this process of microve-
(survival time or life span) ranging from 2 to 5 months in siculation removes patches of damaged membrane that would
domestic animals, depending on the species. Erythrocyte life otherwise bind to macrophages and result in the early removal
spans are related to body weight (and consequently metabolic of otherwise healthy erythrocytes.536
rate), with the smallest animals (highest metabolic rate) Following phagocytosis by macrophages of the spleen, liver,
having the shortest erythrocyte life spans. Greyhound dogs and other organs, erythrocytes and erythrocyte microvesicles
are often used as blood donors. The erythrocyte life span of 6 are lysed and hemoglobin is degraded to heme and globin
greyhound dogs (mean 93 days) was not significantly different (Fig. 4-14). Globin is catabolized to constituent amino acids,
from that of 3 nongreyhound dogs (103 days).158 Aged eryth- and the microsomal heme oxygenase reaction within macro-
rocytes are phagocytized by macrophages of the mononuclear phages degrades heme to iron, biliverdin, and carbon
phagocyte system. Oxidative injury, especially near the end of
their life span, appears to be responsible for normal erythro-
cyte aging and removal.205 Oxidative damage and other stress-
Mononuclear phagocytes
ors can induce suicidal death of erythrocytes (eryptosis), with
Hb
reactions similar to some of those that occur during apoptosis Fe
Globin
of nucleated cells. Eryptosis is characterized by Ca+2 entry, Hb Globin
erythrocyte shrinkage, membrane blebbing (microvesicle for- Heme RBCs Biliverdin
mation), and cell membrane phospholipid scrambling, with Fe
Bilirubin
CO
Porphyrins
phosphatidylserine exposure on the cell surface.271
Surface membrane alterations on aged or damaged cells ALA
that may be recognized by macrophages include exposure of Bilirubin-albumin
phosphatidylserine on the external surface, modified external Glycine complex
membrane carbohydrate residues (e.g., desialation of sialogly-
Succinate
coproteins), and/or modified membrane proteins (e.g.,
Conjugated Nonhemoglobin
partially degraded band 3); these are possible signals for Bone marrow bilirubin
Liver
heme
removal.57,246,265 Phosphatidylserine is normally localized in
the inner leaflet of the lipid bilayer, but with cell damage
phosphatidylserine may be exposed on the outer leaflet of the
lipid bilayer, where it can be bound by phosphatidylserine Urobilinoid compounds
receptors such as CD36 on the surface of macrophages.249
Fecal urobilinogen
Other macrophage receptors can recognize altered carbo
hydrate moieties on the surface of erythrocytes.
Stercobilin
The appearance of a senescent cell antigen may be an
important signal in the removal of aged erythrocytes.246 This FIGURE 4 -14
senescent cell antigen is derived from the band 3 anion trans- Overview of erythrocyte production, erythrocyte phagocytosis by mono-
porter. The specific alteration required for band 3 to become nuclear phagocytes, hemoglobin degradation, and bilirubin metabolism.
monoxide. Biliverdin is reduced to bilirubin via biliverdin humans include sepsis, malaria, hemoglobinopathies, G6PD
reductase in nearly all mammals. Biliverdin reductase activity deficiency, phosphate depletion, iron deficiency, and hemolytic
is low in rabbits and nutria and almost completely lacking in uremic syndrome.270
birds; consequently biliverdin is the predominant bile pigment Almost no lysis of erythrocytes occurs within the circula-
in these species.101,476 Considerable heterogeneity exists in tion of normal individuals, but intravascular hemolysis can be
reptiles, amphibians, and fish in the production of bilirubin present when severe membrane damage occurs in disease
versus biliverdin.101 Bilirubin is released from macrophages states (Fig. 4-16). Following lysis, hemoglobin in plasma
and bound to albumin for transport to the liver for conjuga- (hemoglobinemia) reversibly dissociates into dimers that bind
tion and excretion. Approximately 80% of the bilirubin pro- almost irreversibly to haptoglobin, an α2-glycoprotein in
duced in the body comes from the degradation of hemoglobin, plasma. The hemoglobin-haptoglobin complex is too large to
with the remainder coming from the degradation of other be filtered through the kidney and is rapidly removed from
heme-containing proteins.476 the circulation following binding to the hemoglobin scavenger
receptor CD163 on macrophages. Once inside the cell,
Pathologic Destruction of Erythrocytes hemoglobin-haptoglobin complexes are transported to lyso-
Increased membrane injury associated with various pathologic somes for degradation, and receptors are recycled to the cell
disorders can result in increased phagocytosis of erythrocytes surface.229 The hemoglobin is degraded and iron is conserved,
by macrophages (see previous discussion concerning the as discussed previously.
removal of aged erythrocytes). Anemia develops if the rate of Once plasma haptoglobin is saturated (about 50 to 150 mg/
erythrocyte destruction exceeds the ability of the bone marrow dL of hemoglobin-binding capacity in dogs, cats, and horses),
to respond by producing new erythrocytes. Lysis of erythro- remaining free hemoglobin dimers are small enough to be
cytes within macrophages after phagocytosis is sometimes readily filtered by the kidney.196 Some hemoglobin is reab-
referred to as extravascular hemolysis. Hyperbilirubinemia sorbed by the proximal tubules, but once that capacity is
may be present within a few hours following substantial exceeded, hemoglobin appears in the urine (hemoglobin-
erythrocyte destruction. uria).211 Plasma appears red when as little as 50 mg/dL of
Increased eryptosis (similar to the apoptosis of nucleated hemoglobin is present; consequently hemoglobinemia may
cells) may contribute to the development of anemia in some be observed in the absence of hemoglobinuria. Hemoglobin
disorders. As discussed earlier, eryptosis is characterized by absorbed by the proximal tubules is rapidly catabolized, and
Ca+2 entry, erythrocyte shrinkage, and membrane blebbing iron is stored as ferritin and hemosiderin.203 Iron that is not
with ectosome formation (Fig. 4-15).271 Reported triggers of reutilized is lost when tubular epithelial cells slough into the
eryptosis include oxidative stress, energy depletion, osmotic urine.
shock, lipid-derived signaling molecules, certain bacterial exo-
toxins, various drugs, and metals including lead, copper, zinc,
and mercury. Diseases associated with accelerated eryptosis in
Erythrocytes
Intravascular
hemolysis
Hemoglobin (Hb)
Haptoglobin
HpHb
dimers “Free” hemoglobin
Hepatocytes
Methemoglobin
A B Kidney
Globin
Heme
FIGURE 4-15 Hemopexin
Presumptive eryptosis with erythrocyte shrinkage and vesicle formation Hemopexin Heme
Hemoglobinuria
in blood from dogs. A, Blood from a dog with regenerative anemia and
hemangiosarcoma. Frequent acanthocytes and echinocytes and occa- Albumin
Methemalbumin Hemopexin
sional schistocytes and keratocytes were also noted in the stained blood
film. Wright-Giemsa stain. B, Blood from a febrile dog with mild non-
regenerative anemia and lymphoma. Low numbers of eccentrocytes and FIGURE 4-16
pyknoctyes and rare hemoglobin crystals were also noted in the stained Pathophysiology of intravascular hemolysis. Methemalbumin forms in
blood film. Wright-Giemsa stain. primates but not in common domestic animals. Hp, haptoglobin.
Free hemoglobin in plasma can spontaneously oxidize to in species other than horses, cats, or pigs should be noted
form methemoglobin, which tends to dissociate into ferri- as an abnormal finding.
heme (hemin) and globin. Free heme binds to a plasma Prominent rouleaux formation results in rapid erythrocyte
protein called hemopexin. The heme-hemopexin complexes sedimentation in whole blood samples allowed to stand undis-
undergo endocytosis after binding to CD91 on the surface of turbed. This characteristic formed the basis of an erythrocyte
macrophages and hepatocytes.229 The binding to hemopexin sedimentation rate test that was done with special sedimenta-
protects cell membranes from toxic effects of free heme, and tion tubes. Increased sedimentation rates after 1 hour were
it also conserves iron. Albumin from primates can also bind suggestive of increased globulins in plasma, as typically seen
heme to form methemalbumin, but albumin from common with inflammation. Unfortunately the sedimentation rate
domestic animals does not bind heme.163 increases as the hematocrit decreases, so correction factors
were required for the HCT. The sedimentation rate has largely
been replaced by making direct measurements of total globu-
A B N O R M A L ERY T H RO C Y T E
lins and fibrinogen and other acute-phase proteins.
MORPHOLOGY
Rouleaux Agglutination
Erythrocytes on blood films from healthy horses, cats, and Aggregation or clumping of erythrocytes in clusters (not in
pigs often exhibit rouleaux (aggregations of erythrocytes chains, as in rouleaux) is termed agglutination (Fig. 4-18).
grouped together like a stack of coins) formations (see Fig. Agglutination is caused by the occurrence of immunoglobu-
4-2). Rouleaux formation depends on both the nature lins bound to erythrocyte surfaces. Because of their pentava-
of the erythrocytes and the composition of plasma.37 Eryth- lent nature, IgM immunoglobulins have the greatest propensity
rocytes that are more deformable and have greater membrane to produce agglutination.129 EDTA-dependent IgM-mediated
fluidity with less negative charge on their surfaces (weaker erythrocyte agglutination has been reported in a cat without
electrostatic repulsive force) aggregate more readily than evidence of hemolysis. Agglutination did not occur if blood
cells with the opposite characteristics.37,443,460 Rouleaux was collected in heparin or citrate.411 High-dose unfraction-
formation also depends on the presence of high-molecular- ated heparin treatment in horses also causes erythrocyte
weight proteins in plasma.132 Increased concentrations of agglutination by an undefined mechanism.319,322
globulin proteins—including fibrinogen, haptoglobin, and
immunoglobulins—potentiate rouleaux formation in associ Polychromasia
ation with inflammatory conditions.11,532 Rouleaux formation The presence of bluish-red erythrocytes in stained blood films
can also occur in association with some lymphoproliferative is called polychromasia (Fig. 4-19). Polychromatophilic
disorders in which one or more immunoglobulins are secreted erythrocytes are reticulocytes that stain bluish-red owing to
in high amounts (Fig. 4-17). Prominent rouleaux formation the combined presence of hemoglobin (red staining) and
FIGURE 4 -18
FIGURE 4-17 Erythrocyte agglutination and spherocyte formation in blood from a dog
Rouleaux formation in blood from a dog with multiple myeloma and a with von Willebrand’s disease after transfusion. A large basophilic eryth-
monoclonal hyperglobulinemia. The cytoplasm of a neutrophil present is rocyte (macroreticulocyte or stress reticulocyte) is present in the upper
pale compared to the background that stains blue because of the increased left corner and an echinocyte is present in the lower left corner. Wright-
protein present in the blood. Wright-Giemsa stain. Giemsa stain.

Increased polychromasia and anisocytosis in blood from a dog with a Blood from a dog with a nonregenerative aplastic anemia secondary to
hemolytic anemia caused by Mycoplasma haemocanis, although no organ- trimethoprim-sulfadiazine therapy. Erythrocyte morphology is normal
isms are present in this field. Three large polychromatophilic erythrocytes except for several erythrocytes with scalloped borders (echinocytes).
(reticulocytes) are present in the central area. A nucleated erythrocyte Wright-Giemsa stain.
(metarubricyte) is present in the upper left. Wright-Giemsa stain.
individual ribosomes and polyribosomes (blue staining). Low

numbers of polychromatophilic erythrocytes are usually seen
in blood from normal dogs and pigs, because up to 1.5%
reticulocytes may be present in dogs and up to 1% reticulo-
cytes may be present in pigs even when the HCT is normal.238
Slight polychromasia may be present in normal cats, but many
normal cats exhibit no polychromasia in stained blood films.
Polychromasia is absent in stained blood films from normal
cattle, sheep, goats, and horses because reticulocytes with suf-
ficient RNA to impart a bluish color are not normally present
in the blood in these species.
The most useful approach in the classification of anemia is
to determine whether or not evidence of a bone marrow
response to the anemia is present in blood. For all species
except the horse, this involves determining whether absolute
reticulocyte numbers are increased in blood. Horses rarely
release reticulocytes from the bone marrow even when an FIGURE 4 -21
increased production of erythrocytes occurs. When an abso- Two exceptionally large basophilic erythrocytes (macroreticulocytes or
lute reticulocytosis is present, the animal is said to have a stress reticulocytes) are present in blood from a dog with immune-
regenerative anemia. The presence of a regenerative response mediated hemolytic anemia. Wright-Giemsa stain.
suggests that the anemia results from either increased eryth-
rocyte destruction or hemorrhage. A nonregenerative anemia
generally indicates that the anemia is the result of decreased the blood (Fig. 4-21). High concentrations of erythropoietin
erythrocyte production (Fig. 4-20); however, about 3 to 5 days shorten the marrow transit time for erythroid cells, resulting
are required for increased reticulocyte production and release in the early release of immature reticulocytes that are twice
by the bone marrow in response to an acute anemia.15,65,146,364 the normal size.379 There is a direct correlation between
Consequently the anemia appears nonregenerative shortly the percentage of polychromatophilic erythrocytes and the
after hemolysis or hemorrhage has occurred (see Fig. 4-4). percentage of reticulocytes in dogs (and presumably in pigs)
Increased polychromasia is usually present in regenerative and between the percentage of polychromatophilic erythro-
anemias because many reticulocytes stain bluish-red with cytes and percentage of aggregate reticulocytes in cats (Fig.
routine blood stains (see Fig. 4-19). When the degree of 4-22).15,269 Cats with mild anemia may not release aggregate
anemia is severe, basophilic macroreticulocytes or so-called reticulocytes from the marrow but will release punctate reticu-
stress reticulocytes or shift reticulocytes may be released into locytes (Fig. 4-23, A).15 Because punctate reticulocytes do not
A A
B B
FIGURE 4 -23
Blood from a FeLV-positive cat with a macrocytic normochromic anemia
(HCT = 23%, MCV = 70 fL, MCHC = 33 g/dL). A, Low numbers of
aggregate reticulocytes but markedly increased punctate reticulocytes
(83% uncorrected) are present. Methylene blue reticulocyte stain.
B, Increased anisocytosis is present but polychromasia is not apparent,
even though most of the blood cells present are punctate reticulocytes,
because there is insufficient RNA present to impart a blue color to the
cytoplasm of these cells. Wright-Giemsa stain.
contain sufficient numbers of ribosomes to impart a bluish

color to the cytoplasm, mild regenerative anemia in cats may
C lack polychromasia in stained blood films (Fig. 4-23, B).
FIGURE 4-22 Anisocytosis

Agglutination of aggregate reticulocytes in blood from a cat with a Variation in erythrocyte diameters in stained blood films is
Coombs’-positive hemolytic anemia. A, Agglutination of polychromato- called anisocytosis (see Fig. 4-23, B). It is greater in normal
philic erythrocytes and a metarubricyte. Wright-Giemsa stain. B, Agglu- cattle than in other normal domestic animals.238 Anisocytosis
tination of aggregate reticulocytes. New methylene blue reticulocyte is increased when different populations of cells are present,
stain. C, Agglutination of aggregate reticulocytes. New methylene blue
wet mount preparation. as can occur following a transfusion (Fig. 4-24). Anisocytosis
may occur when substantial numbers of smaller than normal
cells are produced, as occurs with iron deficiency, or when
substantial numbers of larger than normal cells are produced,
as occurs when increased numbers of reticulocytes are

Marked anisocytosis in blood from a 6-week-old kitten after a blood Hypochromic erythrocytes in blood from a dog with iron deficiency
transfusion. The RDW (43%) was also markedly increased. The kitten secondary to chronic blood loss resulting from a persistent flea infesta-
presented with marked lipemia and a severe iron-deficiency anemia. The tion. Not only is the center of each cell paler than normal but the
small hypochromic erythrocytes are from the kitten and the larger eryth- diameter of the area of central pallor is increased relative to the red-
rocytes from the blood donor cat. Two lysed erythrocytes (red smudges) staining periphery of the cell. A polychromatophilic erythrocyte (reticu-
are present (left center) because lipemia enhances erythrocyte lysis during locyte) is present near the left edge of the image. Wright-Giemsa stain.
blood film preparation. Wright-Giemsa stain.

Increased anisocytosis in blood from a horse with a regenerative anemia Hypochromic erythrocytes in blood from a dog with iron deficiency
resulting from internal hemorrhage. Horses almost never release reticu- secondary to chronic gastrointestinal blood loss. A microcytic hypochro-
locytes in response to anemia; therefore no polychromasia is present. mic anemia with poikilocytosis (including keratocytes, schistocytes, and
Wright-Giemsa stain. dacryocytes) were present.
produced. Consequently increased anisocytosis is usually Hypochromasia

present in regenerative anemia (see Figs. 4-18, 4-19, 4-21), The presence of erythrocytes with decreased hemoglobin con-
but it may be present in some cases of nonregenerative anemia centration and increased central pallor is called hypochroma-
resulting from dyserythropoiesis.199 Anisocytosis without sia (Figs. 4-26 through 4-32). Not only is the center of the
polychromasia may be seen in horses with intensely regenera- cell paler than normal but the diameter of the area of central
tive anemia (Fig. 4-25). pallor is increased relative to the red-staining periphery of the

Marked poikilocytosis and hypochromasia in blood from a 6-week-old Marked poikilocytosis and hypochromasia in blood from a piglet with
lamb with microcytic hypochromic iron-deficiency anemia. Wright- microcytic hypochromic iron-deficiency anemia resulting from a failure
Giemsa stain. to provide iron injections that are part of the husbandry required in
raising piglets on slated floors. Wright-Giemsa stain.

Marked poikilocytosis (primarily dacryocytes) and hypochromasia in Microcytic erythrocytes in blood from an iron-deficient llama exhibiting
blood from a goat with microcytic hypochromic iron-deficiency anemia irregular or eccentric areas of hypochromasia within the cells. Wright-
secondary to chronic blood loss resulting from Haemonchus gastrointes- Giemsa stain.
tinal parasites. Wright-Giemsa stain.
cell. True hypochromic erythrocytes must be differentiated the cells are thin leptocytes (Fig. 4-34, A). Because these micro-
from torocytes, which have colorless punched-out centers cytic leptocytes have increased diameter-to-volume ratios, they
but wider dense red-staining peripheries (Fig. 4-33).43,238 may not appear as small cells when viewed in stained blood
Torocytes are generally artifacts. Increased hypochromasia is films (Fig. 4-34, B).210 Microcytic erythrocytes from iron-
observed in iron-deficiency anemia. deficient llamas and alpacas exhibit irregular or eccentric areas
Erythrocytes from dogs, ruminants, and pigs with iron- of hypochromasia within the cells (see Figs. 4-31, 4-32).326
deficiency anemia often appear hypochromic on stained blood
smears. Hypochromasia is less prominent (see Fig. 4-24) and Poikilocytosis
generally not recognized in iron-deficient cats and horses. Erythrocytes can assume a wide variety of shapes. Poikilocytosis
Hypochromasia in iron deficiency results from both decreased is a general term used to describe the presence of abnormally
hemoglobin concentration within cells and from the fact that shaped erythrocytes. Although specific terminology is used
FIGURE 4-32
Blood from an iron-deficient alpaca with dacryocytes and spindle-shaped
hypochromic erythrocytes. Wright-Giemsa stain.
FIGURE 4 -34
Comparison of discocytes to leptocytes. A, The profile of a normal dis-
cocyte is compared to a leptocyte with similar diameter but smaller
volume. B, Blood from a dog with a microcytic hypochromic (MCV =
32 fL, MCHC = 23 g/dL) iron-deficiency anemia was mixed with an
equal volume of blood from a normal dog (MCV = 70 fL, MCHC =
FIGURE 4-33 34 g/dL) prior to blood film preparation. Because the hypochromic cells
Two torocytes (center and bottom left) with colorless punched-out centers are leptocytes, they have diameters similar to normal cells even though
and wide dense red-staining peripheries in blood from a dog. Wright- they are microcytic cells. Wright-Giemsa stain.
Giemsa stain.
From Harvey JW, French TW, Meyer DJ. Chronic iron deficiency anemia in
dogs. J Am Anim Hosp Assoc. 1982;18:946-960.
for certain abnormal shapes, it is less important to quantify
each type of shape change than it is to determine the cause
of the shape change.509 Poikilocytosis may be present in clini- in dogs, pigs, and ruminants may exhibit pronounced
cally normal goats and young cattle (Figs. 4-5, 4-35).225,409 In poikilocytosis (see Figs. 4-27 through 4-32).201 Poikilocytes
some instances, these shapes appear to be related to the hemo- can form when oxidant injury results in Heinz body forma-
globin types present, but an abnormality in protein 4.2 in the tion and/or membrane injury. One or more blunt erythrocyte
membrane has been suggested as a causative factor in calves.352 surface projections may form as the membrane adheres to
Increased numbers of poikilocytes are also present in human Heinz bodies bound to its internal surface.392 Various abnor-
preterm and term neonates.403 malities in erythrocyte shape were reported in horses with
Poikilocytosis forms in various disorders associated with intravascular hemolysis resulting from severe cutaneous burn
erythrocyte fragmentation, including disseminated intra injuries. Membrane blebbing with fragmentation was promi-
vascular coagulation (DIC), liver disease, myeloid neoplasms, nent (Fig. 4-36).347 A variety of abnormal erythrocyte shapes
myelofibrosis, glomerulonephritis, and hemangiosarcoma have been reported in dogs and cats with doxorubicin toxic-
(dogs).392 For unknown reasons, severe iron-deficiency anemia ity32,348 and in dogs with dyserythropoiesis.223
A B
FIGURE 4 -37
Echinocytes in blood from a dog with histiocytic sarcoma. A, Spicules
of similar length are regularly spaced. B, Echinocytes, in a thinner area
of the same blood film, appear as erythrocytes with scalloped borders.
FIGURE 4-35
Poikilocytosis (acanthocytes and echinocytes) in blood from a nonanemic
young calf. Wright-Giemsa stain.
FIGURE 4 -38
Echinocytes in blood from a normal pig appear as erythrocytes with
scalloped borders; consequently the term crenation, from Latin meaning
“notched,” has previously been used for echinocytes. Wright-Giemsa
FIGURE 4-36 stain.
Echinocytes with membrane blebbing and fragmentation in blood from
a horse with intravascular hemolysis resulting from severe cutaneous burn
injuries. Wright stain.
(Fig. 4-38), forming in vitro.238 The morphology of echino-
Photograph of a stained blood film from a 2004 ASVCP slide review case
submitted by E. Spangler, M. Johnson, A. Kessell, B. Weeks, T. Norman, and
cytes varies from slightly spiculated echinodiscocytes to highly
K. Chaffin. spiculated spheroechinocytes, which have been called burr
cells (Fig. 4-39, A). The most advanced echinocytes are those
that have lost most of their spicules and have nearly become
spherocytes (Fig. 4-39, B). Echinocytes form when the surface
Echinocytes (Crenated Erythrocytes) area of the outer lipid monolayer increases relative to the inner
Echinocytes are spiculated erythrocytes in which the spicules monolayer.438
are relatively evenly spaced and of similar size.523 Spicules may Echinocytic transformation occurs in the presence of fatty
be sharp or blunt. When observed in stained blood films, acids, lysophospholipids, and amphiphatic drugs that distri
echinocytosis is usually an artifact that results from excess bute preferentially in the outer half of the lipid bilayer.205
EDTA, improper smear preparation, or prolonged sample Transient echinocytosis occurs in horses with Clostridium per-
storage before blood film preparation. The appearance of the fringens infection526 and in dogs following rattlesnake (see Fig.
echinocytes can vary depending on the thickness of the blood 4-39, A),70,191 coral snake (see Fig. 4-39, B),301 water moccasin
film (Fig. 4-37). They are common in normal pig blood smears (Fig. 4-40), and asp viper (Vipera aspis) envenomation,303
A B
FIGURE 4-39
Echinocytes in the blood of dogs following snake bites. A, Highly spicu-
lated echinocytes (burr cells) in blood from a dog following an Eastern
diamondback rattlesnake bite. B, Spheroechinocytes and a lysed eryth-
rocyte “ghost” (bottom center) in blood from a dog following a coral snake
bite. Wright-Giemsa stain. FIGURE 4 -41
An echinocyte (center) and four polychromatophilic erythrocytes in
blood from a pyruvate kinase-deficient beagle dog. Wright-Giemsa stain.
Although the mechanisms are not fully understood, ATP is

required for the maintenance of normal shape and deform-
ability of erythrocytes.205 Phospholipids are asymmetrically
arranged in the plasma membrane, with anionic amino-
containing phospholipids (predominantly phosphatidylserine)
located in the inner leaflet of the bilaminar membrane.
These anionic phospholipids are shuttled (flipped) from the
outer leaflet to the inner leaflet by an ATP-dependent
aminophospholipid-specific translocase or flippase.553 ATP
also provides the energy needed to pump Ca+2 out of cells.
Increased Ca+2 activates neutral proteases (calpains), which can
degrade membrane skeletal proteins; phospholipase C, which
cleaves phosphoinositides; and scramblase, which accelerates
FIGURE 4-40 the bidirectional transbilayer movement of phospholipids.205,540
Highly spiculated echinocytes in blood from a dog following a water The inhibition of the flippase and/or the activation of the
moccasin bite. Two large polychromatophilic erythrocytes (reticulocytes) scramblase can result in increased phosphatidylserine in the
are also present. Wright-Giemsa stain.
outer layer, which appears to promote echinocyte formation, as
well as coagulation and erythrophagocytosis.116,295,406
presumably secondary to the action of phospholipases present Echinocytes and other shape abnormalities have been rec-
in venom.498 Depending on the time course and dose of venom ognized in dogs with erythrocyte pyruvate kinase deficiency
received, either echinocytosis or spherocytosis may be observed (Fig. 4-41), which results in a decreased ability to generate
after these snakebites. ATP. 86,328,412 Occasional echinocytes may also be seen in dogs
Echinocytes may occur in uremic animals and immediately with erythrocyte phosphofructokinase deficiency (Fig. 4-42).
after transfusion of stored blood.392 They have been seen with PFK-deficient erythrocytes have decreased ATP and increased
increased frequency in dogs with glomerulonephritis and neo- Ca+2 concentrations.404
plasia (lymphoma, hemangiosarcoma, mast cell tumor, and
carcinoma).391,523 Echinocytes are the predominant erythro- Acanthocytes
cyte shape abnormality in human burn patients, and this was Erythrocytes with irregularly spaced, variably sized
one of the shape abnormalities recognized in horses with spicules are called acanthocytes or spur cells (Fig. 4-43).43
severe cutaneous burn injuries (see Fig. 4-36).195,347 Acanthocytes form when erythrocyte membranes contain
Echinocytes also form when erythrocytes are dehydrated, excess cholesterol compared to phospholipids. If cholesterol
pH is increased, ATP is depleted, and intracellular calcium is and phospholipids are increased to a similar degree, codocyte
increased.31,205 Echinocytosis occurs in horses when total body formation is more likely than acanthocyte formation.98
depletion of cations has occurred (endurance exercise, furose- Alterations in erythrocyte membrane lipids can result from
mide treatment, diarrhea, systemic disease).162,205 increased blood cholesterol content or due to the presence of

An echinocyte (center) and four polychromatophilic erythrocytes in Acanthocytes in blood from a dog with neoplastic lymphoid infiltrates
blood from a phosphofructokinase-deficient English springer spaniel in the liver. A neoplastic lymphoblast is also present. Wright-Giemsa
dog. Wright-Giemsa stain. stain.
A B
FIGURE 4-43
Acanthocytes in blood from a dog and cat. A, Two acanthocytes (above)
with irregularly spaced, variably sized spicules in blood from a dog with
neoplastic lymphoid infiltrates in the liver. B, An acanthocyte in blood
of a cat with hepatic lipidosis. Wright-Giemsa stain.
FIGURE 4 -45
Three acanthocytes, four polychromatophilic erythrocytes, two metaru-
abnormal plasma lipoprotein composition.99 Acanthocytes bricytes, and a neutrophil in blood from a dog with hemangiosarcoma.
have been recognized in animals with liver disease (Fig. 4-44),
possibly due to alterations in plasma lipid composition, which
can alter erythrocyte lipid composition.91,392,522 They have also
been reported in dogs with disorders that result in erythrocyte the hole results in the formation of one or two projections.
fragmentation, such as hemangiosarcoma (Fig. 4-45), DIC, Keratocytes have been recognized in various disorders
and glomerulonephritis.342,471,522 including iron-deficiency anemia,201 liver disorders,91 doxoru-
Marked acanthocytosis is reported to occur in young goats bicin toxicity in cats,348 myelodysplastic syndrome,524 and in
and some young cattle (see Fig. 4-35). Acanthocytosis of various disorders in dogs having concomitant echinocytosis
young goats occurs as a result of the presence of hemoglobin or acanthocytosis.342,522,523
C at this early stage of development.205
Stomatocytes
Keratocytes Cup-shaped erythrocytes that have oval or elongated areas of
Erythrocytes containing one or more intact holes are called central pallor when viewed in stained blood films are called
prekeratocytes, and erythrocytes with ruptured holes are stomatocytes (Fig. 4-48, A). They most often occur as artifacts
called keratocytes (Figs. 4-46 and 4-47).140 The rupture of in thick blood film preparations. Stomatocytes form when pH
A B A B

Scanning electron photomicrographs of cat keratocytes. A, A prekerato- Stomatocytes with elongated areas of central pallor in blood. A, Blood
cyte contains a hole completely through the cell. B, A keratocyte forms from a cat with hemolytic anemia. The stomatocytes were not uniformly
when the hole in the prekeratocyte ruptures, resulting in the formation present in the blood film and were considered to be an artifact. Wright-
of one or two projections. This cell might be considered an echinokera- Giemsa stain. B, Stomatocytes in blood from an asymptomatic Pomera-
tocyte because of its additional projections. nian dog with persistent stomatocytosis associated with macrocytic
hypochromic erythrocytes. As in dogs reported to have hereditary sto-
Photomicrographs were provided by C. L. Flint and M. A. Scott. matocytosis, these erythrocytes were osmotically fragile and had a low
concentration of reduced glutathione. Wright-Giemsa stain.
A B
FIGURE 4 -49
Scanning electron photomicrograph of a discocyte (left) and spherocyte
(right) in blood from a dog.
C D Courtesy of K. S. Keeton and N. C. Jain.
FIGURE 4-47
Keratocytes in blood from a cat with hepatic lipidosis. A, Prekeratocyte Pomeranians (Fig. 4-48, B) with hereditary stomatocytosis.
with a hole. B, Keratocyte where the hole has ruptured, producing a
single horn. C, Keratocyte where hole has ruptured, producing two horns.
Chondrodysplasia (short-limbed dwarfism) occurs along with
D, Keratocyte where two holes have ruptured. Wright-Giemsa stain. stomatocytosis in Alaskan malamutes.139 Stomatocytosis in
Drentse patrijshond dogs is associated with severe polysys-
temic disease and shortened life span.433
is decreased160 and amphiphatic drugs are present that distrib- Spherocytes

ute preferentially in the inner leaflet of the lipid bilayer.205 Spherical erythrocytes that result from cell swelling and/or
Stomatocytes also form when the water content of eryth- loss of cell membrane are referred to as spherocytes (Fig.
rocytes is increased, as occurs in at least three different 4-49). Spherocytes lack central pallor and have smaller diam-
inherited syndromes in dogs.205 No clinical signs occur in eters than normal on stained blood films (Fig. 4-50, A).
miniature schnauzers,71,170 standard schnauzers,59,355 or Spherical erythrocytes with slight indentations on one side
A B
FIGURE 4 -51
Spherocytes, polychromatophilic erythrocytes, and two rubricytes in
blood from a dog with primary immune-mediated hemolytic anemia.
C D
FIGURE 4-50
Spherocytes and stomatospherocytes in animal blood samples. A, Three
spherocytes (bottom) and a large polychromatophilic erythrocyte or retic-
ulocyte (top) in blood from a dog with immune-mediated hemolytic
anemia. B, Two stomatospherocytes in blood from a dog with primary
immune-mediated hemolytic anemia. The stomatospherocytes are not
perfect spheres; rather, each has a slight indentation on one side. C, Two
stomatospherocytes in blood from a cat with Mycoplasma haemofelis
infection. D, A spherocyte (top) and a discocyte (bottom) in blood from
a foal with immune-mediated neonatal isoerythrolysis. Wright-Giemsa
stain.
A B
FIGURE 4-52
may be called stomatospherocytes (Fig. 4-50, B,C). Since
Schistocytes in dog blood. A, A fragmented erythrocyte (schistocyte)
erythrocytes from other common domestic animals exhibit and two discocytes in blood from a dog with disseminated intravascular
less central pallor than those of dogs, it is difficult to be certain coagulation (DIC). B, A schistocyte (left), discocyte (top), and echinocyte
when spherocytes are present in these noncanine species (bottom) in blood from a dog with DIC. Wright-Giemsa stain.
(Fig. 4-50, D).
Spherocytes and stomatospherocytes occur most frequently
in association with immune-mediated hemolytic anemia in
dogs (Fig. 4-51).33,368 Other potential causes of spherocyte Schistocytes
formation include snake envenomations,301,303,498 bee Erythrocyte fragments with pointed extremities are called
stings,344,543 zinc toxicity,64 erythrocyte parasites,3 transfusion schistocytes or schizocytes (Fig. 4-52), and they are smaller
of stored blood, and a familial dyserythropoiesis in dogs.223 than normal discocytes. Schistocytes may be seen in dogs with
Spherocytes have been reported in cattle with anaplasmosis463 microangiopathic hemolytic anemia associated with DIC,
and Theileria buffeli infection453 and in horses with cutaneous where they can be formed by the impact of erythrocytes
burns.347 with fibrin strands in flowing blood (Fig. 4-53).392 Schisto-
Defects in ankyrin, band 3, protein 4.2, and certain cytes are less consistently seen in cats, horses, and cattle with
defects in α-spectrin and β-spectrin result in hereditary DIC,234,477 possibly because the erythrocytes of these species
spherocytosis in mice and humans.116 A complete absence of are smaller and less likely to be split by fibrin strands in the
band 3 results in hereditary spherocytosis in Japanese black circulation.
cattle.34,232 Hereditary spherocytosis has been reported in Schistocytes have also been seen in severe iron-deficiency
golden retriever dogs with reductions in erythrocyte mem- anemia (see Fig. 4-27),201 myelofibrosis,128,281,420 heart failure,
brane spectrin432; however, spherocytes were not recognized glomerulonephritis, hemolytic uremic syndrome, hemo
in stained blood films.434 phagocytic histiocytic disorders, hemangiosarcoma in dogs,

Drawing of schistocyte and microspherocyte formation when an eryth- Echinocytes, acanthocytes, and echinoschistocytes in blood from a
rocyte impacts a fibrin strand under flow conditions. splenectomized pyruvate kinase-deficient Cairn terrier dog. Wright-
Giemsa stain.
Leptocytes
These cells are thin, often hypochromic-appearing erythro-
cytes with increased membrane-to-volume ratios. Some lep-
tocytes appear folded (Fig. 4-56, A), some appear as triconcave
knizocytes (Fig. 4-57) that give the impression that the eryth-
rocyte has a central bar of hemoglobin (Fig. 4-56, B,C), and
others appear as codocytes (Fig. 4-58). Codocytes (target cells)
are bell-shaped cells that exhibit a central density or “bull’s-
eye” in stained blood films. Small numbers of codocytes are
often seen in normal dog blood, and both codocytes and
knizocytes are increased in regenerative anemia in dogs.
Codocytes are especially increased in dogs with a congenital
dyserythropoiesis.223 Leptocytes may be seen in iron-deficiency
anemia (Fig. 4-58, B)201 and rarely in hepatic insufficiency
(Fig. 4-58, C), resulting in a balanced accumulation of mem-
brane phospholipids and cholesterol.98 Polychromatophilic
FIGURE 4-54
erythrocytes can sometimes appear as leptocytes.
Echinocytes, a polychromatophilic erythrocyte, a metarubricyte and a
schistocyte (bottom) in blood from a dog with intravascular hemolysis, Eccentrocytes
resulting from dirofilariasis and caudal vena cava syndrome. A neutrophil
and band neutrophil are also present. Wright-Giemsa stain. An erythrocyte in which the hemoglobin is localized to part
of the cell, leaving a hemoglobin-poor area visible in the
remaining part of the cell, is termed an eccentrocyte (Figs.
caudal vena cava syndrome of dirofilariasis in dogs (Fig. 4-54), 4-59, 4-60). Other terms used to refer to eccentrocytes include
and congenital and acquired dyserythropoiesis in dogs.* hemighosts, irregularly contracted cells, double-colored cells, and
Marked poikilocytosis with schistocytes and acanthocytes cross-bonded erythrocytes.24 Oxidant damage to erythrocyte
has been recognized in pyruvate kinase-deficient dogs after membranes results in the adhesion of opposing areas of the
splenectomy (Fig. 4-55).378,412 It is assumed that the spleen cytoplasmic face of the erythrocyte membrane.24,137 It has
had previously removed these fragmented erythrocytes. Schis- been suggested that eccentrocytes form in vivo when adhesive,
tocytes have been described in cats with liver disease, in cats damaged membranes are brought together by osmotic
and dogs with doxorubicin toxicity,91,348 and in horses with shrinkage in the kidney and/or from squeezing through the
severe cutaneous burns.347 microcirculation.138
Eccentrocytes have been recognized in dogs secondary to
increased endogenous oxidants associated with ketoacidotic
*References 193, 197, 223, 281, 392, 522, 524. diabetes, inflammation (Fig. 4-61), neoplasia (especially
A B C
FIGURE 4-56
Leptocytes in blood from dogs. A, Two thin, flat, hypochromic-appearing erythrocytes (leptocytes) with
increased membrane-to-volume ratios are present in blood from a dog with severe iron-deficiency anemia.
The bottom leptocyte is folded. B, A discocyte and triconcave knizocyte in blood from a dog with a mild
nonregenerative anemia associated with a hepatocellular carcinoma. C, Leptocytes, including two knizocytes
(top and bottom center), are present in blood of a dog with iron-deficiency anemia. Wright-Giemsa stain.
Eccentrocytosis was recognized in a cow treated with

intravenous hydrogen peroxide as a “home remedy.”291 Eccen-
trocytes have been reported in horses ingesting red maple
leaves and in those with severe cutaneous burns.347,389 Finally,
eccentrocytes have been seen in horses with inherited G6PD
deficiency (see Fig. 4-59, B) and glutathione reductase
deficiency secondary to erythrocyte FAD deficiency (Fig.
4-64).212,452 Both of these hereditary disorders result in the
decreased ability of erythrocytes to protect against endoge-
nous oxidants.
Pyknocytes
Irregularly spherical erythrocytes with small cytoplasmic pro-
jections are called pyknocytes (see Figs. 4-59, D, 4-60, D).
Like eccentrocytes, pyknocytes appear to be a product of
oxidant injury.131,452 They may develop from eccentrocytes fol-
FIGURE 4-57 lowing the loss of much of the fused membrane.212 Pyknocytes
Scanning electron photomicrograph of two discocytes and a knizocyte are appreciably smaller in diameter and stain more densely
(top) in blood from a dog. (especially with new methylene blue) than discocytes (Fig.
4-64).203 These contracted erythrocytes have higher hemoglo-
Courtesy of K. S. Keeton and N. C. Jain.
bin concentrations than normal131 and presumably represent
erythrocytes undergoing eryptosis.
Elliptocytes (Ovalocytes)
lymphoma), and Babesia canis canis infection.77,82 Eccentro- Erythrocytes from nonmammals and animals in the Cameli-
cytes have been seen in dogs ingesting or receiving oxidants dae family are elliptical or oval in shape (see Fig. 4-6). They
including onions and garlic, acetaminophen (see Fig. 4-59, C) are generally flat rather than biconcave. Abnormal elliptocytes
and nonsteroidal anti-inflammatory drugs, vitamin K and have been recognized in cats with bone marrow abnormalities
vitamin K antagonist rodenticides, naphthalene, and pro- (myeloid neoplasms and acute lymphoblastic leukemia),281
longed propofol anesthesia (Fig. 4-62).77,120,200,276 Eccentrocyte hepatic lipidosis,91 portosystemic shunts,410 and doxorubicin
formation also occurs in cats following oxidant damage toxicity348 and in dogs with myelofibrosis,222,420 myelodys
(Fig. 4-63). plastic syndrome,524 and glomerulonephritis, in which the
A B C
FIGURE 4-58
Codocytes in blood from dogs. These erythrocytes exhibit a central density or “bull’s-eye” and are often referred
to as target cells. A, Three codocytes in blood from a Cairn terrier dog with a regenerative anemia and hepatic
hemochromatosis secondary to pyruvate kinase deficiency. B, Two codocytes (top and bottom center) and a
schistocyte (bottom left) in blood from a dog with severe iron-deficiency anemia. C, Codocytes in blood from
a dog with liver disease. Wright-Giemsa stain.
A B
A B
C D
FIGURE 4-60
Scanning and transmission electron photomicrographs of eccentrocytes
and a pyknocyte in blood from a horse with erythrocyte flavin adenine
C D
dinucleotide (FAD) deficiency. A, Scanning electron photomicrograph
of an eccentrocyte showing a fused membrane leaf in the left portion
FIGURE 4-59 and a spheroid region containing hemoglobin in the right portion of an
Eccentrocytes in blood from dogs and a horse. A, An eccentrocyte erythrocyte. B, Scanning electron photomicrograph of an eccentrocyte
(center) in blood from a dog with prolonged propofol anesthesia. B, Two showing fused membranes in different planes. C, Transmission electron
eccentrocytes in blood from a horse with inherited erythrocyte glucose- photomicrograph of an eccentrocyte showing a fused membrane leaf in
6-phosphate dehydrogenase (G6PD) deficiency. C, Two eccentrocytes the left portion and a spheroid region containing hemoglobin in the right
and a discocyte (left) in blood from a dog with oxidant injury induced portion of an erythrocyte. D, Scanning electron photomicrograph of a
by the administration of acetaminophen. The cell at top center appears spheroid pyknocyte with only tags of cytoplasm remaining.
spherical with a small tag of cytoplasm and is classified as a pyknocyte.
D, Two echinocytes (top), an eccentrocyte (bottom left) and a pyknocyte From Harvey JW, Stockham SL, Scott MA, et al. Methemoglobinemia and
(bottom right) in blood from a dog with diabetes mellitus and septic eccentrocytosis in equine erythrocyte flavin adenine dinucleotide deficiency. Vet
hepatitis. Wright-Giemsa stain. Pathol. 2003;40:632-642.
FIGURE 4-61
Blood from a dog with diabetes mellitus and septic hepatitis. Four eccen-
trocytes, two pyknocytes, and a toxic band neutrophil are present.
FIGURE 4 -63
Heinz body and eccentrocyte formation in blood from a cat with acet-
aminophen toxicity. A, Heinz bodies are barely discernible as pale spots
in erythrocytes. A polychromatophilic erythrocyte (left lower quadrant)
and an erythrocyte containing a Howell-Jolly body (top right) are also
present. B, Eccentrocytes and pale-staining Heinz bodies are visible in
blood collected 2 days after the sample shown in (A).
FIGURE 4-62
Eight eccentrocytes in blood from a dog subjected to prolonged propofol myeloid neoplasms,238 dogs with glomerulonephritis, and a dog
anesthesia. A neutrophil is present at bottom right. Wright-Giemsa stain.
with hypersplenism.267 Dacryocytes are common erythrocyte
shape abnormalities in iron-deficient ruminants, including
llamas and alpacas (Fig. 4-67, C,D).326,483
elliptocytes may be spiculated (Fig. 4-65).391 Hereditary Drepanocytes (Sickle Cells)

elliptocytosis has been reported in a dog with a membrane These fusiform or spindle-shaped erythrocytes were first rec-
protein 4.1 deficiency and in another with a mutant β-spectrin ognized in deer blood in 1840 and in a human with sickle cell
(Fig. 4-66).122,439 anemia in 1910.44 Drepanocytes are often observed in blood
from normal deer (Figs. 4-68, A, 4-69).474 They develop sec-
Dacryocytes ondary to hemoglobin polymerization, and drepanocyte shape
These erythrocytes are teardrop-shaped with single elongated or in deer depends on the hemoglobin types present. This is an
pointed extremities (Fig. 4-67, A,B). Dacryocytosis is a common in vitro phenomenon that occurs when oxygen tension is high
feature of myelofibrosis in humans, but dacryocytes are not as and pH is above 7.4. Members of the Cervidae family whose
commonly recognized in dogs with myelofibrosis.281,388,420 Dac- erythrocytes do not sickle in vitro include reindeer, caribou,
ryocytes have also been seen in the blood of dogs and cats with and muntjac deer.479
A B
FIGURE 4 -65
Elliptocytes in blood from a cat and a dog. A, Three elliptocytes (top)
and a discocyte in blood from a diabetic cat with mild anemia. Radio-
A graphs revealed diffuse interstitial lung disease of unknown etiology.
B, A discocyte (left) and an echinoelliptocyte (right) in blood from a dog
with glomerulonephritis. Wright-Giemsa stain.
FIGURE 4-64
Pyknocytes and eccentrocytes in blood from horses with erythrocyte A
flavin adenine dinucleotide (FAD) deficiency. A, Two eccentrocytes
(center), three pyknocytes (upper left quadrant), and an erythrocyte con-
taining an eccentrically located hemoglobin crystal (bottom) in a stained-
blood film from a mustang mare with FAD deficiency. Wright-Giemsa.
B, Dark blue-staining eccentrocytes and pyknocytes in a blood film from
a Kentucky mountain saddle horse gelding with FAD deficiency. New
methylene blue reticulocyte stain.
From Harvey JW. Pathogenesis, laboratory diagnosis, and clinical implications

of erythrocyte enzyme deficiencies in dogs, cats, and horses. Vet Clin Pathol.
2006;35:144-156.
FIGURE 4 -66
Elliptocytosis in blood from a dog with a mutant β-spectrin. A, Ellip-
tocytes and echinoelliptocytes in blood from a dog. Wright-Giemsa
stain. B, Scanning electron photomicrograph of elliptocytes and echino-
elliptocytes in blood from a dog.
From Di Terlizzi R, Gallagher PG, Mohandas N, et al. Canine elliptocytosis

due to a mutant beta-spectrin. Vet Clin Pathol. 2009;38:52-58.
A B
FIGURE 4 -69
Transmission electron photomicrograph of a matchstick-shaped drepa-
nocyte in blood from a deer showing filamentous aggregates of
hemoglobin.
C D
From Taylor WJ, Simpson CF. Ultrastructure of sickled deer erythrocytes. II.
FIGURE 4-67 The matchstick cell. Blood. 1974;43:907-914.
Dacryocytes in the blood of animals. A, A dacryocyte (bottom) and two
discocytes in blood from a cat. Wright-Giemsa stain. B, A dacryocyte
(top) and elliptocyte (bottom) in blood from a dog with glomerulo
nephritis. C, Hypochromic dacryocytes in blood from a goat with severe
iron-deficiency anemia. D, Hypochromic dacryocytes in blood from a depending on the individual goat and on in vitro alterations
llama with severe iron-deficiency anemia. The presence of the normal in temperature, pH, and oxygenation. The number of these
llama elliptocyte (above right) is the result of a blood transfusion. Wright-
Giemsa stain.
cells decreases during anemia, probably because of the synthe-
sis of hemoglobin C.239
Crystallized Hemoglobin
The presence of large hemoglobin crystals within erythrocytes
is occasionally recognized in blood films from cats (Figs. 4-70,
A, 4-71, A)16,17,428 and dogs (Fig. 4-70, B)287 and frequently in
the blood of llamas and alpacas (Fig. 4-70, C).483 No hemo-
globin abnormalities have been recognized by hemoglobin
electrophoresis in animals with hemoglobin crystals.
In addition to eccentrocytes and pyknocytes, intraerythro-
cytic crystals have been recognized in low number in horses
A B with erythrocyte G6PD and FAD deficiencies (Figs. 4-70, D,
4-71, B).212 Large hemoglobin crystals were reported in horses
FIGURE 4-68 with experimental Babesia equi infections treated with imido-
Drepanocytes and fusiform erythrocytes in a deer and goat. A, Elongated
carb dipropionate (Fig. 4-71, C).429 Low numbers of intraeryth-
drepanocyte (sickle cell) in blood from a white-tailed deer. B, Erythro- rocytic hemoglobin crystals have been reported in two dogs
cytes containing hemoglobin inclusions in blood from a mixed-breed with multiple erythroid abnormalities including prominent
goat. Some erythrocytes appeared as rectangles but most appeared more siderotic inclusions in their erythrocytes (Fig. 4-71, D).81,209
fusiform and may represent polymerization of hemoglobin in tubular Hemoglobin crystals may also develop as sample storage
filaments, as occurs in drepanocytes. Wright-Giemsa stain.
artifacts.38
The mechanism or mechanisms of hemoglobin crystal for-
Polymerization of hemoglobin in tubular filaments occurs mation are unknown, but based on studies of humans with
in some normal adult Angora goats239,240 and some breeds of hemoglobin C disease, crystal formation may be influenced
British sheep.130 The resultant fusiform or spindle-shaped by factors such as pH, degree of oxygen saturation, and cellular
erythrocytes resemble drepanocytes in deer; they have been dehydration/increased intracellular hemoglobin concentra-
called acuminocytes by some authors (Fig. 4-68, B).238 tion.76 Their potential relationships to oxidant injury, eccen-
The proportion of fusiform cells in Angora goats varies trocytes, and eryptosis need further study.212
A B A B
C D C D

Hemoglobin crystals in stained blood films. A, Crystallized hemoglobin Scanning and transmission electron photomicrographs of hemoglobin
in an erythrocyte from a cat. B, Discocyte and erythrocyte with crystal- crystals. A, Scanning electron photomicrograph of a cat erythrocyte
lized hemoglobin in the blood from a dog with a mild nonregenerative containing crystallized hemoglobin. B, Scanning electron photomicro-
anemia associated with a hepatocellular carcinoma. C, Crystallized graph of an erythrocyte containing a hemoglobin crystal in blood from
hemoglobin in two erythrocytes from a llama. D, An eccentrocyte a mustang mare with FAD deficiency. C, Transmission electron photo-
(bottom) and hemoglobin crystal in blood from mustang mare with FAD micrograph of a horse erythrocyte containing both a hemoglobin crystal
deficiency. Wright-Giemsa stain. and a Babesia equi organism. The horse had been treated with imidocarb.
D, Transmission electron photomicrograph of a dog erythrocyte contain-
ing a hemoglobin crystal. The dog had microcytic hypochromic erythro-
cytes that also contained siderotic inclusions and Heinz bodies.
A, From Jain NC. Schalm’s Veterinary Hematology. 4th ed. Philadelphia,

Lysed Erythrocytes Lea & Febiger, 1986. B, From Harvey JW, Stockham SL, Scott MA, et al.
Methemoglobinemia and eccentrocytosis in equine erythrocyte flavin adenine
The presence of erythrocyte “ghosts” in peripheral blood films dinucleotide deficiency. Vet Pathol. 2003;40:632-642. C, From Simpson CF,
indicates that the cells lysed prior to blood film preparation Taylor WJ, Kitchen H. Crystalline inclusions in erythrocytes parasitized with
(Fig. 4-72, A). Erythrocyte membranes are rapidly cleared from Babesia equi following treatment of ponies with imidocarb. Am J Vet Res.
the circulation following intravascular hemolysis; consequently 1980;41:1336-1340. D,Courtesy of W. L. Clapp.
the presence of erythrocyte ghosts indicates either recent intra-
vascular hemolysis or in vitro hemolysis in the blood tube after
collection. If the hemolysis is caused by an oxidant, Heinz
bodies may be visible within erythrocyte ghosts (Fig. 4-72, B). Erythrocyte Vesicles
When erythrocytes lyse during blood film preparation, they Erythrocyte injury can result in the formation of variably
appear as red smudges (see Figs. 4-24, 4-72, C). These smudged sized erythrocyte vesicles (Fig. 4-74, A–C). They are classified
erythrocytes are commonly seen in lipemic samples. as microvesicles when their diameter is less than 1 µm. The
small size of microvesicles makes them difficult to identify
in stained blood films and difficult to count using flow
Erythroid Loops cytometers.
In addition to echinocytes, spherocytes, and erythrocyte Erythrocytes, leukocytes, platelets, and endothelial cells all
ghosts, moderate numbers of unusual erythrocyte membrane- release membrane-bound microvesicles (also called micro
like structures (termed erythroid loops) have been found in particles) into the circulation, even in healthy individuals.
the blood of dogs bitten by the asp viper (Fig. 4-73).303 Depending on the cell type and stimulus, these microvesicles
Although generally round, these loops were sometimes dis- may be either exosomes or ectosomes. Exosomes are micro
rupted and appeared as thin, pale reddish-blue bands. They vesicles that form following inward budding of membranes
are believed to be a consequence of erythrocyte hemolysis, but to form multivesicular bodies, and multivesicular bodies sub-
the mechanism of their formation is unknown. sequently fuse with the external cell membrane, releasing
A B C
FIGURE 4-72
Lysed erythrocytes in blood films. A, Red-staining intact erythrocytes (echinocyte in the center) and
pale-staining erythrocyte ghosts in blood from a horse in which intravascular hemolysis was produced by
the intravenous and intraperitoneal administration of hypotonic fluid believed isotonic at the time of
administration. B, Erythrocyte ghosts, each containing a single red-staining Heinz body, in erythrocytes
from a cat with intravascular hemolysis caused by acetaminophen administration. C, A lysed erythrocyte (red
smudge at top) and discocyte in blood from a dog with lipemia. The lysis occurred during smear preparation.
A B
FIGURE 4-73
Three erythroid loops in blood from a dog bitten by a snake (Vipera
aspis).
From Masserdotti C. Unusual “erythroid loops” in canine blood smears after C D

viper-bite envenomation. Vet Clin Pathol. 2009;38:321-325.
FIGURE 4 -74
Erythrocyte vesicles in blood films. A, Erythrocyte vesicle (center) in
blood from a febrile dog with mild nonregenerative anemia and lym-
phoma. Low numbers of eccentrocytes and pyknoctyes and rare hemo-
globin crystals were also noted in the stained blood film. B, Erythrocyte
vesicle formation in blood from a cat with eccentrocytes and erythrocytes
containing Heinz bodies. The cat was a ketoacidotic diabetic with hepatic
lipidosis. C, Elongated erythrocyte vesicle (top left) and eccentrocyte
(bottom right) in blood from a dog subjected to prolonged propofol
anesthesia. D, Microvesiculation from an acanthocyte in blood from a
dog with neoplastic lymphoid infiltrates in the liver.
preformed exosomes.405 As presented in Chapter 3, as reticu- transfusion when ATP depletion results in echinocyte forma-
locytes mature into erythrocytes, this process is used by reticu- tion and blebbing of ectosomes from the tips of echinocytic
locytes to eliminate unneeded proteins, such as transferrin spicules.186
receptors and Na+, K+ ATPase transporters (in dogs) from Erythrocyte microvesicles are rapidly cleared by macro-
their surfaces.263 phages in the liver.537 The spleen also appears to be important
Ectosomes are shed by many cells, including erythrocytes, in this regard because erythrocyte microvesicles are higher in
following the budding of microvesicles directly from the cell splenectomized compared with nonsplenectomized humans
membrane (Fig. 4-74, D). The mean numbers of erythrocyte with immune-mediated thrombocytopenia.143
microvesicles in healthy humans are reported to vary from
about 30 to 300/µL of plasma, depending on the methods
used to measure them.40,335,536 Erythrocytic ectosomes have Nucleated Erythrocytes
both procoagulant and anti-inflammatory/immunosuppres- Rubricytes and metarubricytes (Fig. 4-75, A,B) are seldom
sive activities.405 They are released in increased numbers during present in the blood of normal adult mammals, although low
conditions that result in eryptosis (see Fig. 4-15), as discussed numbers may occur in some normal dogs and cats.490 Meta-
earlier in this chapter.270 Increased total microvesicles, includ- rubricytosis (generally called normoblastemia in human
ing erythrocyte microvesicles, are present in the blood of hematology) is often seen in blood in association with a
humans with metabolic syndrome and oxidative stress.216 regenerative anemia; however, their presence does not neces-
Large numbers of microvesicles can form in blood stored for sarily indicate that a regenerative response is present.296
FIGURE 4-75 Nucleated erythroid cells and nuclear fragmentation
A B C D
A, Polychromatophilic rubricyte in blood from a dog with regenerative anemia. B, Metarubricyte in blood
from a beagle with erythrocyte pyruvate kinase deficiency. C, Nuclear lobulation in a polychromatophilic
rubricyte in blood from a cat with acute myelogenous leukemia (AML-M6) and a nonregenerative anemia.
D, Exceptionally large basophilic rubricyte in blood from a cat with myelodysplastic syndrome and a nonre-
generative anemia.
E F G H
E, Rubriblast in blood from a cat with acute myelogenous leukemia (AML-M6) and a nonregenerative
anemia. F, Polychromatophilic metarubricyte with elongated nucleus and erythrocyte containing nuclear
fragments in blood from a dog with immune-mediated hemolytic anemia and thrombocytopenia 5 days after
treatment with vincristine. G, Erythrocyte containing nuclear fragments in blood from a dog with immune-
mediated hemolytic anemia and thrombocytopenia 5 days after treatment with vincristine. H, Mitotic rubri-
cyte in blood from a dog with hemangiosarcoma and a regenerative anemia. Wright-Giemsa stain.
Nucleated erythrocytes are rarely seen in horses with regen- when the nucleus is expelled (Fig. 4-77).43,76 They have gener-
erative anemia. ally been called Howell-Jolly bodies by hematologists and
Nucleated erythrocytes may be seen in animals with lead micronuclei by toxicologists.194,552 The latter term is advanta-
poisoning, in which there is minimal or no anemia,164,325 and geous because it describes their composition. Howell-Jolly
in nonanemic conditions in which the bone marrow is bodies are removed or “pitted” by the spleen as reticulocytes
damaged, such as septicemia, endotoxic shock, and drug and erythrocytes squeeze through interendothelial slits of the
administrations.117,490,534 Nucleated erythrocytes are present in splenic sinus.76,528 They may be present in low numbers in
the blood of most dogs presented with heat stroke.25 erythrocytes of normal horses and cats (Fig. 4-76, A). They are
Low numbers of nucleated erythrocytes are seen in a wide often present in association with regenerative anemia or
variety of conditions in dogs, including cardiovascular disease, following splenectomy in other species.76,384,494 Howell-Jolly
trauma, hyperadrenocorticism, and various inflammatory bodies may be increased in animals receiving glucocorticoid
conditions.296 therapy (Fig. 4-76, B),238 in benign poodle macrocytosis (Fig.
When frequent nucleated erythrocyte precursors are 4-76, C),238 and in animals being treated with chemothera-
present in the blood of an animal with nonregenerative anemia peutic agents that induce nuclear fragmentation, such as
(Fig. 4-75, C-E), conditions including myelodysplasia, hema- vincristine, colchicine, cytosine arabinoside, and cyclophos-
topoietic neoplasia,238,296 infiltrative marrow disease,217,490 phamide (see Fig. 4-75, F, G).194,552
impaired splenic function,490 and inherited dyserythropoietic
disorders223,446 should be considered. The presence of rubri- Heinz Bodies
blasts in blood from an animal with nonregenerative anemia These inclusions are large aggregates of oxidized, precipitated
strongly suggests that a myeloid neoplasm is present (Fig. hemoglobin attached to the internal surfaces of erythrocyte
4-75, E). Nucleated erythrocytes may be excessively large membranes. In contrast to Howell-Jolly bodies, which stain
(Fig. 4-75, D) and erythrocyte nuclei may be lobulated or dark blue, Heinz bodies stain red to pale pink with
fragmented in animals with myeloid neoplasms (Fig. 4-75, Romanowsky-type stains (Fig. 4-78, A,C). Heinz bodies stain
C)105,125,416 or following vincristine therapy (Fig. 4-75, F, G). lighter blue than Howell-Jolly bodies when stained with
Nucleated erythroid precursors are capable of division earlier reticulocyte stains (Fig. 4-78, B,D). They can also be visualized
than metarubricytes; consequently, mitotic nucleated erythro- as dark refractile inclusions in new methylene blue wet mount
cytes may be seen in blood (Fig. 4-75, H). preparations (Fig. 4-78, E). Heinz bodies may also be visible
within eccentrocytes (Fig. 4-78, F). If they bind extensively to
the inner surface of erythrocyte membranes, they may be
I N C LU S I O N S O F ERY T H RO C Y T E S recognized as small surface projections when the membrane
Howell-Jolly Bodies (Micronuclei) binds around much of an inclusion (Figs. 4-78, F,G, 4-79,
These small spherical nuclear remnants (Fig. 4-76) form in 4-80). When intravascular hemolysis occurs, they may be
the bone marrow following nuclear fragmentation or rupture visible as red inclusions within erythrocyte ghosts (Figs. 4-72,
of the nuclear membrane, with nuclear material left behind B, 4-78, G, 4-79, C).
A B C
FIGURE 4-76
Howell-Jolly bodies (spherical nuclear remnants) in erythrocytes. A, Erythrocyte (left) containing a Howell-
Jolly body in blood from a cat. B, Three Howell-Jolly bodies in erythrocytes in blood from a dog being treated
with glucocorticoid steroids. C, Three Howell-Jolly bodies in a single erythrocyte from a poodle with benign
macrocytosis.
formation can also occur in cats with repeated propofol anes-

thesia.21,304 Although erythrocyte survival tends to be short-
ened, anemia is either absent or mild in the above conditions.
Causes of Heinz body hemolytic anemia are presented under
“Hemolytic Anemias,” below.
Basophilic Stippling
Reticulocytes usually stain as polychromatophilic erythrocytes
with Romanowsky-type blood stains owing to the presence of
dispersed ribosomes and polyribosomes; however, sometimes
the ribosomes and polyribosomes aggregate together, forming
blue-staining punctate inclusions referred to as basophilic
stippling (Fig. 4-81).43 These aggregates are similar to those
produced using reticulocyte stains, but they form during the
A process of cell drying prior to staining with Romanowsky-
type blood stains. Diffuse basophilic stippling commonly
occurs in regenerative anemia in ruminants (Fig. 4-81, A,B)
and occasionally in regenerative anemia in other species (Fig.
4-81, C).238 It may be prominent in any species in the presence
of lead poisoning (Fig. 4-81, D).164,325 Diffuse basophilic stip-
pling was reported in a dog with dyserythropoiesis and Cabot
rings within erythrocytes.286
Siderotic Inclusions
Anucleated erythrocytes containing siderotic (iron-positive)
inclusions are called siderocytes. Nucleated siderocytes have
been called sideroblasts in human hematology. In contrast to
diffuse basophilic stippling, which is distributed throughout
the erythrocyte, siderotic inclusions generally appear as focal
B basophilic inclusions located near the periphery of erythro-
cytes (Fig. 4-81, E,F). A Prussian blue staining procedure is
FIGURE 4-77 used to verify the presence of iron-positive material (Fig. 4-81,
Formation of Howell-Jolly bodies. A, Apparent rupture of the nuclear G). Siderotic inclusions in erythroid cells may consist of cyto-
membrane of a metarubricyte, allowing nuclear material to enter the plasmic ferritin aggregates or iron-loaded mitochondria. Fer-
cytoplasm and potentially form a Howell-Jolly body. B, Transmission ritin aggregates can occur normally in nucleated erythroid cells
electron photomicrograph of a nucleated erythroid cell with expanded
nuclear pore or rupture of the nuclear membrane, allowing nuclear mate-
of humans, dogs, and pigs, but the presence of iron-loaded
rial to enter the cytoplasm and potentially form a Howell-Jolly body. mitochondria (Fig. 4-82) is a pathologic finding.205 Iron-
loaded degenerative mitochondria may be contained within
From Simpson CF, Kling JM. The mechanism of denucleation in circulating autophagic vacuoles (lysosomes).43 These inclusions have been
erythroblasts. J Cell Biol. 1967;35:237-245. called Pappenheimer bodies in human hematology.60 Electron
microscopy is used to identify the nature of siderotic inclu-
sions; however, the location of iron-positive inclusions in a ring
around the nucleus of a nucleated siderocyte (termed ringed
In contrast to other domestic animal species, normal cats sideroblast in human hematology) strongly suggests the pres-
may have up to 5% Heinz bodies within their erythrocytes.90 ence of iron-loaded mitochondria.60
Small Heinz bodies may be seen in other species following Except for iron deficiency, disorders of mitochondrial iron
splenectomy.238 Not only is cat hemoglobin more susceptible metabolism have the potential to cause excess iron accumula-
to denaturation by endogenous oxidants,205 but the cat spleen tion in mitochondria. Erythrocytes may be microcytic and/or
is less efficient in the removal (pitting) of Heinz bodies from hypochromic secondary to defective heme synthesis.233 Exper-
erythrocytes than are spleens of other species.55 Increased imental pyridoxine deficiency and experimental chronic
numbers of Heinz bodies may occur in cats with spontaneous copper deficiency have both resulted in mitochondrial iron
diseases such as diabetes mellitus (especially when ketoacido- overload in nucleated erythroid cells in the bone marrow of
sis is present), hyperthyroidism, and lymphoma.89,90 Increased deficient pigs.205 Drugs or chemicals reported to cause sidero-
numbers of Heinz bodies have been seen in kittens fed cytes and/or nucleated siderocytes in dogs include chloram-
fish-based diets218 and in cats fed commercial soft-moist phenicol (see Fig. 4-81, E), lead, hydroxyzine (Fig. 4-83), zinc,
diets containing propylene glycol.92,218 Increased Heinz body and an oxazolidinone antibiotic.213,288,370
F IGURE 4 -78 Staining characteristics and appearance of Heinz bodies
A B C D
A, The erythrocyte at the top contains a pale-pink staining Heinz body at the margin at the 7 o’clock position.
This cat had a Heinz body hemolytic anemia resulting from acetaminophen toxicity. In contrast, the erythrocyte
at the bottom contains a Howell-Jolly body that stains dark blue. Wright-Giemsa stain. B, For comparison,
erythrocytes from the same cat presented in (A) were stained with a new methylene blue reticulocyte stain. The
erythrocyte on the left contains a dark blue-staining Howell-Jolly body and the other two cells each contain a
light blue-staining Heinz body. C, Heinz bodies in blood from a cat appearing as pale “spots” within erythro-
cytes. Wright-Giemsa stain. D, Heinz bodies in blood from a cat. New methylene blue reticulocyte stain.
E F G
E, Heinz bodies in blood from a cat. New methylene blue wet mount preparation. F, Eccentrocyte with a
visible Heinz body in blood of a cat with acetaminophen toxicity. G, A large polychromatophilic erythrocyte
(top), erythrocyte “ghost” containing a Heinz body (bottom), and an intact erythrocyte containing a Heinz
body projecting from its surface (right) in blood from a dog with a hemolytic anemia resulting from the
ingestion of several pennies containing zinc. Wright-Giemsa stain.
A B C
FIGURE 4 -79
Transmission and scanning electron photomicrographs of Heinz bodies. A, Transmission electron photomi-
crograph of a Heinz body bound to the inner membrane of a horse erythrocyte. B, Scanning electron pho-
tomicrograph of a Heinz body protruding from a cat erythrocyte. C, Scanning electron photomicrograph of
two erythrocyte ghosts, each containing a Heinz body.
A, From Simpson CF. The ultrastructure of Heinz bodies in horse, dog, and turkey erythrocytes. Cornell Vet.
1971;61:228-238. B, From Jain NC. Schalm’s Veterinary Hematology. 4th ed. Philadelphia: Lea & Febiger, 1986.
C, From Jain NC. Schalm’s Veterinary Hematology. 4th ed. Philadelphia: Lea & Febiger; 1986.
Cabot Rings
Cabot rings are reddish purple-staining threadlike loops or
figure-eight structures that are primarily found in reticulo-
cytes in humans.189 Diffuse basophilic stippling is usually also
present. Spectral analysis indicates that Cabot rings contain
arginine-rich histones but not DNA.401 They may be remnants
of mitotic spindle microtubules. Cabot rings have been
reported in erythrocytes from humans with cobalamin and
folate deficiency, dyserythropoiesis, and myelodysplastic
syndrome.189
Cabot rings have been observed in normal camel and llama
erythrocytes stained with reticulocyte stains (Fig. 4-84, A,B)
and less frequently in llama erythrocytes prepared with May-
A Grunwald-Giemsa stain.30,321,483 Studies of lysed camelid
erythrocytes demonstrate thin elliptical and figure-eight-
shaped structures in reticulocytes that presumably represent
Cabot rings. When examined by transmission electron micros-
copy, these structures were described as marginal bands that
are composed of microtubules and associated proteins. These
marginal bands disappear as the cell matures.94
Cabot rings have been reported in a May-Grunwald-
Giemsa-stained blood film from a dog with dyserythropoiesis
(Fig. 4-84, C,D).286 As in humans, most Cabot rings occurred
in polychromatophilic erythrocytes.
I N F E C T I O U S AG EN T S O F
ERY T H RO C Y T E S
B
A number of infectious agents are recognized to occur in or
FIGURE 4-80 on erythrocytes. These include intracellular protozoal parasites
Idiopathic Heinz body hemolytic anemia in a horse. A, Heinz bodies (Babesia species, Theileria species, and Cytauxzoon felis), intra-
protruding from erythrocytes, including two erythrocytes that also cellular rickettsial organisms (Anaplasma species), and epicel-
contain hemoglobin crystals. A large platelet is located on the left side. lular Mycoplasma species. The erythrocyte protozoal organisms
Wright-Giemsa stain. B, Heinz bodies protruding from erythrocytes.
Apparently shrunken erythrocytes with larger Heinz bodies stain more
(piroplasms) of the order Piroplasmida each have a nucleus
basophilic. New methylene blue reticulocyte stain. within their cytoplasm. In contrast to Plasmodium and Hae-
moproteus genera, these organisms do not form pigment in
infected erythrocytes even though they also consume hemo-
globin.491 The rickettsia and mycoplasma organisms are bac-
teria and therefore lack nuclei. These infectious agents
Siderotic inclusions in erythroid cells have been recognized generally cause mild to severe hemolytic anemia depending
in some dogs and cats with myeloid neoplasms.56,524 Acquired on the pathogenicity of the organism and the susceptibility of
dyserythropoiesis with siderocytes has been reported in dogs the host. Distemper virus inclusions may also be seen in dog
in which specific etiologies could not be determined, although erythrocytes.
some of these animals had inflammatory disorders.81,513 Con-
genital anemias with ringed nucleated siderocytes have been Babesia Species
reported in humans.60,233 Persistent siderotic inclusions have More than 100 species of Babesia are recognized to infect
been recognized in microcytic hypochromic erythrocytes domestic animals, wild animals, and humans worldwide.228,491
from an English bulldog; they were composed of degenerate When stained with Romanowsky-type blood stains, a babesial
iron-loaded mitochondria (see Fig. 4-82). Echinocytes and organism (piroplasm) generally has colorless to light-blue
acanthocytes were present and erythrocytes contained Heinz cytoplasm with a red to purple nucleus (Fig. 4-85). Babesial
bodies and low numbers of hemoglobin crystals (see Fig. parasites vary considerably in size from large (2.5-5.0 µm),
4-83).209 A congenital defect resulting in mitochondrial iron easily visualized Babesia canis parasites (Fig. 4-85, A) to small
overload and secondary oxidant injury was suspected but not (1.0-2.5 µm in diameter), difficult-to-see Babesia gibsoni
identified. (Fig. 4-85, B) and Babesia felis (Fig. 4-85, C) parasites. Large
FIGURE 4-81 Diffuse and focal basophilic stippling
A B C D
A, Diffuse basophilic stippling (bottom left) in a macrocytic polychromatophilic erythrocyte, a macrocytic
erythrocyte (top right), and three normal-sized erythrocytes in blood from a cow with anaplasmosis (no organ-
isms present) and a subsequent regenerative anemia. Wright-Giemsa stain. B, Diffuse basophilic stippling in
a large erythrocyte (left) in blood from a sheep with a regenerative anemia. Wright-Giemsa stain. C, Eryth-
rocytes containing a Howell-Jolly body (top), diffuse coarse basophilic stippling (middle), and diffuse fine
basophilic stippling (bottom) in blood from a cat with Mycoplasma haemofelis infection (no organisms present)
and a regenerative anemia. Wright-Giemsa stain. D, A polychromatophilic erythrocyte with basophilic stip-
pling (left) and a polychromatophilic metarubricyte (right) in blood from a dog with lead toxicity. Wright-
Giemsa stain.
E F G
E, Focal basophilic stippling in an erythrocyte in blood from a dog treated with chloramphenicol. The inclu-
sions were shown to contain iron using the Prussian blue staining procedure and can therefore be called
siderocytes. Wright-Giemsa stain. F, Focal basophilic stippling in two erythrocytes (siderocytes) in blood from
a male Sheltie dog that had many siderocytes in his blood when examined several times over 4 years. Eryth-
rocytes were microcytic but the dog was not anemic. The dog was treated with hydroxyzine. Abnormalities in
copper, zinc, and pyridoxine metabolism were ruled out, as was lead toxicity. Wright-Giemsa stain. G, Iron-
positive inclusions in erythrocytes (siderocytes) in blood from the same dog as shown in (F), Prussian blue
stain.
F-G, Blood samples and case information provided by M. Plier.

A B
FIGURE 4-82
Transmission electron photomicrograph of a cluster of iron-loaded (dark
material) degenerating mitochondria in a circulating erythrocyte from an
English bulldog, which produced focal basophilic stippling seen in eryth-
rocytes by light microscopy in this animal (see Fig. 4-83). C D
Courtesy of W. L. Clapp. FIGURE 4 -84

Cabot rings in erythrocytes. A, Cabot ring with a figure-eight shape in
an erythrocyte from an adult alpaca with a moderately regenerative
hypochromic iron-deficient anemia secondary to gastric ulceration and
hemorrhage. New methylene blue reticulocyte stain. B, Erythrocyte with
a Cabot ring at the periphery and a reticulocyte (bottom right) in blood
from an adult alpaca with a moderately regenerative hypochromic iron-
deficient anemia secondary to gastric ulceration and hemorrhage. New
methylene blue reticulocyte stain. C, Cabot ring with a figure-eight shape
in an erythrocyte from a dog with dyserythropoiesis. D, Cabot rings in
an erythrocyte from a dog with dyserythropoiesis. May-Grunwald-
Giemsa stain.
A, From a stained blood film from a 2009 ASVCP slide review case submitted
by B. Fierro and M. Scott. B, From a stained blood film from a 2009 ASVCP
slide review case submitted by B. Fierro and M. Scott. C and D, From
Lukaszewska J, Lewandowski K. Cabot rings as a result of severe dyseryth-
ropoiesis in a dog. Vet Clin Pathol. 2008;37:180-183.
FIGURE 4 -83
Siderotic inclusions in microcytic hypochromic erythrocytes in a blood
film from an English bulldog with a presumptive defect in mitochondrial
iron metabolism. A, Basophilic stippling is diffuse in one erythrocyte (top
center) but focal in other erythrocytes. Echinocytes and acanthocytes are
present, and two erythrocytes contain hemoglobin crystals. Wright-
Giemsa stain. B, Focal (iron-positive) basophilic inclusions are present
in a blood film from the same dog presented in (A). These siderotic
inclusions were composed of degenerate, iron-loaded mitochondria (see
Fig. 4-82). Prussian blue stain.
B
A B C
D E F
FIGURE 4-85 Babesia organisms in erythrocytes

A, Two pear-shaped Babesia canis organisms in each of four erythrocytes in blood from a puppy with hemolytic
anemia. Infected erythrocytes often were seen to be adhering to one another. Wright-Giemsa stain. B, Single
small Babesia gibsoni organisms in two erythrocytes from a dog. C, Single Babesia felis organisms in three
erythrocytes in blood from a domestic cat from South Africa. Wright stain. D, Two Babesia bigemina organ-
isms in an erythrocyte from a cow. Wright-Giemsa stain. E, Two pear-shaped Babesia caballi organisms in a
horse erythrocyte. Wright-Giemsa stain. F, A single Babesia equi organism in one erythrocyte (top) and a
Maltese cross of four organisms in another erythrocyte (bottom) in blood from a horse. Wright-Giemsa stain.
babesial organisms generally appear pear-shaped and com- dogs. B. microti organisms exhibit characteristics of Theileria
monly occur in pairs. Small babesial organisms are more often species, and a B. microcoti-like organism in dogs has been
round in shape. Aside from size, the morphology of protozoal called Theileria annae, but a new genus may be required for
organisms infecting erythrocytes is similar; therefore poly- this group of organisms.491
merase chain reaction (PCR) and 18S rRNA gene sequencing B. felis (Fig. 4-85, C) causes hemolytic anemia in cats in
is required to specifically identify these organisms.228 New South Africa.362 Additional piroplasms have been identified
organisms will continue to be identified using genetic by PCR in domestic cats, but their clinical significance remains
analysis. to be determined.491 The two most important Babesia species
Three genetically distinct large B. canis subspecies (B. canis infecting cattle appear to be Babesia bigemina (Fig. 4-85, D)
canis, B. canis vogeli, and B. canis rossi) have been identified as and Babesia bovis. The two most important piroplasms infect-
causing disease in dogs.467 In reality, these organisms appear ing horses are the large Babesia caballi parasite (Fig. 4-85, F)
to be different species; consequently they may be renamed B. and the smaller, more pleomorphic Babesia. equi parasite (Figs.
canis, B. vogeli (Fig. 4-85, A), and B. rossi, respectively.491 In 4-85, E, 4-86). Some investigators have recommended that
addition, a large Babesia species was originally described in B. equi be renamed Theileria equi based on genetic studies as
dogs in North Carolina; it has yet to be named.51,226,427 At least well as finding schizonts transiently in lymphocytes.318 Other
three small protozoal species, B. gibsoni (Fig. 4-85, B), Babesia investigators believe B. equi belongs in a new genus that is
conradae, and a Babesia microti-like organism cause disease in distinct from both Babesia and Theileria.14,491
A B
FIGURE 4-86
Babesia equi organisms in erythrocytes from a horse with clinical disease.
Although pleomorphism (including variation in organism size) was
prominent, PCR and gene sequencing of the 18S rDNA gene identified
a single infective agent. Wright-Giemsa stain. C D
FIGURE 4 -87
Theileria and Cytauxzoon organisms in erythrocytes. A, Theileria buffeli
Theileria Species organism in an erythrocyte in blood from a cow. Wright stain. (Photo-
graph from a 2001 ASVCP slide review case submitted by A. Boisvert and
Theilerial organisms appear similar to babesial organisms R. Pillars.) B, Theileria cervi organisms in erythrocytes in blood from a
when observed on stained blood films (Fig. 4-87, A). The white-tailed deer. Several drepanocytes (sickle erythrocytes) are present.
genus Theileria differs from the genus Babesia in that the Wright-Giemsa stain. C, A Cytauxzoon felis organism in a domestic cat
erythrocyte. Wright-Giemsa stain. D, Single Cytauxzoon felis organisms
former species has a tissue phase as well as an erythrocyte in several domestic cat erythrocytes. Wright-Giemsa stain.
stage of development. Schizonts develop in lymphoid cells
and, when mature, release merozoites, which enter erythro-
cytes. Babesia organisms proliferate only in erythrocytes.
Theilerial species (Theileria parva and Theileria annulata)
cause important diseases in ruminants in Africa, Asia, and the is directly related to the number of subunits present. Unlike
Middle East274; however, the theilerial organisms present in Howell-Jolly bodies, Anaplasma organisms are generally not
ruminants in the United States are usually nonpathogenic.245,453 perfect spheres, and most are smaller than Howell-Jolly bodies.
Piroplasms are commonly observed in deer blood in the Anaplasma marginale is an important pathogen of cattle.
United States (Fig.4-87, B). They are generally considered of Organisms are often located at the margin in erythrocytes
low pathogenicity but may cause hemolytic anemia under when viewed on stained blood films (see Fig. 4-88). Anaplasma
some circumstances.544 centrale is less pathogenic and organisms are more often located
more centrally in erythrocytes when viewed on stained blood
Cytauxzoon felis films.159 Anaplasma ovis (see Fig. 4-88, C) is pathogenic for
Cytauxzoon felis (Fig. 4-87, C,D), as its name implies, infects sheep, goats, and some wild ruminants.112
feline erythrocytes.185 It is similar in morphology to B. felis in
erythrocytes (see Fig. 4-85, C). Like Theileria, the genus
Cytauxzoon has both a tissue phase and an erythrocyte phase. Distemper Inclusions
In contrast to Theileria, the schizonts of Cytauxzoon develop Viral inclusions may be seen in the blood cells of some dogs
in macrophages rather than in lymphocytes. during the viremic stage of canine distemper virus infec-
tion.181,504 These inclusions can be difficult to visualize when
Anaplasma Species routine Wright or Giemsa stains are used. In erythrocytes,
Anaplasma organisms appear as round to oval basophilic inclu- they appear as variably sized round, oval, or irregular blue-gray
sions in ruminant erythrocytes (Fig. 4-88), which must be inclusions that most often occur in polychromatophilic cells
differentiated from Howell-Jolly bodies.500 Although morulae (Fig. 4-90, A). For an unknown reason, distemper inclusions
are not appreciated by light microscopy, the inclusions consist typically stain red and are easier to see in erythrocytes stained
of one to several subunits within a membrane-lined vesicle with the aqueous Diff-Quik stain (Fig. 4-90, B), which is a
(Fig. 4-89). The size of an inclusion seen by light microscopy rapid modified Wright stain.13,198
A B C
FIGURE 4-88
Anaplasma organisms in erythrocytes. A, Anaplasma marginale organism located within an erythrocyte in blood
from a Holstein cow. Three platelets are also visible (right). Wright-Giemsa stain. B, An erythrocyte contain-
ing an Anaplasma marginale organism (bottom left), a macrocytic erythrocyte (top left), and an abnormally
shaped erythrocyte with basophilic stippling (right) in blood from a Holstein cow. Wright-Giemsa stain.
C, Anaplasma ovis organisms in blood from a 6-month-old goat with esophageal perforation and intestinal
Trichostrongylus infestation. Wright-Giemsa stain.
A B A B

Transmission electron photomicrograph of Anaplasma marginale organ- Distemper inclusions in dog erythrocytes. A, Round blue-gray distemper
isms in bovine erythrocytes. A, Binary fission of an Anaplasma organism inclusion in a polychromatophilic erythrocyte (top left). Wright-Giemsa
within an erythrocytic vacuole. B, Six Anaplasma organisms within an stain. B, Two round reddish distemper inclusions in erythrocytes in
erythrocytic vacuole. blood from the same dog as shown in (A). The inclusion at top right is
in a large polychromatophilic erythrocyte. Diff-Quik stain.
From Simpson CF, Kling JM, Love JN. Morphologic and histochemical nature
of Anaplasma marginale. Am J Vet Res. 1967;28:1055-1065.
Distemper inclusions are composed of aggregates of viral Hemoplasmas has been proposed as a trivial name for these
nucleocapsids.310 The presence of viral inclusions in anucleated hemotropic mycoplasmas.340
cells is explained by the fact that they form within nucleated Hemoplasmas appear as small (generally 0.5 to 1 µm)
erythroid precursors in the bone marrow and persist following blue-staining cocci, rods, or rings on erythrocytes in blood
expulsion of the nucleus.181 films stained with Wright-type blood stains (Fig. 4-94, A,B).
Reticulocyte stains cannot be used to search for mycoplasmas
Hemotropic Mycoplasmas (Hemoplasmas) because the basophilic ribosomal material in reticulocytes can
Hemotropic mycoplasmas are gram-negative non-acid-fast appear similar to the organisms. These organisms were classi-
bacteria that attach to the external surfaces of erythrocytes fied as rickettsia in the genus Haemobartonella or the genus
(Figs. 4-91, 4-92, 4-93), although evidence that Mycoplasma Eperythrozoon for many years.202 Organisms that were tightly
suis can enter erythrocytes has recently been published.187 bound to erythrocyte surfaces, with prominent cocci and rod

Scanning electron photomicrograph of erythrocytes from a cat infected Scanning electron photomicrograph of a bovine erythrocyte parasitized
with Mycoplasma haemofelis. with M. wenyoni.
From Harvey JW. Hemotrophic mycoplasmosis (hemobartonellosis). In: Greene From Keeton KS, Jain NC. Eperythrozoon wenyoni: a scanning electron
CE, ed. Infectious Diseases of the Dog and Cat. 3rd ed. Philadelphia: microscopy study. J Parasitol. 1973;59:867-873.
Saunders Elsevier; 2006:252-260.
manner in which a blood film is prepared. In addition, organ-

isms detach from erythrocytes over time in blood samples
collected using EDTA as an anticoagulant (Fig. 4-94, C).12
Results from sequencing of the 16S rRNA gene indicate
that all of these epicelluar erythrocyte parasites are mycoplas-
mas.202,539 Consequently the Haemobartonella and Eperythro-
zoon genera have been discarded and organisms in these
genera were moved to the genus Mycoplasma. Species names
often include the prefix haemo (e.g., Mycoplasma haemofelis)
to identify these unique mycoplasmas that attach to
erythrocytes.
Three different hemoplasmas have been identified in cats
based on 16S rRNA gene sequences, Mycoplasma haemofelis
(formerly called the large form of Haemobartonella felis), Can-
didatus Mycoplasma haemominutum (formerly called the small
form of Haemobartonella felis), and Candidatus Mycoplasma
FIGURE 4-92 turicensis.473,539 M. haemofelis is more pathogenic than either
Transmission photomicrograph of Mycoplasma haemofelis organisms of the other organisms.473
attached to the external surface of a cat erythrocyte. M. haemofelis organisms appear as small blue-staining
cocci, rods, or rings on feline erythrocytes (see Fig. 4-94, A).
From Simpson CF, Gaskin JM, Harvey JW. Ultrastructure of erythrocytes
parasitized by Haemobartonella felis. J Parasitol. 1978;64:504-511.
Chains of organisms may be present on the surface of heavily
parasitized erythrocytes (Fig. 4-95). Ring- and rod-shaped
organisms are seen more readily in thin blood films. Organ-
isms are pleomorphic and vary in size, but most are between
forms, were classified in the Haemobartonella genus. Organ- 0.5 and 1.5 µm in diameter or length. They appear to be
isms that were often found between erythrocytes, as well as partially buried in indented foci on the surface of the eryth-
adhered to erythrocytes, with prominent ring forms, were rocytes (Fig. 4-96). Parasitized erythrocytes may lose their
classified in the Eperythrozoon genus. These criteria seemed normal biconcave shape and become spherocytes or stomato-
inadequate for the establishment of two genera, especially spherocytes (see Fig. 4-91). Organisms occur in cyclic para-
since the frequency of ring forms and the number of free sitemias; consequently they are not always identifiable in
organisms can be influenced to some degree simply by the blood even during acute infections.202
FIGURE 4-94 Hemotrophic mycoplasmas (hemoplasmas)
A B C D
A, Mycoplasma haemofelis organisms located on the surface of erythrocytes in blood from a cat. Some organ-
isms appear as rings, including the unattached one at the bottom right. Wright-Giemsa stain. B, Mycoplasma
haemominutum organisms on the outside of an erythrocyte. C, An aggregate of free Mycoplasma haemofelis
organisms that have detached from erythrocytes following 3 days in transit to the laboratory of infected blood
collected with EDTA as the anticoagulant. Wright stain. D, Mycoplasma haemocanis organisms located on
the outside of erythrocytes in blood from a dog. One erythrocyte (center) has a rod-shaped structure on its
surface that may be composed of two closely associated organisms, while another erythrocyte (top right) has
many organisms forming filamentous chains in deep grooves on its surface. A platelet and large polychro-
matophilic erythrocyte are also present. Wright-Giemsa stain.
E F G H
E, Mycoplasma suis organisms on the surface of erythrocytes and between erythrocytes in blood from a
splenectomized pig. Erythrocytes appear as echinocytes, a normal finding in pig blood. Wright stain.
F, Mycoplasma ovis organisms between erythrocytes in blood from a sheep. Wright-Giemsa stain. G, Myco-
plasma wenyoni organisms between erythrocytes in blood from a Charolais bull. Wright stain. H, Mycoplasma
haemolamae organisms on the surface of erythrocytes and between erythrocytes in blood from a llama. Wright-
Giemsa stain.
B, Courtesy of J. B. Messick. C, From Allison RW, Fielder SE, Meinkoth JH. What is your diagnosis? Blood film
from an icteric cat. Vet Clin Pathol. 2010;39:125-126. E, Photograph of a stained blood film from a 1980 ASVCP
slide review case submitted by G. Searcy. F, Photograph of a stained blood film from a 1993 ASVCP slide review case
submitted by E. G. Welles, J. W. Tyler, and D. F. Wolfe.
Candidatus M. haemominutum organisms (see Fig. 4-94, B) identification. Candidatus M. turicensis has been documented
are rarely recognized in stained blood films. When seen, they in blood from many cats using PCR-based assays, but it has
appear as small rods or coccoid organisms and infrequently as not yet been identified in stained blood films from infected
ring forms, which stain less densely and measure about half cats, presumably because of the low numbers of organisms
the size (approximately 0.3 µm) of M. haemofelis.142,165 present.539
However, this reported size difference has been challenged by Three hemoplasmas have been reported in dogs based on
the finding of a Candidatus M. haemominutum isolate in 16S rRNA gene sequences. Organisms are generally observed
Great Britain that is about 0.6 µm in diameter.472 Conse- only in splenectomized or immunosuppressed dogs.202 Single
quently morphologic appearance is not reliable in distinguish- Mycoplasma haemocanis (formerly Haemobartonella canis)
ing these isolates; genetic analysis is needed for specific organisms dimple the surface of host erythrocytes in a manner
recent studies evaluating RNase P gene sequences demon-

strated a lower degree of sequence homology between the two
organisms (about 95%), suggesting that the organisms may
represent different species.48
A hemoplasma with 99% 16S rRNA gene homology to
Candidatus M. haemominutum has been identified in a dog
in China.551 Another hemoplasma with somewhat less 16S
rRNA gene homology has been identified in multiple coun-
tries and classified as Candidatus Mycoplasma haematopar-
vum.35,465,539 Candidatus Mycoplasma haematoparvum appeared
as small (0.3 µm), basophilic coccoid bodies on the surface of
erythrocytes in a stained blood film from a dog.464
Hemoplasmas, previously classified as Eperythrozoon
organisms, infect pigs (Mycoplasma suis, Fig. 4-94, E),221 sheep
and goats (Mycoplasma ovis, Fig. 4-94,G),339 cattle (Myco-
plasma wenyoni [Figs. 4-93, 4-94, F], and Candidatus Myco-
FIGURE 4-95 plasma haemobos or the synonymous Candidatus Mycoplasma
Large numbers of Mycoplasma haemofelis organisms (including ones haemobovis),343,457,493 and llamas and alpacas (Candidatus
forming chains) located on the outside of erythrocytes in blood from a Mycoplasma haemolamae, Fig. 4-94, H).315,390
cat. A large polychromatophilic erythrocyte (aggregate reticulocyte) is
Hemoplasma infections have been recognized for many
also present (upper right). Wright-Giemsa stain.
years in animals but, using PCR assays, have only recently
been documented in humans. These human organisms were
related to M. suis, M. wenyoni, and M. ovis. 95,466 Based on the
widespread use of PCR assays, new hemoplasmas will
undoubtably be identified in animals and humans.
Bartonella Species
Members of the Bartonella species are small gram-negative
bacteria. Cats and dogs are infected with multiple Bartonella
species.88 Bartonella henselae appears to be the primary cause
of cat-scratch disease in humans. This organism causes mild
illness and anemia in cats during the initial infection; subse-
quently, however, cats generally become carriers without evi-
dence of disease.63 This small rod-shaped bacterium occurs
within erythrocytes264 but is rarely appreciated in blood films
of bacteremic cats (Fig. 4-97, A), even though the organism
can be cultured from the blood of many healthy cats.63
Artifacts Resembling Infectious Agents

FIGURE 4-96 Erythrocyte parasites (especially hemoplasmas) must be dif-
Blood film from a cat with Mycoplasma haemofelis infection demonstrat- ferentiated from precipitated stain, refractile drying or fixa-
ing features that help distinguish this parasite from stain precipitation. tion artifacts (Fig. 4-97, B), poorly staining Howell-Jolly
An organism indents the membrane of the erythrocyte in the upper left. bodies, and basophilic stippling. Platelets overlying erythro-
A second organism (center right) binds two areas of an erythrocyte mem-
brane together. Two polychromatophilic erythrocytes (aggregate reticu- cytes (Fig. 4-97, C) may also be confused with erythrocyte
locytes) are also present. parasites, especially Babesia species.
ERY T H RO C Y T E A S S AY S
similar to M. haemofelis and chains of organisms are frequently Erythrocyte Counts, Hematocrit, and
found in grooves or deep folds, which can markedly distort Hemoglobin Content
the erythrocyte shape and appear as filamentous structures on Erythrocytes in blood are quantified by cell counting (number
the surface of erythrocytes (see Fig. 4-94, D). M. haemocanis of cells per microliter), by determining blood hemoglobin
has long been considered to be a distinctly different organism content (grams per deciliter), and by determining the hema-
from M. haemofelis, but the sequence of the 16S rRNA gene tocrit (HCT) as a percentage of blood volume. Because essen-
of a M. haemocanis isolate from one dog was remarkably tially all hemoglobin is present within erythrocytes, the
similar (99% homology) to that of M. haemofelis.67 More erythrocyte count or red blood cell (RBC) count, HCT, and
A B C
FIGURE 4-97
Bartonella organisms and artifacts that might be confused with erythrocyte parasites. A, Bartonella henselae
organisms in an erythrocyte (center right) from a confirmed bacteremic cat. In addition to blood culture,
organisms were identified in fixed erythrocytes using a fluorescent-labeled antibody. Wright-Giemsa stain.
B, Drying artifact and precipitated stain present in this blood film from a cat may be confused with blood
parasites. Wright-Giemsa stain. C, A platelet overlying an erythrocyte (bottom left) may be confused with a
blood parasite in this blood film from a dog. Wright-Giemsa stain.
A, Courtesy of R. E. Raskin.
hemoglobin content parallel each other when a change occurs. sympathetic tone keeps the spleen slightly contracted during
The term packed cell volume (PCV) is often used when the the awake resting state.110 A slight postprandial increase in
HCT is measured by centrifugation of blood in a microhe- HCT has been reported after feeding dogs and sheep; this
matocrit centrifuge. The PCV is the easiest and most repro- persists for several hours, and most of this change in dogs and
ducible test available for quantifying erythrocytes in clinical about half of this change in sheep is attributable to splenic
practice. The RBC count and hemoglobin content need to be contraction.123,268 HCT increases associated with splenic con-
measured only when erythrocyte indices are to be calculated. traction in ruminants and pigs are smaller than increases in
The RBC count and mean cell volume (MCV) are accurate cats, dogs, and horses.376,448 Conversely, anesthesia (especially
if they are measured using an electronic cell counter that has with barbiturates) can produce splenic enlargement, causing
been designed or adjusted to measure the variably sized eryth- the HCT to drop below reference intervals.66
rocytes of animals. Hemoglobin concentration is measured Maximum information can be gained by interpretating the
spectrophotometrically. Modern electronic cell counters cal- HCT and plasma protein concentrations simultaneously.
culate the HCT using the measured RBC count and MCV. Various combinations of low, normal, or high HCT values
This efficiency negates the need to centrifuge a microhema- may occur with low, normal, or high plasma protein concen-
tocrit tube of blood. Unfortunately, useful information con- trations. The various combinations and examples of how they
cerning the appearance of plasma is missed when the HCT can be interpreted are given in Box 4-1.
is determined electronically unless a serum sample is also
prepared for clinical chemistry tests. Reticulocyte Counts
Cats, dogs, hot-blooded horses, and some marine diving Manual methods used in performing reticulocyte staining and
mammals (e.g., seals) have large (as much as one-third of the counting are given in Chapter 2. Reticulocytes in cats are
total blood volume in horses) contractile spleens.449 This pro- classified as aggregate (if coarse clumping is observed) or
vides a blood reservoir that can be released into the general punctate (if small individual inclusions are present). In healthy
circulation in response to sympathetic stimulation induced by cats as well as cats with regenerative anemia, the number of
exercise, hypoxia, hemorrhage, or excitement. The capacity of punctate reticulocytes is much greater than that seen in other
the spleen to expand and contract results in substantial changes species.15 In contrast to those of the cat, most reticulocytes in
in the peripheral blood HCT in these species because the other species are of the aggregate type; consequently no
HCT in the spleen (about 80%-90%) is much higher than attempt is made to differentiate stages of reticulocytes except
that in peripheral blood.110,449 Maximal splenic contraction in cats. The higher number of punctate reticulocytes occurs
increases the HCT 1.3- to 1.5-fold above resting levels in in cats because the maturation (loss of ribosomes) of
these species.66,110,309 The HCT is slightly higher in spleen- reticulocytes in cats is slower than that in other species.
intact versus splenectomized dogs, presumably because basal Aggregate reticulocytes in the circulation mature to punctate
reticulocyte count would be 9 divided by 45 (mean normal

Box 4-1 HCT for dogs) times 6% = 1.2%. The corrected reticulocyte
Concomitant Interpretation
of Hematocrit (HCT) and count is used to determine whether reticulocytes are truly
Total Plasma Protein (TPP) increased in blood. Although the raw reticulocyte count (6%)
Concentration suggests that the reticulocyte numbers were increased sub-
stantially in blood, the corrected reticulocyte count demon-
Normal HCT with
strates that little increase in reticulocyte numbers was present
Low TPP: Gastrointestinal protein loss, proteinuria, severe
liver disease, vasculitis in blood. Normal dogs generally have no more than about 1%
Normal TPP: Normal reticulocytes when this value is determined by manual
High TPP–Increased globulin synthesis, dehydration-masked methods.
anemia If the total RBC count is known, an absolute reticulocyte
High HCT with count (reticulocytes per microliter) can be determined. This is
Low TPP: A combination of splenic contraction and a source done by multiplying the percentage of reticulocytes counted
of protein loss (expressed as a fraction) by the total RBC count.
Normal TPP: Splenic contraction, primary or secondary eryth-
rocytosis, dehydration-masked hypoproteinemia
Absolute reticulocyte count ( per microliter)
High TPP: Dehydration
Low HCT with = RBC count ( per microliter )
Low TPP: Substantial ongoing or recent blood loss, × raw reticulocyte count ( fraction)
overhydration
Normal TPP: Increased erythrocyte destruction, decreased It should be noted that the calculations of absolute reticu-
erythrocyte production, chronic blood loss locyte counts and corrected reticulocyte counts are indepen-
High TPP: Anemia of inflammatory disease, multiple myeloma, dent calculations, with each one using the original raw
lymphoproliferative diseases, hepatocellular carcinoma (one reticulocyte count.
report in a dog)97 Absolute reticulocyte counts in normal dog are generally
less than 80 × 103/µL when manual counts are done. Deter-
mined manually, absolute aggregate reticulocyte counts in 41
reticulocytes in a day or less, but a week or more is required normal cats were between 0 and 95 × 103/µL and absolute
for maturation (total disappearance of ribosomes) of punctate punctate reticulocyte counts were between 0 and 650 × 103/µL.
reticulocytes in cat blood.133 Percentages of both types should Absolute reticulocyte counts can also be determined
be reported separately in cats. Manual reticulocyte counts directly by flow cytometry with some automated hematology
were done in blood from 41 healthy cats. Aggregate reticulo- analyzers. These instruments provide more rapid results with
cyte counts were between 0% and 0.9% and punctate reticu- better precision than the manual method. Their use is also
locyte counts were between 0% and 7.4%. much less labor-intensive. However, it is essential that these
Raw (uncorrected) manual reticulocyte counts can be mis- automated counts be validated by comparing automated
leading when moderate to severe anemia is present because counts against manual counts for accuracy. Automated refer-
reticulocytes are quantified as a percentage of the total number ence intervals can vary considerably depending on the instru-
of erythrocytes (reticulocytes plus mature erythrocytes) ment used. In cats, it is especially important to determine
counted. The raw reticulocyte count (percent) would be higher whether some punctate reticulocytes are counted by the
in an anemic animal (with a lower number of mature eryth- machine, along with the aggregate reticulocytes, before
rocytes) than it would be in a normal animal (with a higher the reticulocyte counts can be interpreted appropriately. The
number of mature erythrocytes), even if the actual number of Advia 120 (Siemens Healthcare Diagnostics, Inc., Tarrytown,
reticulocytes per microliter in the circulation was the same in NY) hematology analyzer appears to count primarily aggre-
each animal. Consequently reticulocyte counts should either gate reticulocytes in cats; reticulocyte counts from the same
be corrected for the degree of anemia using the HCT or an 41 normal cats listed above were between 8 × 103/µL and 57
absolute reticulocyte count should be calculated using the × 103/µL, when measured by the Advia 120. This was similar
total RBC count. The reticulocyte count is corrected by divid- to the absolute reticulocyte counts in 58 normal dogs, which
ing the patient’s HCT by the mean normal HCT for the were between 8 × 103/µL and 65 × 103/µL when measured
species and then multiplying this value by the raw reticulocyte using the Advia 120.
count to obtain a corrected reticulocyte count. The corrected reticulocyte response to blood loss anemia in
the cat is shown in Figure 4-98.15 As in other species, about
Corrected reticulocyte count 4 days are required to obtain a maximal aggregate reticulocyte
= ( patient’s HCT/mean normal HCT for species) response to anemia because of the time needed for the pro-
× raw reticulocyte count in percent duction of aggregate reticulocytes from progenitor cells. The
maximal punctate response occurs considerably later, primar-
For example, if the raw reticulocyte count was determined ily because of the long time required for punctate reticulocytes
to be 6% in a dog with an HCT of 9%, the corrected to mature to erythrocytes. As can be seen in Figure 4-98,
Controlled bleeding in cats

45 45
Erythrocyte Indices
40
Before After bleeding
40 Determination of erythrocyte indices can assist in the dif-
ferential diagnosis of anemia. Of the erythrocyte parameters
Corrected reticulocytes (%)

35 35
routinely determined or calculated, the MCV is the most
30 30
useful.
PCV (%)
25 25
20 PCV 20 Mean Cell Volume
15 15 The MCV represents the average volume of a single erythro-
Punctate reticulocytes cyte in femtoliters (10−15 L) in a population of erythrocytes
10 10
Aggregate reticulocytes
(typically whole blood). Erythrocytes lose volume and hemo-
5 5
globin through vesiculation as they age. The MCV measured
0
642 0 2 4 6 8 10 12 14 16 18 20 22 24
0 in an aged human erythrocyte population was decreased by
30%, while the mean cell hemoglobin concentration (MCHC)
Days
was increased by 15%.535 The MCV is determined most accu-
FIGURE 4-98 rately by direct measurement with electronic cell counters. It
Reticulocyte response following controlled bleeding in cats. Reticulocyte can be determined indirectly by dividing the HCT (as a per-
counts have been corrected using packed cell volume (PCV) values. centage) by the RBC count (in millions of cells per microliter)
and multiplying by 10, but this calculated value is less accurate
Data from Alsaker RD, Laber J, Stevens JB, et al. A comparison of polychro- because two separate measurements are required. The MCV
masia and reticulocyte counts in assessing erythrocyte regenerative response in
varies greatly depending on species. Mammals have smaller
the cat. J Am Vet Med Assoc. 1977;170:39.
erythrocytes than birds, reptiles, or amphibians.215 Erythro-
cytes (and other blood cell types) are especially large in
amphibians with MCVs in excess of 10,000 fL in the Amphi-
punctate reticulocyte continue to be released from bone uma salamander (see Fig. 4-8).175 Species with larger erythro-
marrow after the HCT begins to increase and aggregate retic- cytes have lower RBC counts, resulting in similar HCTs and
ulocyte release has ceased. Consequently cats with mild regen- hemoglobin concentrations in mammals and birds.215 MCVs
erative anemia may have increased punctate reticulocyte can vary with age, with higher MCVs reported in older horses
counts and normal aggregate reticulocyte counts (see Fig. and cattle.84,238 Slight increases in MCVs are reported with
4-23, A). exercise in horses.302,329
Lower HCTs typically result in higher plasma erythropoi- When identified, high MCV values (macrocytosis) are
etin concentrations,351 which lead to higher absolute blood usually associated with regenerative anemias because the
reticulocyte counts (except in horses) if the marrow is able to volumes of individual reticulocytes (especially stress reticulo-
respond appropriately. High erythropoietin concentrations cytes) are larger than the volumes of mature erythrocytes.
also lead to the early release of reticulocytes from the bone However, it is important to remember that many macrocytic
marrow into the blood.7 Instead of continuing to mature in cells must be present to increase the MCV above the normal
the bone marrow, these more immature reticulocytes mature reference interval; consequently the MCV is usually within
in the circulation. This reticulocyte maturation time or resi- reference intervals in animals with regenerative anemia.119
dence time increases from about 0.5 day to 1 day in sheep Some dogs with nonregenerative immune-mediated anemia
following a phlebotomy removing more than half of the cir- and/or myelofibrosis also have a macrocytosis.454 High MCVs
culating erythrocyte mass.146 Absolute reticulocyte counts are may occur in animals with myeloid neoplasms and nonregen-
generally higher in response to hemolytic anemia than they erative anemia.125,257,527 Macrocytosis is often seen in feline
are in response to hemorrhage, presumably because plasma leukemia virus (FeLV)-positive cats with nonregenerative
iron concentration is high in animals with hemolytic anemia anemias.507 Folate deficiency has been reported as a cause of
and normal or low in animals following hemorrhage. Conse- macrocytic nonregenerative anemia in a cat.332 Macrocytosis
quently the presence of a marked reticulocytosis indicates the (without anemia or reticulocytosis) occurs in some apparently
likelihood that increased erythrocyte destruction is the cause healthy miniature and toy poodle dogs that have variable
of the anemia. When the degree of anemia is severe, basophilic megaloblastic abnormalities in the bone marrow and normal
macroreticulocytes or so-called stress reticulocytes may be serum folate and cobalamin values.80 Dogs with hereditary
released into the blood (see Fig. 4-21). It is proposed that one stomatocytosis may have high MCVs, with normal or only
less mitotic division occurs during production and that large slightly increased reticulocyte counts.59,231 Some cats with
immature reticulocytes are released. Although a portion of hyperthyroidism have slightly increased MCVs with normal
these macroreticulocytes may be rapidly removed from the or increased HCTs.365 Macrocytic anemia has been reported
circulation, it appears from studies in cats that some can in Hereford calves with congenital dyserythropoiesis. In these
mature into large (macrocytic) erythrocytes with relatively calves, many nucleated erythrocytes are present in blood but
normal life spans.506 reticulocyte counts are only slightly increased.447
High MCVs may occur as an artifact secondary to (rcEPO) therapy.387 Microcytosis apparently occurs because
agglutination of erythrocytes, as can occur in immune- iron delivery to the developing erythroid cells is not sufficient
mediated disorders or following heparin administration to to fully support the accelerated erythropoiesis accompanying
horses.322,374,489,550 MCVs may also be spuriously increased in rcEPO administration.
cats and dogs with persistent hypernatremia because the cells Copper is needed for optimal iron absorption and release
can swell in vitro when diluted with counting fluid prior to from body iron stores. Consequently prolonged copper defi-
sizing in an electronic cell counter.58 Finally, MCVs increase ciency results in microcytic anemia in some species. Pyridox-
with prolonged storage of blood samples; however, the increase ine is required for the first step in heme synthesis. Although
may not be sufficient to elevate the value beyond the reference natural cases of pyridoxine deficiency have not been docu-
interval.156,311 mented in domestic animals, microcytic anemias with high
Macrocytosis is more likely to occur in response to hemo- serum iron values have been produced experimentally in dogs,
lytic anemia than to hemorrhage, at least in part because cats, and pigs with dietary pyridoxine deficiency.205
serum iron concentration is increased in animals with hemo- The MCV may be slightly decreased in association with
lytic anemia. While iron does not stimulate erythropoiesis, the anemia of inflammatory disease, but the MCV is at the
decreased iron availability may limit the erythropoietic low end of the reference interval in most cases. Microcytosis
response following hemorrhage. Reticulocytes, especially is common in dogs with portosystemic shunts.243 In these
those produced in response to severe anemia (stress reticulo- cases, the MCV is seldom more than 7 fL below the reference
cytes), are larger than mature erythrocytes. A week or more is interval and the HCT is within the reference interval or only
required before macrocytosis occurs in response to hemolytic slightly decreased.201 This modest decrease in MCV may be
anemia. Although the bone marrow normally contains some masked if blood is stored for 24 hours before being assayed.178
reticulocytes undergoing maturation, most reticulocytes Some cats with portosystemic shunts and hepatic lipidosis
released from the bone marrow in response to anemia must be exhibit slight microcytosis.85,280 Drugs or chemicals that inter-
formed de novo. A minimum of 4 days is required for a peak fere with heme synthesis, such as chloramphenicol, lead, and
reticulocyte response to occur,15,136,364 and then the newly pro- probably hydroxyzine (dogs) have the potential for causing
duced, larger cells must comprise a high enough percentage of the formation of microcytic erythrocytes with siderotic inclu-
the total erythrocytes present to increase the MCV above the sions.205 Microcytic anemia may also occur in myeloid
reference interval. Although there is a reduction in size as neoplasms exhibiting iron accumulation in erythroid cells.524
reticulocytes mature into erythrocytes, larger-than-normal Persistent siderotic inclusions have been recognized in micro-
reticulocytes produce larger-than-normal erythrocytes.506 cytic hypochromic erythrocytes from an English bulldog (see
The erythrocytes produced in the fetus are larger than those Fig. 4-83, A,B). These erythrocytes also contained Heinz
produced in the adult.10,62,334 There is a gradual decrease in bodies and rare hemoglobin crystals.209 A congenital defect
MCV during fetal development. The MCV is within adult resulting in mitochondrial iron overload and secondary
reference intervals in horses and cattle at birth.207,238 The oxidant injury was suspected but not identified. A nonregen-
MCV is above adult reference intervals in dogs and cats at erative microcytic anemia with many circulating nucleated
birth, and it declines as the larger erythrocytes formed in the erythrocytes has been reported in related English springer
fetus are replaced by smaller erythrocytes produced after spaniels with dyserythropoiesis, polymyopathy, and heart
birth.126,508 disease.223 Microcytosis has been described in a crossbred dog
Microcytic (low-MCV) anemias usually indicate the pres- with persistent elliptocytosis resulting from a lack of erythro-
ence of chronic iron deficiency.204 Microcytic iron-deficiency cyte membrane protein band 4.1. Although the animal was
anemia in adult animals is almost always due to chronic hem- not anemic, the reticulocyte count was about twice normal in
orrhage. Depending on the initial MCV and the magnitude response to a shortened erythrocyte life span.205 Some Japa-
of ongoing blood loss, one or more months are required before nese breeds (Akita and Shiba) normally have MCV values
the MCV decreases below the reference interval. Body iron below the reference intervals established for other breeds of
stores must be depleted and then the microcytes formed must dogs, but they are not anemic.115,179
comprise a high enough percentage of the total erythrocytes Spurious microcytosis may occur when platelets are
present to decrease the MCV below the reference interval. included with erythrocytes in MCV calculations in severely
Microcytic anemia rarely occurs as a result of dietary iron anemic animals or animals with marked thrombocytosis.550
deficiency in adult animals. However, iron deficiency without MCVs may also be spuriously decreased in dogs with persis-
blood loss is common in nursing animals, because milk is low tent hyponatremia because the erythrocytes shrink when they
in iron and these rapidly growing animals have an increased are diluted in vitro with counting fluid prior to sizing in an
demand for iron.204 Although microcytes are often formed in electronic cell counter.58
nursing animals, the MCV may not be reduced in iron-
deficient neonatal dogs and cats because of the persistence of Mean Cell Hemoglobin Concentration
macrocytes formed before birth.508 Even if body iron stores The MCHC represents the average hemoglobin concentra-
have not been depleted, erythrocytes may become microcytic tion within erythrocytes. It is calculated by dividing the
in dogs given long-term recombinant canine erythropoietin whole blood hemoglobin value (in grams per deciliter) by the
HCT (as a percentage) and multiplying by 100. The MCHC dogs with hereditary stomatocytosis because the increased
is expressed as grams per deciliter of erythrocytes. (Note: intracellular water, which occurs in this condition, dilutes the
Hemoglobin values in blood are expressed as grams per deci- hemoglobin concentration within the cells.59,231 MCHCs may
liter of whole blood.) Electronic cell counters calculate the be spuriously decreased in cats and dogs with persistent
MCHC using three measured parameters (RBC count, MCV, hypernatremia because the cells can swell when they are
and hemoglobin concentration); consequently the MCHC diluted with counting fluid prior to analysis in an electronic
can provide a form of quality control for these measured cell counter.58
parameters.
High MCHC values are artifacts. They may result from in Mean Cell Hemoglobin
vivo or in vitro hemolysis, lipemia, the presence of Heinz The mean cell hemoglobin (MCH) is calculated by dividing
bodies within erythrocytes, cryoproteins that precipitate when the hemoglobin value (in grams per deciliter) by the RBC
the sample is cooled, or paraprotein precipitation in the ana- count (in millions of cells per microliter) and multiplying
lyzer diluent.100,297,550 In the case of hemolysis, some hemo- by 10. The MCH provides no added value because it depends
globin is free in plasma; but the formula used to calculate the on the MCV and MCHC. It usually correlates directly with
MCHC assumes that all measured hemoglobin is contained the MCV except in animals with macrocytic hypochromic
within erythrocytes. Lipemia, protein precipitation, and Heinz erythrocytes.
bodies cause turbidity in the spectrophotometric assay for
hemoglobin, thereby giving erroneously elevated hemoglobin Red Cell Distribution Width
values. A high MCHC may also occur if there is agglutination The red cell distribution width (RDW) is an electronic
of erythrocytes when the specimen is assayed in an electronic measure of anisocytosis or erythrocyte volume heterogeneity.
cell counter, as can happen with cold-acting autoantibodies or A histogram of the volume of individual erythrocytes reveals
following heparin therapy in some horses.322,489,550 Large a plot approximating a Gaussian distribution. Consequently
erythrocyte aggregates are too large to be considered erythro- one can calculate the degree of size variation by determining
cytes; consequently cell counters are programmed to exclude the standard deviation (SD) of erythrocyte volumes. However,
them from erythrocyte measurements. This results in errone- the SD depends on the size of the cells as well as the degree
ously low HCT values and consequently erroneously high of size variation around the MCV. To provide a measure of
MCHC values. Agglutination should not interfere with HCT size variation that does not depend on how large the cells are,
values determined by centrifugation as long as the blood the coefficient of variation of erythrocyte volume is calculated
samples are well mixed before the microhematocrit tubes are by dividing the SD by the MCV and then multiplying by 100.
filled. In addition to the standard MCHC calculation, the In short, the RDW is the SD of erythrocyte volumes expressed
Advia 120 (Siemens Healthcare Diagnostics, Inc., Tarrytown, as a percentage of the mean erythrocyte volume.
NY) determines the hemoglobin concentration within indi- Reference values vary depending on the instrument used
vidual erythrocytes based on deflection of light that occurs to measure the RDW. Cattle and horses normally have some-
when a laser beam strikes individual cells. The mean of hemo- what higher RDW values than cats and dogs.505 One need
globin concentrations within erythrocytes determined in this only refer to the upper limit of a reference interval in examin-
manner is calculated and referred to as the cell hemoglobin ing data from a patient, because there is no pathologic state
concentration mean (CHCM). The CHCM provides an accu- in which erythrocytes have greater volume homogeneity
rate measure of mean hemoglobin concentration within (lower RDW) than in the normal state.
erythrocytes even when hemoglobinemia and lipemia are Examination of the RDW has not been extensively utilized
present.297 For example, a kitten with marked lipemia and a in veterinary medicine. It is expected to be increased in cases
microcytic anemia (HCT 12%, MCV 33 fL) had a calculated where the degree of anisocytosis (as estimated on the stained
MCHC of 122 g/dL but a CHCM of 26 g/dL. blood film) is increased. It is often increased in regenerative
MCHC values may be decreased in animals with regenera- anemias because reticulocytes and young erythrocytes are
tive anemia, especially those with high percentages of stress larger than mature erythrocytes.127,338,382 Like the MCV, the
reticulocytes. Hemoglobin synthesis is not complete until late number of large erythrocytes in blood must reach a certain
in reticulocyte maturation. Consequently hemoglobin synthe- level before the RDW of a given patient exceeds the reference
sis is less complete in stress reticulocytes because these cells interval. As an animal responds to anemia and young eryth-
are released from the bone marrow earlier than would occur rocytes become the predominant population, the RDW will
normally.205 begin to decline and may return to the reference interval even
Low MCHC values may also occur in animals with chronic though the MCV is still high.
iron-deficiency anemia. When it is determined using an elec- The RDW is also expected to increase in iron-deficiency
tronic cell counter, the MCHC may be normal in animals anemia, where smaller than normal erythrocytes are produced.
with slight microcytosis, but it is usually low when the MCV As in regenerative anemia, the increase is most likely to be
is markedly reduced.201 The MCHC is low in iron deficiency seen during the phase of disease when a significant number
because there is inadequate iron to support the synthesis of of normally and abnormally sized erythrocytes are present
normal amounts of hemoglobin. Low MCHC values occur in simultaneously.201 In severe chronic iron-deficiency anemia,
the RDW might decrease toward normal once the whole

population of erythrocytes is small. The RDW may increase
again following iron therapy, as normally sized erythrocytes
are produced.
Other potential causes of increased RDW include condi-
tions in which substantial fragmentation of erythrocytes is
occurring and following transfusion of blood from a donor
animal in which the MCV is substantially different from that
of the recipient. The RDW is also increased in dogs with
hereditary stomatocytosis.59 Animals with nonregenerative
anemias will have normal RDW values unless significant dys-
erythropoiesis is present. Spuriously increased RDW values
may occur when agglutination is present or platelets are A Erythrocyte volume
included with erythrocytes in calculations of cell volume dis-
tribution in severely anemic animals.183
Erythrocyte Volume Histograms and

Erythrocyte Cytograms
Although quite useful when abnormal, MCV and MCHC
values are relatively insensitive in identifying the presence of
erythrocytes with abnormal volumes or hemoglobin concen-
trations. Many microcytic or macrocytic erythrocytes are
required to move the MCV below or above the reference
interval, and many hypochromic erythrocytes are needed to
move the MCHC below the reference interval. In addition to
counting cells, electronic cell counters can determine and plot
the volume of individual erythrocytes, and examination of B Hemoglobin concentration
these erythrocyte volume histograms can reveal the presence
of increased numbers of microcytes or macrocytes even when FIGURE 4 -99
the MCV is within the reference interval (Fig. 4-99, A). Some Erythrocyte histograms from a 6-week-old kitten after a blood transfu-
electronic cell counters, such as the Advia 120, can also deter- sion. The kitten presented with marked lipemia and a severe iron defi-
ciency anemia. The animal was obtained from a shelter and was no longer
mine the hemoglobin concentration of individual erythrocytes
nursing. An image from the stained blood film prepared from this cat is
from the deflection of light that occurs when a laser beam shown in Figure 4-24. Most erythrocytes from normal animals are
strikes individual cells. This allows for the generation of expected to be between the vertical lines. A, Erythrocyte volume histo-
hemoglobin concentration histograms (Fig. 4-99, B). Inspec- gram demonstrates a population of small erythrocytes from the patient
tion of hemoglobin concentration histograms can reveal the and a population of large erythrocytes from the blood donor. The MCV
was 47 fL and the RDW was 43%. B, Erythrocyte hemoglobin concen-
presence of increased numbers of hypochromic erythrocytes
tration histogram revealed hypochromic cells but did not demonstrate
even when MCHC has not decreased below the reference two distinct cell populations. Histograms were generated using an Advia
interval. Individual erythrocytes can be further characterized 120 hematology analyzer.
by creating a cytogram in which the erythrocyte volumes of
individual cells are plotted against their respective hemoglobin
concentrations (Fig. 4-100).
Direct Antiglobulin Test

A direct antiglobulin (Coombs’) test is done when autoag- content in blood above 1.5% of total hemoglobin. Clinical
glutination is absent but immune-mediated hemolytic anemia signs associated with methemoglobinemia are the result of
is still suspected. Species-specific antisera against IgG, IgM, tissue hypoxia, because methemoglobin cannot bind O2. Both
and the third component of complement (C3) are used to low blood O2 tension and methemoglobinemia can result in
detect the presence of one or more of these factors on the cyanotic-appearing mucous membranes and dark-colored
surface of erythrocytes.502 This test is discussed in greater blood samples. Hypoxemia is documented by measuring a low
detail in Chapter 6. PO2 in arterial blood (PaO2). Methemoglobinemia is sus-
pected when arterial blood with normal or increased PaO2 is
Methemoglobin Determination dark-colored. Methemoglobin is quantified spectrophoto-
Methemoglobin differs from hemoglobin only in that the iron metrically, but a spot test can be used to determine whether
moiety of heme groups has been oxidized to the ferric (+3) clinically significant levels of methemoglobin are present (see
state. The term methemoglobinemia refers to methemoglobin Fig. 2-4).
in cattle, copper toxicity in sheep and goats, and red maple

toxicity in horses.53,205 These oxidants can also produce Heinz
body hemolytic anemia. Methemoglobinemia without Heinz
Erythrocyte volume bodies or eccentrocytes has been reported in a dog with
hydroxycarbamide (hydroxyurea) toxicity.542 Nitrite produces
methemoglobinemia without Heinz body formation or devel-
opment of anemia. Methemoglobinemia occurs in ruminants
eating nitrate-accumulating plants, especially when those
plants have been fertilized with nitrogenous compounds.
Nitrate is relatively nontoxic, but it is reduced to nitrite by
ruminal microorganisms.205 Nitrite toxicity has been reported
in dogs and cats fed a commercial pet food that had sodium
A Hemoglobin concentration nitrite added as a preservative.541
Cytochrome-b5 Reductase Deficiency

Persistent methemoglobinemia resulting from erythrocyte
Cb5R deficiency has been recognized in many breeds of dogs
and in domestic shorthaired cats.203 Methemoglobin content
Erythrocyte volume
is generally higher in cats (44% to 52%) than in dogs (13% to

51%) with this deficiency because of lower enzyme activity in
deficient cats compared with deficient dogs. Flavin adenine
dinucleotide (FAD) is a cofactor for the Cb5R enzyme, and
persistent methemoglobinemia (26% to 48%) has also been
recognized in horses with Cb5R deficiency secondary to
erythrocyte FAD deficiency.203 Animals with Cb5R deficiency,
in contrast to those with methemoglobinemia produced by
B Hemoglobin concentration
oxidant drugs and compounds, usually exhibit few or no clini-
FIGURE 4-100 cal signs of illness. The diagnosis of this deficiency is made by
Erythrocyte volume versus hemoglobin concentration (V/HC) cyto- measuring enzyme activity within erythrocytes.
grams. Most erythrocytes from normal animals are expected to be within
the square formed by the double vertical and horizontal lines. A, Eryth-
rocyte V/HC cytogram from a 6-week-old kitten with marked lipemia S ER U M I RO N A S S AY S
and a severe iron-deficiency anemia that was a littermate of the cat Serum Iron
presented in Figures 4-24 and 4-99. A population of microcytic hypo-
chromic cells is clearly visible. The HCT was 12%, MCV was 33 fL, Serum iron concentration is generally increased in animals
RDW was 30%, MCHC was 122 g/dL, and CHCM was 26 g/dL. The with hemolytic anemia, dyserythropoiesis, hypoplastic or
MCHC was spuriously increased because of the lipemia. B, Erythrocyte aplastic anemia, iron overload, acute iron toxicity, chronic
V/HC cytogram from a 6-week-old kitten with lipemia and severe iron hepatopathy (dogs), experimental pyridoxine deficiency (pigs),
deficiency that is also presented in Figures 4-24 and 4-99. The kitten was
and following the administration of glucocorticoid steroids to
given a whole-blood transfusion prior to sample analysis. The HCT was
23%, MCV was 47 fL, RDW was 43%, MCHC was 40 g/dL, and dogs and horses.192,205 Serum iron values may be spuriously
CHCM was 27 g/dL. Two populations of erythrocytes are visible. The increased if laboratory tubes or pipettes that are used to handle
kitten’s cells are concentrated in the bottom left area of the cytogram. blood or serum are contaminated with iron.
Erythrocyte V/HC cytograms were generated using an Advia 120 hema- Serum iron concentration is generally low in iron defi-
tology analyzer.
ciency and with inflammation.108,204,236 It is also low in about
half of the dogs with portosystemic shunts.204 Serum iron may
be decreased when demands for erythropoiesis exceed the iron
flow from the diet and storage pools, such as might occur with
Toxic Methemoglobinemia erythropoietin administration or following acute hemor-
Methemoglobinemia results from either increased production rhage.204,356 Serum iron concentration is decreased following
of methemoglobin by oxidants or decreased reduction of met- glucocorticoid administration to cattle and goats.204
hemoglobin associated with a deficiency in the erythrocyte True iron deficiency may be differentiated from other
Cb5R (also called methemoglobin reductase) enzyme.203 causes of hypoferremia by examination of bone marrow for
Experimental studies indicate that many drugs can produce stainable iron, which is minimal or absent in iron deficiency
methemoglobinemia in animals. Significant methemoglobin- and normal or high in other disorders (see Chapter 8).
emia has been associated with clinical cases of benzocaine However, stainable iron is not present in the bone marrow of
toxicity in several species, acetaminophen and phenazopyri- normal cats; consequently a lack of stainable iron does not
dine toxicities in cats, skunk musk in dogs, chlorate toxicity suggest iron deficiency in this species.204
used to measure EPO, but commercial tests developed for

Total Iron-Binding Capacity human assays may not always cross-react sufficiently for use in
The total iron-binding capacity (TIBC) of serum is a measure other species. Consequently individual tests require validation
of serum transferrin concentration because insignificant for each species to be assayed before they can be used clinically.
amounts of circulating iron are bound to other proteins. TIBC Serum EPO is increased in response to various anemias except
is calculated by measuring serum iron and serum unsaturated the anemia of chronic renal disease, in which EPO production
iron-binding capacity and summing these values. Serum is decreased.96,350,361 Serum EPO concentration appears to be
TIBC is low normal or decreased in association with inflam- regulated not only by the rate of renal production but also by
matory disorders and increased in iron-deficient humans, the rate of utilization by erythroid cells. At any given blood
rabbits, pigs, horses, and cattle.204 A slight increase in serum hemoglobin concentration, the serum EPO concentration is
TIBC was reported in an experimental study of diet-induced likely to be highest in disorders with low marrow erythroid
iron-deficiency anemia in young growing dogs,151 but serum activity (e.g., erythroid aplasia).83 The EPO assay has received
TIBC is generally normal in dogs with naturally occurring limited use in veterinary medicine. EPO has been assayed in
iron-deficiency anemia.204 About half of the dogs with porto- serum to assist in differentiating primary erythrocytosis (where
systemic shunts exhibited hypoferremia with normal or EPO values should be normal or low) from secondary eryth-
slightly decreased serum TIBC.204 TIBC may be increased in rocytosis (where EPO values should be high).182,214,248,408
some animals with iron overload and in dogs with chronic Unfortunately there is considerable overlap among patients
hepatopathy.204,441 with primary and secondary erythrocytosis, thus limiting the
diagnostic value of the EPO assay.96
Serum Ferritin
Ferritin is secreted by cells into the blood rather than leaking
from the cytoplasm of damaged cells. Serum ferritin typically D I F F ER EN T I A L D I AG N O S I S
has a much lower iron content than does intracellular ferritin. O F A N EM I A
The macrophage appears to be the primary cell involved in
secreting ferritin into the blood, at least under steady-state True or absolute anemia is defined as a decrease in erythrocyte
conditions, but other cells, including kidney proximal tubular mass within the body. HCT, hemoglobin, and RBC count
cells and hepatocytes, may also secrete ferritin into the values are usually below their reference intervals; however, the
blood.93,486 anemia can sometimes be masked by concomitant dehydra-
Serum ferritin concentration correlates with tissue iron tion. Low erythrocyte parameters may also be present in blood
stores in domestic animals, including cats. Consequently when the total-body erythrocyte mass is normal (relative
serum ferritin concentration can help differentiate true iron anemia). This can result from overhydration resulting in
deficiency (serum ferritin is low) from the anemia of inflam- erythrocyte dilution and from splenic sequestration of eryth-
matory disease (serum ferritin is normal or high).204 Increased rocytes as occurs with splenic relaxation during anesthesia,
serum ferritin occurs in animals with increased storage iron, heparin-induced erythrocyte agglutination in horses, and
as may occur with hemolytic anemia, hemophagocytic histio- various causes of splenomegaly.66,319
cytic sarcoma (in dogs), and repeated blood transfusions. An Anemia is a condition, not a diagnosis. Anemia is classified
increase has also been reported in dogs with inflammation, in various ways to assist in determining its specific cause so
liver disease, and lymphoma.148 It is transiently increased in that effective therapy can be given. In addition to past history,
horses after moderate to severe exercise and in foals following presenting complaints, and laboratory findings, results of
consumption of colostrum. Serum ferritin is an acute-phase other test procedures (e.g., diagnostic imaging) are important
protein; consequently increased values are expected in inflam- in reaching a final diagnosis.
matory conditions in addition to conditions with increased Anemia may occur following blood loss, increased eryth-
iron stores.204 It should be remembered that true iron defi- rocyte destruction, or decreased erythrocyte production.
ciency can be missed if concomitant inflammation is present, Factors that can be useful in categorizing anemia into these
resulting in increased ferritin secretion into blood. Commer- broad causes (and often into more specific causes) include
cial assay kits are not available for serum ferritin assays in reticulocyte counts, erythrocyte indices, erythrocyte morphol-
animals, but ferritin assays may be performed for several ogy on stained blood films, the appearance of the plasma,
species at the Kansas State University College of Veterinary plasma protein concentration, serum iron measurements,
Medicine. serum bilirubin determination, direct antiglobulin test, and
bone marrow evaluation.
Anemia may also develop as a result of the expansion of
ERY T H RO P O I E T I N A S S AY the vascular space faster than the expansion of the total-body
erythrocyte mass. This hemodilution contributes to the anemia
Erythropoietin (EPO) is a glycoprotein hormone that stimu- of the neonate (to be discussed later) and to the mild anemia
lates erythropoiesis in a number of ways. Radioimmunoassays that develops during pregnancy in most domestic animals, the
or enzyme-linked immunosorbent assays (ELISAs) may be horse being an exception.8,41,208,531
development than normally occurs. These large “stress” reticu-

R E G EN ER AT I V E V ER S U S locytes (see Fig. 4-21) apparently remain in the circulation
N O N R E G EN ER AT I V E A N EM I A longer than other reticulocytes before maturation is com-
plete.146 Factors have been utilized in an attempt to correct for
The most useful approach in the classification of anemia is to this longer reticulocyte circulation time in humans, and some
determine whether evidence of a bone marrow response to the veterinary authors have empirically applied these same factors
anemia is present in blood. For all common domestic animals to anemic dogs to calculate what has been called the reticulo-
except the horse, this involves determining whether absolute cyte index. This approach has not been validated in dogs.
reticulocyte numbers are increased in blood. Horses rarely Hemolytic anemia usually elicits a more dramatic regen-
release reticulocytes from the bone marrow even when an erative response than hemorrhagic anemia at least in part due
increased production of erythrocytes occurs. MCV and/or to the greater availability of iron. There are also species differ-
RDW values are increased in some horses responding to anemia, ences in the ability to increase erythrocyte production. The
but others recover from anemia without having these parame- HCT increases about 1 percentage point per day following
ters exceed reference intervals.127,382 Consequently bone marrow experimental phlebotomy in dogs and cats, with a slightly
evaluation is often needed to determine whether an appropriate lower response in cattle and horses.15,65,364
response to anemia is present in a horse. Myeloid to erythroid Anemias with no or minimal increase in blood reticulocyte
(M : E) ratios below 0.5 and bone marrow reticulocyte counts counts are classified as nonregenerative and poorly regenera-
above 5% suggest a regenerative response to anemia.238 tive respectively. The lack of a reticulocyte response in none-
Increased polychromasia is usually present in regenerative quine species generally indicates that the anemia results from
anemias because many reticulocytes stain bluish red with insufficient erythrocyte production in the marrow. A minimal
routine blood stains (see Figs. 4-18, 4-19, 4-21, 4-22). Cats reticulocyte response may be present if the anemia develops
with mild anemias may not release aggregate reticulocytes acutely following hemorrhage or hemolysis because about 4
from the marrow but will release punctate reticulocytes. days are required for a substantial reticulocyte response to
Because punctate reticulocytes do not contain sufficient occur. Mild anemias may have minimally increased reticulo-
numbers of ribosomes within them to impart a bluish color cyte counts.
to the cytoplasm, mild regenerative anemias in cats may lack
polychromasia in stained blood films (see Fig. 4-23, A,B). Classification of Anemia Using Erythrocyte Indices
Increased anisocytosis is often present in regenerative anemias An anemia can also be classified using the MCV and MCHC
because of the presence of large immature erythrocytes (see values to assist in determining its cause. The terms used to
Figs. 4-18, 4-19, 4-21), although anisocytosis may be marked indicate size are macrocytic (increased MCV), normocytic
in some nonregenerative anemias as well. (normal MCV), and microcytic (decreased MCV). The terms
Except in horses, some nucleated erythrocytes (rubricytes used to describe MCHC values are normochromic (normal
and metarubricytes) are often seen on blood films in associa- MCHC) and hypochromic (decreased MCHC). Anemias are
tion with regenerative anemia; however, nucleated erythro- not classified as hyperchromic because high MCHC values
cytes may also be present in anemic and nonanemic disorders are artifacts. A comparison of erythrocyte indices and causes
with minimal or no reticulocytosis (see Fig. 4-75, C-E). There- of anemia is given in Box 4-2.
fore the presence of nucleated erythrocytes in blood is a much
less reliable indicator of a regenerative response to anemia
than is an increased reticulocyte count. H EM O LY T I C A N EM I A S
Howell-Jolly bodies are often present within erythrocytes
in regenerative anemias, but they also occur in normal cats Hemolytic anemias occur as a result of increased erythrocyte
(see Fig. 4-76, A) and horses and in splenectomized animals destruction within the body. Causes of hemolytic anemia in
of other species. Basophilic stippling occurs in regenerative animals are given in Box 4-3. Erythrocytes may be lysed
anemias in ruminants (see Fig. 4-81, A,B) but rarely in other within the circulation (intravascular hemolysis), but more fre-
species (see Fig. 4-81, C). Basophilic stippling can also occur quently they are lysed following phagocytosis by cells of the
in erythrocytes of any species with lead toxicity whether or mononuclear phagocyte system (extravascular hemolysis).
not anemia is present (see Fig. 4-81, D). Hemolytic anemias are generally regenerative if sufficient
The presence of compensatory reticulocytosis indicates that time has elapsed for a bone marrow response to the anemia.
the anemia has resulted from either blood loss or increased They are initially normocytic normochromic but may be mac-
erythrocyte destruction. Several factors should be kept in mind rocytic hypochromic or macrocytic normochromic if sufficient
in interpreting the magnitude of a reticulocyte response. In time has elapsed for the release of a significant number of
regenerative anemias, animals with lower HCTs should large reticulocytes from the bone marrow. Macrocytic hypo-
have higher absolute reticulocyte counts. Severe anemia evokes chromic erythrocytes may also occur in hereditary stomatocy-
a greater stimulus for increased erythrocyte production tosis in dogs as a result of membrane abnormalities and
than does mild anemia.351 Also, in response to severe anemia, erythrocyte swelling.205 An example of a hemolytic anemia
reticulocytes can be released from the marrow earlier in their that is usually nonregenerative and normocytic normochromic
Box 4-2 Comparison of Classification of Anemias by Erythrocyte Indices and Etiology
Normocytic Normochromic 4. Nonregenerative immune-mediated anemia and/or

1. Hemolytic anemia if reticulocyte response is mild or if myelofibrosis in dogs
sufficient time has not elapsed for a prominent reticulocyte 5. Poodle macrocytosis (healthy miniature poodles with no
response to occur. anemia)
2. Hemorrhage if reticulocyte response is mild or if sufficient 6. Hyperthyroid cats (slight macrocytosis without anemia)
time has not elapsed for a prominent reticulocyte response to 7. Folate deficiency (rare)
occur. 8. Congenital dyserythropoiesis of Hereford calves
3. Early iron-deficiency anemia before microcytes predominate 9. Spurious with erythrocyte agglutination
4. Chronic inflammation and neoplasia (sometimes slightly 10. Spurious in cats and dogs with persistent hypernatremia (may
microcytic) be hypochromic)
5. Chronic renal disease
6. Endocrine deficiencies Microcytic Normochromic/Hypochromica
7. Selective erythroid aplasia 1. Chronic iron deficiency (months in adults, weeks in nursing
8. Aplastic and hypoplastic bone marrows animals)
9. Lead toxicity (may not be anemic) 2. Portosystemic shunts in dogs and cats (often not anemic)
10. Cobalamin deficiency 3. Anemia of inflammatory disease (usually normocytic)
4. Hepatic lipidosis in cats (usually normocytic)
Macrocytic Hypochromic 5. Normal Akita and Shiba dogs (not anemic)
1. Regenerative anemias with marked reticulocytosis 6. Prolonged recombinant erythropoietin treatment (mild)
2. Hereditary stomatocytosis in dogs (often slight 7. Copper deficiency (rare)
reticulocytosis) 8. Drugs or compounds that inhibit heme synthesis
3. Abyssinian and Somali cats with increased erythrocyte 9. Myeloid neoplasms with abnormal iron metabolism (rare)
osmotic fragility (a reticulocytosis is usually present) 10. Pyridoxine deficiency (experimental)
4. Spurious with prolonged storage of blood sample 11. Familial dyserythropoiesis of English springer spaniel dogs
(rare)
Macrocytic Normochromic 12. Hereditary elliptocytosis in dogs (rare)
1. Regenerative anemias (decreased MCHC is not always 13. Spurious when platelets are included in erythrocyte
present) histograms
2. FeLV infections with no reticulocytosis (common) 14. Spurious in dogs with persistent hyponatremia (not typically
3. Erythroleukemia (AML-M6) and myelodysplastic syndromes anemic)
a
The presence of low MCHC along with low MCV strongly suggests iron-deficiency anemia.
is cytauxzoonosis in cats. Most cats die before there is time

for a regenerative response to the anemia to occur.185 Increased
erythrocyte phagocytosis occurs in animals with hemophago-
cytic syndrome (macrophage activation syndrome), but the
anemia may not be regenerative because the associated release
of inflammatory mediators interferes with normal
erythropoiesis.499,516
An increase in the plasma bilirubin concentration imparts
a yellow color to the plasma (see Fig. 2-9, B). Mucous mem-
branes and skin may also appear yellow (icteric) in extreme
cases of hyperbilirubinemia (Fig. 4-101). Hyperbilirubinemia
associated with a substantial decrease in the HCT suggests
increased phagocytosis of erythrocytes.
If substantial intravascular hemolysis occurs rapidly,
hemoglobinemia (see Fig. 2-9, C, D) and subsequently hemo-
globinuria may be observed. Disorders where significant
intravascular hemolysis sometimes occurs include immune- FIGURE 4 -101
mediated hemolytic anemia, oxidant chemical and plant tox- Icteric mucous membranes in an English springer spaniel dog with a
icities, severe hypophosphatemia, leptospiral and clostridial hemolytic crises associated with PFK deficiency.
infections, coral snake and rattlesnake envenomation, zinc
toxicity, copper toxicity, severe babesiosis, hypo-osmolality,
Box 4-3 Causes of Hemolytic Anemias in Domestic Animals
1. Immune-mediated erythrocyte destruction: Primary or in horses), crude oil (marine birds), venoms (snakes, bees,
autoimmune hemolytic anemia (common in dogs); neonatal wasps, and spiders)
isoerythrolysis (primarily horses and cats); lupus erythematosus 5. Fragmentation: Disseminated intravascular coagulation
(primarily dogs); incompatible blood transfusions; drugs, (primarily dogs), dirofilariasis (especially posterior vena cava
including penicillin (horses), cephalosporins (dogs), levamisole syndrome) in dogs, hemangiosarcoma (dogs), vasculitis,
(dogs), sulfonamides (horses and dogs), pirimicarb (insecticide hemolytic uremia syndrome
in dogs), and propylthiouracil (cats) 6. Hypo-osmolality: Hypotonic fluid administration, water
2. Erythrocyte parasites (may have an immune-mediated intoxication (primarily in cattle)
component): Anaplasma spp. (ruminants), erythrocytic 7. Hypophosphatemia: Postparturient hemoglobinuria (cattle),
Mycoplasma spp. (except horses), Babesia spp., Cytauxzoon ketoacidotic diabetic animals following insulin therapy (cats
felis, Theileria spp. (ruminants) and dogs), hepatic lipidosis (cats)
3. Other infectious agents (may have an immune-mediated 8. Hereditary erythrocyte defects: Pyruvate kinase deficiency
component): Leptospira and Clostridium spp. (primarily (dogs and cats), phosphofructokinase deficiency (dogs),
ruminants and horses), FeLV (seldom hemolytic), glucose-6-phosphate dehydrogenase deficiency (horses),
equine infectious anemia virus (multifactorial, also hereditary stomatocytosis (mild anemia in dogs), erythropoietic
with decreased production), Sarcocystis spp. (cattle porphyria (cattle and cats), hereditary nonspherocytic
and sheep), Trypanosoma spp. (primarily outside the hemolytic anemias of unknown etiology (poodle and beagle
United States) dogs), idiopathic increased erythrocyte osmotic fragility (cats),
4. Chemicals and plants (most are oxidants): Onions, red maple erythrocyte flavin adenine dinucleotide deficiency in horses
(horses), Brassica spp. (ruminants), lush winter rye (cattle), (methemoglobinemia and sometimes mild anemia), hereditary
copper (sheep and goats), phenothiazine (horses), spherocytosis in cattle
acetaminophen (cats and dogs), methylene blue (cats and 9. Miscellaneous: Liver failure (horses), hypersplenism,
dogs), benzocaine (cats and dogs), phenazopyridine (cats), hemophagocytic histiocytic sarcoma, splenic torsion (dogs),
methionine (cats), vitamin K (dogs), propylene glycol (cats), selenium deficiency in cattle grazing on St. Augustine grass,
naphthalene (dogs?), zinc (dogs and ruminants), indole postparturient hemoglobinuria in cattle not associated with
(experimental in cattle and horses), tryptophan (experimental hypophosphatemia
caudal vena cava syndrome of dirofilariasis in dogs, hepatic primary, disorder. Primary IMHA may be associated with
failure in horses, phosphofructokinase deficiency in dogs, immune-mediated thrombocytopenia; it may also be part of
inherited idiopathic increased erythrocyte osmotic fragility in systemic lupus erythematosus, a multisystemic autoimmune
Abyssinian and Somali cats, postparturient cattle without disease.167,480
hypophosphatemia, and splenic torsion in dogs. In most of
these disorders, however, erythrocyte destruction occurs pri- Neonatal Isoerythrolysis
marily by increased phagocytosis. Neonatal isoerythrolysis is an IMHA that develops in neona-
tal animals following ingestion of colostrum containing anti-
Immune-Mediated Hemolytic Anemia bodies against antigens on their erythrocytes.480 It occurs
The binding of antibodies and/or complement to erythrocyte primarily in horses, mules, and cats. In horses, dams become
surfaces can result in phagocytosis by macrophages and, sensitized to foreign erythrocyte antigens from leakage of fetal
in some cases, intravascular hemolysis. Immune-mediated erythrocytes through the placenta during pregnancy or from
hemolytic anemia (IMHA) may be primary or it may occur exposure to fetal erythrocytes of the same blood type during
secondarily to rickettsial, bacterial, protozoal, viral, or hemo- a previous parturition. Antibodies are produced against these
tropic mycoplasmal infections; neoplasia (especially lympho- antigens and secreted in colostrum.61,371 Neonatal isoeryth-
mas); and toxin or drug exposure. Additional information rolysis can occur in kittens with blood type A born to queens
concerning IMHA is given in Chapter 6. with blood type B who have had no prior exposure to blood
type A antigens, because all adult cats with type B blood
Primary IMHA naturally have high anti-A antibody titers.171
A diagnosis of primary IMHA, also called autoimmune
hemolytic anemia, is reached by ruling out other disorders Transfusion Reactions
known to have concomitant IMHA. Primary IMHA is Hemolytic transfusion reactions may occur when plasma of
common in dogs,33 less common in cats,261 and rare in other the recipient contains antibodies against one or more antigens
domestic animal species. About two-thirds of the dogs with on the surface of donor erythrocytes. Erythrocyte destruction
IMHA appear to have primary IMHA.394 In contrast, IMHA may also occur when plasma of the donor contains antibodies
in noncanine species is usually a secondary rather than a against one or more antigens on the surface of recipient
erythrocytes, but the amount of antibody present to react with B. canis canis occurs primarily in Europe and Asia and is of
erythrocytes is considerably less. With the exception of cats, intermediate pathogenicity.467 In addition, an unnamed large
naturally occurring antierythrocyte antibodies of clinical sig- Babesia species has been described in the United States in
nificance seldom occur in animals. Rather, antibody formation immunosuppressed dogs.51,226,427
results from prior exposure to different erythrocyte antigens B. gibsoni is a small Babesia species infecting dogs; it is
via transfusion, pregnancy, or vaccination with products con- endemic in Africa, the Middle East, and Asia. There has been
taining blood group antigens. Consequently severe hemolytic a rapid increase in the number of cases reported in various
transfusion reactions generally do not occur at the time of the parts of the United States, predominantly in American pit bull
first blood transfusion. Cats with blood type B have naturally terriers.547 Although ticks are considered the primary vector
occurring anti-A antibodies with high hemolytic titers. There- in much of the world, it appears that transmission of infected
fore the transfusion of type A blood into a type B cat can blood through dog bites is a major mechanism of transmission
result in a life-threatening intravascular hemolytic reaction of this organism in the United States.49 A second small Babesia
the first time such a transfusion is given. species called Babesia conradae has been recognized in Cali-
fornia. Like B. gibsoni, it causes severe clinical disease in adult
dogs.252 It is closely related to piroplasms isolated from wild-
Erythrocyte Parasites life and humans in the western United States. A third small
Erythrocyte parasites include intracellular protozoal parasites piroplasm has been identified causing hemolytic anemia in
(Babesia species, Theileria species, and Cytauxzoon felis), intra- dogs; it is closely related to Babesia microti, a previously rec-
cellular rickettsial organisms (Anaplasma species), and epicel- ognized parasite of rodents and humans. A provisional name
lular Mycoplasma species. The morphology of these organisms of Theileria annae has been assigned to this organism based
is presented earlier in this chapter. These infectious agents on its 18S rRNA gene sequence.79
generally cause mild to severe hemolytic anemia, depending B. felis causes hemolytic anemia in cats in South Africa.362
on the pathogenicity of the organism and the susceptibility of Additional piroplasms have been identified by PCR in
the host. Some damage to erythrocytes is caused directly by a domestic cats, but their clinical significance remains to be
parasite, but secondary immune-mediated injury may be more determined.491
important in the pathogenesis of anemia in some cases.314 The two most important Babesia species infecting cattle are
Antierythrocyte antibodies are often present and spherocytes B. bigemina and B. bovis. Both are capable of causing life-
may be seen in stained blood films; therefore it is important threatening hemolytic anemia.176 The two most important
to differentiate these infectious diseases from primary piroplasms infecting horses are the large B. caballi parasite,
IMHA.82,176 Macrophage activation with increased erythro- and the smaller, more pleomorphic B. equi parasite.491 The
phagocytosis may also contribute to the development of anemia is generally more severe in B. equi-infected than in
anemia.330 B. caballi-infected horses.78
Except in horses, reticulocytosis is generally present if suf-
ficient time (about 4 days) has elapsed for a bone marrow Theileria Species
response to occur. The anemia may be nonregenerative if a Theileria parva and Theileria annulata cause important diseases
concomitant inflammatory response inhibits erythropoiesis in ruminants in Africa, Asia, and the Middle East.274 The
(see “Anemia of Inflammatory Disease,” below), and/or if the associated anemia may be regenerative or nonregenera-
infection results in decreased hematopoietic precursors, as tive.353,430 Theilerial organisms present in domestic ruminants
reported with Theileria parva infection in cattle.305 Thrombo- in the United States are usually nonpathogenic.245,453 Like-
cytopenia is usually present with protozoal infections of wise, Theileria in deer are generally considered of low patho-
erythrocytes. Platelet consumption may occur in association genicity, but may cause hemolytic anemia under some
with DIC in severe disorders,185,467 but thrombocytopenia is circumstances.544 Some investigators have proposed that B.
probably more often associated with increased phagocytosis equi and B. microti be moved to the genus Theileria, but others
of platelets in response to antibodies on their surfaces and/or believe that they belong in a new genus that is distinct from
because of macrophage activation by inflammatory cytokines both Babesia and Theileria.14,491
such as M-CSF and IFN-γ.2,159,331 Thrombocytopenia is not
generally present with erythrocytic Anaplasma and hemo- Cytauxzoon felis
plasma infections. Like Theileria, the genus Cytauxzoon has both a tissue phase
and an erythrocyte phase. In contrast to Theileria, the schiz-
Babesia Species onts of Cytauxzoon develop exclusively in macrophages
Three genetically distinct large B. canis subspecies (B. canis rather than in lymphocytes.459 Most domestic cats with acute
canis, B. canis vogeli, and B. canis rossi) have been identified. cytauxzoonosis die.185 Cats generally have icteric plasma
B. canis vogeli in the United States generally causes mild or in the terminal stage of the disease. The HCT may be in
inapparent disease in adults (unless they are immunosup- the low thirties but is usually in the twenties. Reticulocyte
pressed) but severe hemolytic anemia in pups. B. canis rossi in counts are not increased in response to the anemia. Cats
South Africa can cause severe disease and death in adult dogs. become thrombocytopenic during the late stage of disease.
Coagulation tests may be prolonged or remain normal, tests been associated with growth retardation and mild anemia in
for fibrin degradation products (FDPs) may be positive, and feeder pigs and unthriftiness with poor reproductive perfor-
the total serum protein concentration is variably decreased. mance in sows.221,397
White blood cell counts are variable, but leukopenia generally M. ovis generally causes hemolytic anemia in young sheep
develops terminally. Parasitemia occurs late in the disease. and goats, with mild or unapparent disease in adults.314,339
Since domestic cats with acute cytauxzoonosis generally die However, severe hemolytic anemia and death have been
in a matter of days, they are believed to be dead-end hosts. described in an outbreak in adult sheep that were also infected
Bobcats, however, usually do not die when infected with C. with A. ovis.227
felis and serve as a reservoir of infection for transmission to Systemic illness with edema has been reported in cattle
domestic cats by ticks. A low percentage of cats have been infected with M. wenyoni435,530; however, this syndrome could
reported to survive.190 This may, in large part, be because of not be reproduced experimentally.529 M. wenyoni is not
infection with a less virulent strain of the organism; however, reported to produce significant anemia in cattle unless they
studies are needed to determine if these cats can serve as a have been splenectomized.
reservoir for C. felis transmission that would result in disease Variable anemia has been reported in llamas and alpacas
in other domestic cats.72 infected with Candidatus M. haemolamae, but most infections
appear to be subclinical. Camelids that are immune-
Anaplasma Species compromised or have other concurrent disorders are more
Anaplasma marginale is an important pathogen causing mild likely to be ill and anemic.483 The anemia may not be regenera-
to severe hemolytic anemia and sometimes death in naive tive if it is complicated by inflammatory disease.390
adult cattle.255 Infected calves generally do not become ill but
become carriers of the organism. Anaplasma centrale is much
less pathogenic in cattle and has been used as a vaccine against Other Infectious Agents
A. marginale.159,313 Anaplasma ovis is pathogenic for sheep, In addition to erythrocyte parasites, infections with other
goats, and some wild ruminants.112 agents may result in hemolytic anemia. As in the case of
erythrocyte parasites, the enhanced erythrocyte destruction
Hemotropic Mycoplasmas (Hemoplasmas) may have an immune-mediated component. Leptospira species
Hemotropic mycoplasmas are bacteria that attach to the have been reported to cause hemolytic anemia in cattle, sheep,
external surfaces of erythrocytes. In contrast to erythrocyte and pigs.47,114,451 Clostridium species have been reported to
protozoal parasites, thrombocytopenia is not a feature of cause hemolytic anemia in cattle,469 sheep,307,385 and horses.393,526
hemoplasma infections. Three different hemoplasmas have Hemolytic anemia has been described in experimental
been identified in cats based on 16S rRNA gene sequences.473,539 acute Sarcocystis infections in cattle,147,293 goats,124 and pigs.36
Mycoplasma haemofelis generally produces anemia and clinical Several Trypanosoma species cause hemolytic anemia in cattle,
signs of disease, while Candidatus Mycoplasma haemominutum sheep, and goats in tropical and subtropical regions, except in
generally results in unapparent infection and minimal change the United States, even though the parasite is not directly
in HCT unless the infection is complicated by other disorders associated with erythrocytes.23,52,230 Nonpathogenic trypano-
such as FeLV infection, feline immunodeficiency virus (FIV) somes, including T. theileri, occur worldwide in cattle.134,413
infection, and neoplasia.113,141,165,242,533 Candidatus Mycoplasma The anemia associated with equine infectious anemia
turicensis can be pathogenic in cats, causing a moderate to (EIA) virus infection is multifactorial. It includes both an
severe hemolytic anemia. Early reports suggest that the immune-mediated destruction of erythrocytes and bone
clinical signs of infection are worsened in cats having concur- marrow suppression.308,417,461 Macrophages within the marrow
rent FIV infections or following the administration of are important components of the erythropoietic microenvi-
corticosteroids.538 ronment, where they can produce erythropoietic stimulatory
Three hemoplasmas have been reported in dogs based on and inhibitory factors (see Chapter 3). The EIA virus repli-
16S rRNA gene sequences. Mycoplasma haemocanis is closely cates in macrophages, including those in the bone marrow,462
related to M. haemofelis.48,67 A hemoplasma closely related to and macrophages produce proinflammatory cytokines in
Candidatus M. haemominutum has been identified in a dog in response to EIA infection that may inhibit both erythrocyte
China.551 Another hemoplasma has been identified in mul- and platelet production.283,461,484
tiple countries and classified as Candidatus Mycoplasma A transient regenerative anemia was reported in some
haematoparvum.35,465,539 Splenectomy, splenic pathology, or kittens experimentally infected with FeLV,290 and immune-
immunosuppression is generally required before hemoplasmas mediated hemolytic anemia has infrequently been reported
are recognized in stained blood films, and hemolytic anemia in naturally infected cats,261 although some of these animals
develops in dogs.202 might have had concomitant hemoplasma infections.165,174
Mycoplasma suis causes hemolytic anemia with icterus in However, the anemia seen in FeLV-infected cats is typically
young piglets. Hemolytic anemia may also occur in pigs at nonregenerative because it is secondary to bone marrow pro-
weaning, in feeder pigs under stress, and in pregnant sows liferative abnormalities, discussed later in this chapter and in
immediately prepartum.314 However, M. suis has more often Chapter 9.174,421
Chemicals and Plants Drugs Causing Immune-Mediated Anemia

Oxidants A number of drugs are reported to produce secondary IMHA
Most chemicals and plants that cause hemolytic anemia are in animals. These drugs include penicillin (horses),306,398 ceph-
oxidants. Consequently Heinz bodies, eccentrocytes, and/or alosporins (dogs),54 levamisole (dogs),28 sulfonamides (horses
methemoglobinemia may be present. Dietary causes of Heinz and dogs),478,487 pirimicarb (insecticide in dogs),235 and pro-
body hemolytic anemia (with or without eccentrocytosis and/ pylthiouracil (cats).29
or methemoglobinemia) include consumption of onions and
garlic in small and large animals, consumption of kale and Fragmentation
other Brassica species by ruminants, consumption of lush Microangiopathic hemolytic anemia may occur when eryth-
winter rye by cattle, and consumption of red maple leaves by rocytes are forced to flow through altered vascular channels
horses and alpacas.* Heinz bodies have been recognized in or exposed to turbulent blood flow. Erythrocyte fragments
erythrocytes from selenium-deficient Florida cattle grazing with pointed extremities are called schistocytes. Erythrocyte
on St. Augustine grass pastures and in postparturient New fragmentation may be seen in animals (especially dogs) with
Zealand cattle grazing primarily on perennial ryegrass.205 DIC. Mechanical fragmentation occurs as the cells pass
Copper toxicity results in Heinz body formation in sheep and through the fibrin meshwork of a microthrombus (see Fig.
goats.39 Heinz body formation has been reported in dogs 4-53). Fragmentation anemia is especially common in dogs
ingesting zinc-containing objects (e.g., U.S. pennies minted with hemangiosarcoma and in dogs with caudal vena cava
after 1982).46,205 Naphthalene ingestion may have caused syndrome, resulting from a rapid blockage of the posterior
Heinz body formation in a dog.120 vena cava with large numbers of adult heartworms.197,392,456
Heinz body hemolytic anemia has occurred following the Erythrocyte fragmentation is a component of the
administration of a variety of drugs including acetaminophen hemolytic-uremic syndrome. This rare syndrome is character-
and methylene blue in cats and dogs, methionine and phenazo- ized by hemolytic anemia, thrombocytopenia, and acute renal
pyridine in cats, menadione (vitamin K3) in dogs, and pheno- failure.345 In humans (and probably animals), the syndrome is
thiazine in horses. The application of benzocaine to inflamed usually initiated by certain toxins produced by Escherichia coli
dog skin can result in Heinz body formation, but methemo- (and less often other bacteria) infecting the gastrointestinal
globinemia is more prominent.205 tract; however, atypical cases have also been associated with
Heinz body hemolytic anemia has been reported in a dog nongastrointestinal infections, cancer, parturition, immuno-
that had been sprayed with skunk musk.549 The ingestion of suppressive drug therapy, organ transplants, and autoimmune
crude oil by marine birds results in Heinz body hemolytic diseases.26,118,224,345 Absorbed bacterial toxins can result in
anemia.278,488 endothelial injury, activation of hemostasis, the formation of
Heinz bodies were consistently present in erythrocytes microthrombi, erythrocyte fragmentation, and reduced blood
from a male English bulldog with multiple erythroid abnor- flow with injury to the kidney and other affected organs.345
malities including prominent siderotic inclusions in his eryth- Atypical cases have been attributed to the overactivation of
rocytes.209 A prominent Heinz body hemolytic anemia (see the alternative complement pathway. The glomerular capillary
Fig. 4-80, A,B), with an HCT of 14% to 24% and a mild bed of the kidney may be at increased risk because of its
methemoglobinemia (about 13%), persisted in a rescued fenestrated endothelium, which continually exposes the sub-
quarter horse colt during 3 months of study. A source of endothelial matrix to a variety of circulating proteins and
dietary oxidants was not identified, and hemoglobin electro- peptides.346 Additional disorders where erythrocyte fragmen-
phoresis appeared normal. Erythrocyte reduced glutathione tation may be observed are described under “Schistocytes,”
concentration was below the reference interval. An erythro- above.
cyte metabolic defect was suspected, but erythrocyte enzyme
assays failed to identify a cause. ( J. W. Harvey, unpublished). Hypo-osmolality
Intravascular hemolysis can occur in calves that drink exces-
Venoms sive amounts of water following a period of water depriva-
Venoms from snakes, bees, wasps, and brown recluse spiders tion.172 Water intoxication decreases plasma osmolality, and
are complex mixtures of proteins, peptides, enzymes, and water moves into erythrocytes causing them to swell and lyse.
chemicals that have multiple pathologic effects, including Hemolysis has been reported in juvenile pygmy goats fed
hemolytic anemia and other forms of tissue injury.301,303,344,431 water using a nipple bottle for human infants,316 and hemo-
Phospholipase enzymes (especially phospholipase A2) in lysis may occur when hypotonic fluid is administered intra
venoms appear to be important in causing erythrocyte venously (see Fig. 4-72, A).
injury.396,407,431,498
Hypophosphatemia
Hemolytic anemia can occur secondary to hypophosphatemia
because hypophosphatemia decreases the erythrocyte glyco-
lytic rate, which results in a decreased ATP concentration.
*References 18, 27, 121, 205, 470, 545. Hemolytic anemia resulting from hypophosphatemia has
been reported in diabetic cats and in a diabetic dog following anemias and sporadic episodes of intravascular hemolysis with
insulin therapy, in a cat with hepatic lipidosis, and in post hemoglobinuria. HCTs are generally between 30% and 40%,
parturient cattle and buffaloes.109,205 In addition to having except during hemolytic crises, when the HCT may decrease
low ATP concentrations, dog erythrocytes might hemolyze to 15% or less. MCVs are usually between 80 and 90 fL and
as a result of decreased erythrocyte 2,3DPG concentration, reticulocyte counts are generally between 10% and 30% even
because dog erythrocytes with low 2,3DPG are more alkaline- when the HCT is within the reference range. Lethargy, weak-
fragile than those of normal dogs and may hemolyze at physi- ness, pale or icteric mucous membranes, mild hepatospleno-
ologic pH values.205 megaly, muscle wasting, and fever as high as 41oC may
occur during hemolytic crises, which occur secondary to
Hereditary Erythrocyte Defects hyperventilation-induced alkalemia.
Pyruvate Kinase Deficiency in Dogs and Cats Affected dogs appear to tire more easily than normal, and
Pyruvate kinase (PK) deficiency is transmitted as an autoso- a myopathy with cramping is infrequently observed. Progres-
mal recessive trait in many breeds of dogs, with highest preva- sive cardiac disease was reported in two whippets. In contrast
lence reported in basenji and beagle dogs. Homozygously to PK deficiency, myelofibrosis and liver failure have not been
affected animals have decreased exercise tolerance, pale recognized in dogs with PFK deficiency.
mucous membranes, tachycardia, and splenomegaly. Affected Homozygous affected animals more than 3 months of age
animals have a macrocytic hypochromic anemia (HCT 16% can easily be identified by measuring erythrocyte PFK activity.
to 28%) with marked reticulocytosis (15% to 50% uncorrected A DNA test using PCR technology has been developed that
reticulocyte count) when young. Myelofibrosis and osteoscle- can clearly differentiate normal, carrier, and affected dogs of
rosis develop in the bone marrow, and hemochromatosis and all breeds except wachtelhunds, which have a different genetic
cirrhosis develop in the liver as the dogs age. HCT and reticu- mutation.205
locyte counts decrease as myelofibrosis and osteosclerosis
become severe. Affected dogs generally die between 1 and 5
years of age because of bone marrow failure and/or liver Increased Erythrocyte Osmotic Fragility in Cats
failure.169,203 A hemolytic anemia with markedly increased osmotic fragil-
PK deficiency has been reported in Abyssinian, Somali, ity occurs in Abyssinian and Somali cats.260 Splenomegaly and
and domestic shorthaired cats. Affected cats are often asymp- polyclonal hyperglobulinemia are common. The HCT is gen-
tomatic, but lethargy, pale mucous membranes, and erally between 15% and 25%, but values as low as 5% have
inappetence may be recognized. The HCT is generally been recognized. A macrocytosis with mild to moderate retic-
normal or mildly decreased, but severe anemia may occur ulocytosis is present in most cats. Most samples exhibit
during intermittent hemolytic crises. The MCV is usually extreme hemolysis after 1 day of refrigeration; however, in
mildly increased. The MCHC is sometimes decreased. An vivo intravascular hemolysis also occurs, as evidenced by
aggregate reticulocytosis is present in 90% of cases.259 Sple- hemoglobinuria in some cats. An erythrocyte membrane
nectomy may reduce the severity of the anemia in cats. In defect is suspected.
contrast to dogs, bone marrow and liver failure have not been
reported in cats; consequently the life expectancy in cats with
PK deficiency is generally longer than that in dogs with this Hereditary Spherocytosis in Cattle
defect.169 Severe hemolytic anemia with icterus and splenomegaly is
PK deficiency can be diagnosed by measuring erythrocyte present shortly after birth in Japanese black cattle that lack
enzyme activity in affected cats, but it can be difficult to band 3 in their erythrocyte membranes. The mortality rate is
diagnose by measuring enzyme activities in affected dogs high in affected animals, especially during the first week of
because their erythrocytes contain an unstable M2 isozyme life. Those that survive exhibit a persistent mild hemolytic
that is usually lost as erythroid precursors develop into eryth- anemia (HCT 25% to 35%), with marked spherocytosis and
rocytes.206 All PK-deficient cats identified thus far have had anisocytosis but no reticulocytosis. This defect is inherited as
the same mutation; therefore a single DNA-based diagnostic an autosomal dominant trait, and osmotic fragility is increased
test may be used to identify deficient cats.259 Unfortunately in both heterozygous and homozygous cattle.231
several different genetic mutations have been identified in
dogs with PK deficiency. Consequently different DNA-based
diagnostic assays must be developed and/or validated for each Glucose-6-Phosphate Dehydrogenase Deficiency in a Horse
affected dog breed.169 Persistent hemolytic anemia and hyperbilirubinemia have
been described in an American standard-bred colt with severe
Phosphofructokinase Deficiency in Dogs G6PD activity. Morphologic abnormalities of erythrocytes
Autosomal recessive inherited erythrocyte PFK deficiency included eccentrocytosis, pyknocytosis, increased anisocytosis,
occurs in English springer spaniel, American cocker spaniel, increased Howell-Jolly bodies, and rare hemoglobin crystals.
mixed-breed, wachtelhund, and whippet dogs.166,205 Homozy- Heinz bodies were not observed in erythrocytes stained with
gously affected dogs have persistent compensated hemolytic new methylene blue.452
Erythrocyte Flavin Adenine Dinucleotide Deficiency Postparturient Hemoglobinuria in Dairy Cattle

in Horses Postparturient hemoglobinuria in dairy cattle has been associ-
Eccentrocytosis, pyknocytosis, and variable numbers of hemo- ated with hypophosphatemia.450 The anemia appears to
globin crystals have also been seen in FAD-deficient horses, develop because affected animals have decreased erythrocyte
but HCTs were normal or only slightly decreased. Affected ATP concentrations.205 An apparently different syndrome of
horses have undetectable glutathione reductase and reduced postparturient hemoglobinuria has been reported in cattle in
Cb5R enzyme activities because FAD is a cofactor for these New Zealand grazing primarily on perennial ryegrass (Lolium
enzymes. A persistent methemoglobinemia (25% to 46%) is perenne). The presence of Heinz bodies in these animals indi-
present because of the Cb5R deficiency.203 cates an oxidant etiology. Postparturient cattle may be more
susceptible to the development of anemia because increased
food consumption associated with lactation could increase
Hereditary Stomatocytosis in Dogs exposure to an unidentified dietary oxidant. Copper deficiency
Stomatocytosis is recognized in association with three differ- may contribute to the severity of the anemia in these cattle by
ent inherited syndromes in dogs that result in one or more rendering their erythrocytes more susceptible to oxidants.292
membrane defects causing erythrocytes to swell. Hemoglobin
values and erythrocyte counts are low-normal or slightly
reduced, but HCTs are normal in Malamutes, schnauzers, and B L O O D - L O S S A N EM I A S
Pomeranians. The MCV is increased and MCHC decreased
even though reticulocyte counts are normal or only slightly Causes of blood-loss anemia are given in Box 4-4.* In some
increased. Affected Drentse patrijshond dogs have lower cases the diagnosis of blood-loss anemia and its cause is
HCTs and higher reticulocyte counts than those found in the apparent from the history and/or physical findings. In other
other breeds. Erythrocytes from all breeds have increased
osmotic fragility and shortened erythrocyte survival.205
*References 6, 107, 145, 161, 168, 284, 298, 312, 380, 414, 444.
Additional Hereditary Defects

Familial nonspherocytic hemolytic anemia has been reported
in poodles. Despite extensive studies, the defect in this disor-
der could not be determined, but PK deficiency cannot be Box 4-4
ruled out. A mild hemolytic anemia with reticulocytosis, Causes of Blood-Loss Anemias in
slightly increased erythrocyte osmotic fragility, shortened Domestic Animals
erythrocyte life span, and normal erythrocyte morphology has
1. Trauma: Accidents, fights, gastrointestinal foreign bodies,
been reported in beagle dogs. A membrane defect was sus-
surgery
pected. Persistent elliptocytosis and microcytosis have been 2. Parasites: Hookworms, fleas, blood-sucking lice,
described in a crossbred dog that lacked erythrocyte mem- Haemonchus spp. (small ruminants), liver flukes,
brane protein 4.1. Although the animal was not anemic, the Coccidia spp.
reticulocyte count was about twice normal in compensation 3. Coagulation disorders: Vitamin K deficiency, sweet clover
for a shortened erythrocyte life span.205 (dicoumarol) toxicity (cattle), rodenticide toxicity, bracken
fern toxicity (cattle and sheep), disseminated intravascular
Miscellaneous Causes of Hemolytic Anemia coagulation, inherited coagulation factor deficiencies (see
Splenic Disorders Chapter 7)
Disorders that cause splenomegaly may result in a syndrome 4. Platelet disorders: Thrombocytopenia and inherited platelet
function defects (see Chapter 7)
called hypersplenism, where phagocytosis of blood cells is
5. Neoplasia: Gastric tumors including carcinomas,
increased.442 Increased erythrophagocytosis and anemia occur
leiomyosarcoma, and lymphoma; transitional cell carcinoma
in various hemophagocytic disorders,516 most notably in and transitional cell papilloma of urogenital system; and
hemophagocytic histiocytic sarcoma, which typically involve ruptured hemangioma, hemangiosarcoma, and adrenal gland
the spleen.149,324 Erythrocyte destruction can occur secondary tumors with bleeding into body cavities and tissues
to splenic torsion, where stagnation and breakdown of blood 6. Gastrointestinal ulcers: Glucocorticoids, nonsteroidal
results in hemoglobinuria.336 anti-inflammatory drugs, mast cell tumors, gastrinoma,
stress, metabolic diseases (uremia, liver failure,
hypoadrenocorticism)
Liver Failure in Horses 7. Vascular abnormalities: Arteriovenous fistula and vascular
Marked intravascular hemolysis has been reported in horses ectasia in the gastrointestinal or urogenital tracts
8. Phenylephrine-induced hemorrhage: Presumably associated
with liver failure. The mechanism of this hemolysis is unknown,
with hypertension in aged horses treated for nephrosplenic
but bile acids or their salts have been considered possible
entrapment of the large colon
hemolytic factors in horses.383
cases hemorrhage is apparent but its cause must be deter-

mined. Finally, blood-loss anemia and its cause may not be A N EM I A S R E S U LT I N G F RO M
recognized until laboratory tests and other diagnostic tests are D E C R E A S ED ERY T H RO C Y T E
done. The gastrointestinal and urogenital tracts are common P RO D U C T I O N
sites of occult hemorrhage. Tests that may assist in the diag-
nosis of gastrointestinal hemorrhage include the occult blood Anemia resulting from decreased erythrocyte production
test in feces, fecal examination for parasite ova, and diagnostic lacks evidence of bone marrow response to the anemia (e.g.,
imaging to identify tumors or ulcers. Urinalysis and diagnostic the absolute reticulocyte count in blood is not increased or
imaging of the urinary system may assist in the diagnosis of only minimally raised for the degree of anemia). Nonregen-
renal or bladder hemorrhage. erative anemias result from reduced or defective erythropoi-
Although total blood volume is decreased, HCT and esis (Box 4-5). They are usually normocytic. Exceptions
plasma protein concentration are normal immediately after include microcytic anemia associated with chronic iron-
substantial acute blood loss has occurred because there is a deficiency anemia, copper deficiency, pyridoxine deficiency,
balanced loss of erythrocytes and plasma. The HCT may even and dyserythropoiesis in English springer spaniel dogs and
be increased shortly after acute blood loss in horses and dogs macrocytic anemia associated with folate deficiency, FeLV
because splenic contraction occurs, which releases blood with infection in cats, erythroleukemia, some myelodysplastic dis-
a higher HCT into the general circulation.237 After several orders, and dyserythropoiesis in polled Hereford calves.201,257
hours, the HCT and plasma protein concentration decrease Bone marrow biopsies are often required to delineate the
as fluid moves from the digestive tract and extravascular spaces nature of nonregenerative anemias.
into the circulation to return the blood volume toward normal.
If no further hemorrhage occurs, the plasma protein concen-
tration will return to normal within a few days. Consequently Box 4-5
the occurrence of a low plasma protein concentration in asso- Anemias Resulting from
Decreased Erythrocyte
ciation with anemia suggests the presence of recent or ongoing Production in Domestic Animals
hemorrhage. Considerably more time is required for the HCT
to return to normal than is required for the plasma protein Reduced Erythropoiesis
concentration to return to normal. The HCT increases about 1. Chronic renal disease: Primarily lack of erythropoietin
1 percentage point per day following experimental phlebot- 2. Endocrine deficiencies: Hypothyroidism,
omy in dogs and cats, with a slightly lower response in cattle hypoadrenocorticism, hypopituitarism, hypoandrogenism
and horses.15,65,364 3. Inflammatory disease: Inflammation and neoplasia
The anemia appears nonregenerative shortly after blood 4. Cytotoxic damage to the marrow: Bracken fern poisoning
loss because approximately 4 days are required for production (cattle), cytotoxic anticancer drugs, estrogen toxicity (dogs
and ferrets), chloramphenicol (cats, usually not anemic),
of reticulocytes by the marrow. The MCV may not be increased
phenylbutazone (dogs), trimethoprim-sulfadiazine (dogs),
following blood loss in animals because the reticulocyte radiation, albendazole (dogs, cats, alpacas), griseofulvin
response may not be of sufficient magnitude to result in a high (cats), trichloroethylene (cattle)
MCV. Few reticulocytes are released from the marrow in 5. Infectious agents: Ehrlichia spp. (dogs, horses, and cats),
response to blood-loss anemia in cattle and no reticulocytes FeLV, nonbloodsucking trichostrongyloid parasites
are released following hemorrhage in horses. (ruminants), parvovirus (pups)
Chronic external blood loss can result in iron deficiency. 6. Immune-mediated: Nonregenerative anemia, selective
Iron-deficiency anemia is common in adult dogs and rumi- erythroid aplasia, continued treatment with recombinant
nants but seldom occurs in adult cats and horses because human erythropoietin, idiopathic aplastic anemia (?)
parasitism causing significant blood loss is uncommon in 7. Congenital/inherited: Foals and dogs?
these species. If iron deficiency persists for several weeks, the 8. Myelophthisis: Myeloid leukemias, lymphoid leukemias,
myelodysplastic syndromes, multiple myeloma, myelofibrosis,
anemia can become microcytic and hypochromic. Reticulo-
osteosclerosis, metastatic lymphomas, metastatic mast cell
cyte counts may be slightly to moderately increased in early tumors
iron-deficiency anemia in dogs; however, as iron deficiency
becomes more severe, the regenerative response will be attenu- Defective Erythropoiesis
ated (see discussion of iron deficiency under “Abnormalities 1. Disorders of heme synthesis: Iron, copper, and pyridoxine
in Heme Synthesis,” later in this chapter).204 Internal hemor- deficiencies; lead toxicity; drugs
rhage can share some characteristics of hemolytic anemias. 2. Disorders of nucleic acid synthesis: Folate and cobalamin
Iron is conserved so that hypoferremia does not occur. Slight deficiencies
hyperbilirubinemia may occur due to phagocytosis and deg- 3. Abnormal maturation: Erythroleukemia or AML-M6 (primarily
radation of erythrocytes at the sites of widespread hemor- cats), myelodysplastic syndromes with erythroid
rhage. Some plasma proteins may be reabsorbed when predominance (MDS-Er), inherited dyserythropoiesis of
Hereford calves, inherited dyserythropoiesis of English
hemorrhage occurs in body cavities, thus shortening the return
springer spaniels
of plasma protein concentrations to normal.
Nonregenerative Anemias without Leukopenia Abnormalities in Heme Synthesis

or Thrombocytopenia Iron deficiency in adult domestic animals usually results from
A nonregenerative anemia without an accompanying leuko- blood loss. The absolute reticulocyte count may be increased
penia or thrombocytopenia in blood suggests a bone marrow early in response to hemorrhage, but as iron deficiency
abnormality affecting only erythroid cells. Mild to moderate becomes more severe, a minimal regenerative response is
anemia of this type may occur in association with chronic present. Microcytic erythrocytes form when iron becomes
renal disease, endocrine deficiencies, and the anemia of inflam- limiting because erythroid cells apparently undergo additional
matory disease. Erythroid production is reduced in these dis- divisions, resulting in smaller-than-normal cells. If sufficient
orders, but often not enough to result in an M : E ratio in the time has elapsed for these small cells to account for a substan-
marrow that is increased above the reference interval. tial portion of the total erythrocyte population, the MCV will
decrease below the normal reference interval. When the MCV
Hormone Deficiencies is only slightly decreased, the MCHC is usually normal.
Because the kidney is the major site of EPO production in When the MCV is substantially below normal, the MCHC
the body, chronic renal disease can result in a mild to moderate will also be decreased.204 Erythrocytes in these microcytic
nonregenerative anemia secondary to reduced EPO produc- hypochromic anemias will appear hypochromic (pale cells
tion.250,360 Disorders such as hypopituitarism, hypoadrenocor- with prominent areas of central pallor) on stained blood films.
ticism, and hypothyroidism may result in mild nonregenerative A low MCHC is seldom present and hypochromasia is usually
anemia because these hormones enhance the growth of ery- not apparent in stained blood films from iron-deficient horses
throid progenitor cells in the presence of EPO.205,282,357 Glu- and adult cats. Hematologic aspects of iron deficiency are
cocorticoids appear to be important in stress erythropoiesis compared to the anemia of inflammatory disease in Table 4-1.
(e.g., following hemorrhage or increased erythrocyte destruc- Milk contains little iron; consequently nursing animals can
tion, when a substantial increase in erythropoiesis is required). deplete body iron stores as they grow. Microcytic erythrocytes
Thyroid hormones may also promote the synthesis of EPO in are produced in response to iron deficiency, but a low MCV
the kidney.205 may not develop postnatally in species such as dogs and cats
in which the MCV is above adult values at birth. The potential
Anemia of Inflammatory Disease for development of severe iron deficiency in young animals
(Anemia of Chronic Disease) appears to be less in species that begin to eat food at an early
A mild to moderate nonregenerative anemia often accom age. Piglets are especially susceptible to the development of
panies chronic inflammatory and neoplastic disorders. The iron deficiency when they are not raised on dirt; thus the
cause of the anemia is multifactorial and only partially under- practice of iron injections of piglets.204
stood. Abnormalities that can contribute to the anemia Prolonged copper deficiency usually results in anemia in
include low serum iron, the production of inflammatory mammals. Because copper is required for normal iron
mediators that can inhibit erythropoiesis, and shortened
erythrocyte life spans, presumably secondary to membrane
damage caused by endogenous oxidants generated during
inflammation.204
Disorders of Nucleic Acid Synthesis Table 4-1

Anemia resulting from folate deficiency is rarely reported in Laboratory Findings in Chronic Iron-
animals. Macrocytic anemia has been produced experimen- Deficiency Anemia versus the Anemia of
tally in pigs and a clinical case of folate deficiency has been Inflammatory Disease
recognized in a cat.205 Cobalamin (vitamin B12) deficiency in Anemia of
humans, who require cobalamin for normal folate metabolism, Chronic Iron Inflammatory
causes hematologic abnormalities similar to folate deficiency. Parameter Deficiency Disease
In contrast, cobalamin deficiency does not cause macrocytic HCT Slight to marked Slight to moderate
anemia in any animal species. Anemia has been reported in decrease decrease
some experimental animal studies, but erythrocytes were of MCV Slight to marked Normal to slight
normal size. Cobalamin deficiency occurs secondarily to an decrease decrease
inherited malabsorption of cobalamin in dogs. Affected Serum iron Slight to marked Slight to moderate
decrease decrease
animals have normocytic, nonregenerative anemia with
Serum TIBC Normal to increased Normal to decreased
increased anisocytosis. Additional findings include neutrope- Serum ferritin Decreased Normal to increased
nia with hypersegmented neutrophils and giant platelets. A Marrow Decreased or absent Normal to increased
normocytic nonregenerative anemia was also present in a hemosiderin
cobalamin-deficient cat that probably had an inherited defect
in cobalamin absorption.205 HCT, Hematocrit; MCV, mean cell volume; TIBC, total iron-binding capacity.
metabolism, the anemia that develops is generally microcytic, may be a prominent feature of some hematopoietic neoplasms,
but it may be normocytic. Pyridoxine (vitamin B6) is required especially erythroleukemia, and some myelodysplastic syn-
for the first step in heme synthesis. While natural cases of dromes.56,125,518 However, leukopenia and/or thrombocytope-
pyridoxine deficiency have not been documented in domestic nia are usually present along with nonregenerative anemia in
animals, microcytic anemias with high serum iron values have these latter disorders, which in cats are usually associated with
been produced experimentally in dogs, cats, and pigs with FeLV infections.56
dietary pyridoxine deficiency.205
Nonregenerative Anemias with Leukopenia
Nonregenerative Immune-Mediated Anemia and/or Thrombocytopenia
Erythroid cellularity in the marrow varies from hypocellular The pattern of a pancytopenia with a nonregenerative anemia
to hypercellular in dogs and cats with nonregenerative suggests a defect in the production of blood cells in bone
immune-mediated anemia. Erythroid maturation may be marrow. The bone marrow may be hypocellular, with low
complete to the polychromatophilic erythrocyte stage or a numbers of cells (including hematopoietic precursors) present,
maturation arrest may occur at an earlier stage of erythrocyte or high numbers of abnormal cells may have replaced the
development.244,261,454,517 A nonregenerative immune-mediated normal hematopoietic precursors in marrow (myelophthi-
anemia with maturation arrest has also been reported in a sis).520 However, pancytopenia with nonregenerative anemia
ferret.294 An antibody or cell-mediated response may be may sometimes be present in disorders other than marrow
directed against one or more maturation antigens present on hypoplasia/aplasia and myelophthisis. These disorders gener-
nucleated erythrocyte precursors and/or reticulocytes. Ery- ally have increased numbers of macrophages (histiocytes) in
throid hypoplasia occurs when the immune response is bone marrow and/or peripheral tissues, and the destruction of
directed at earlier stages of erythroid development.517 blood cells is a component of the pathogenesis of the cytope-
nias in these histiocytic disorders. Histiocytic inflammatory
Selective Erythroid Aplasia conditions that may have accompanying pancytopenia include
Pure red cell aplasia or selective erythroid aplasia can result the terminal stage of cytauxzoonosis in cats,50 histoplasmo-
in severe anemia in dogs and cats. Most cases appear to be sis,157 leishmaniasis,300 and mycobacteriosis.363 Bicytopenia or
acquired, but congenital erythroid aplasia may occur in dogs.320 pancytopenia has been reported in dogs classified as having
Some cases in adult dogs and cats appear to be immune- the hemophagocytic syndrome (activated macrophage syn-
mediated.511,517 Selective erythroid aplasia occurs in cats drome) with greater than 2% hemophagocytic macrophages
infected with FeLV subgroup C, but not in cats infected only in bone marrow.516 This syndrome develops secondary to some
with subgroups A or B.155 Colony-forming-unit-erythrocyte infectious, neoplastic, and immune-mediated conditions, but
(CFU-E) numbers are markedly decreased but burst-forming- an underlying disease may not always be identified.499,516
unit-erythrocyte (BFU-E) numbers are normal in infected Cytopenias may also occur with hypersplenism, associated
cats.1 FeLV-C binds to a heme exporter on bone marrow with splenomegaly and increased phagocytosis of blood cells
CFU-E cells, and it is hypothesized that this binding inhibits by the spleen.87,420,442,520 Hemophagocytic histiocytic sarcoma
heme export from these cells, resulting in their destruction may also result in multiple cytopenias.149,514,520 The anemia is
because free heme is toxic to cells.381 High doses of chloram- generally nonregenerative in most of these histiocytic disor-
phenicol cause reversible erythroid hypoplasia in some dogs ders because of accompanying inflammation (see “Anemia of
and erythroid aplasia in cats.205 Marked erythroid hypoplasia Inflammatory Disease,” above), but the anemia may be regen-
has been reported in dogs, cats, and horses given recombinant erative in response to erythrocyte phagocytosis in some his-
human EPO.106,369 Antibodies made against this human tiocytic disorders.
recombinant glycoprotein apparently cross-react with the
animals’ endogenous EPO.369 A recombinant cat EPO pro- Hypocellular/Aplastic Bone Marrow
duced in a hamster cell line has also caused erythroid aplasia A marrow is classified as hypoplastic when 5% to 25% of the
in cats.386 hematopoietic space consists of bone marrow cells. General-
ized necrosis may result in hypocellular marrow,247,420,426,525 but
Dyserythropoiesis generally fat replaces lost hematopoietic cells in hypocellular
The term dyserythropoiesis refers to various disorders in which marrow. When all hematopoietic cell types—erythrocytic,
abnormal erythrocyte maturation and/or morphology is asso- granulocytic, and megakaryocytic—are absent or markedly
ciated with ineffective erythropoiesis (see dyserythropoiesis reduced (less than 5% of the hematopoietic space consists of
section in Chapter 9). Dyserythropoiesis is a prominent com- hemic cells), the marrow is said to be aplastic. Anemic animals
ponent of some congenital (presumably inherited) disorders with generalized marrow aplasia in which nearly all of the
in Hereford calves and English springer spaniel dogs.223,447 marrow space is occupied by fat are reported to have aplastic
Dyserythropoiesis has been reported in association with anemia. When only one cell line is reduced or absent, more
immune-mediated disorders, drug toxicities, and myelofibro- restrictive terms, such as granulocytic hypoplasia or erythroid
sis.9,518 Dyserythropoiesis with Cabot rings has been reported aplasia, are used to describe the abnormalities present. Hypo-
in a dog with a metastatic carcinoma.286 Dyserythropoiesis cellular or aplastic bone marrow may result from insufficient
numbers of stem cells, abnormalities in the hematopoietic cases of aplastic anemia are immune-mediated, and activated
microenvironment or abnormal humoral or cellular control of type-1 cytotoxic T cells have been implicated.548 Consequently
hematopoiesis. These factors are interrelated, and the specific most cases of idiopathic aplastic anemia in animals may be
defect in a given disorder is usually unknown. immune-mediated, if the pathophysiology is similar to that in
Drug-induced causes of aplastic anemia or generalized humans.
marrow hypoplasia in animals include estrogen toxicity in Congenital aplastic anemia, renal abnormalities, and skin
dogs,440 phenylbutazone toxicity in dogs (and possibly lesions have been reported in newborn foals whose mothers
horses),521 trimethoprim-sulfadiazine administration in were treated for equine protozoal myeloencephalitis with sul-
dogs,144 bracken fern poisoning in cattle and sheep,496 fonamides and/or pyrimethamine during pregnancy.482 Aplas-
trichloroethylene-extracted soybean meal in cattle,180 alben- tic anemia has been described in 11- and 14-day-old Holstein
dazole toxicity in dogs, cats, and alpacas,188,455 griseofulvin calves that may have also developed in utero.19,424 An in utero
toxicity in cats,402 methimazole toxicity in cats,515 various toxic insult was suspected in a 9-week-old Clydesdale foal with
cancer chemotherapeutic agents, immunosuppressive drugs, aplastic anemia.317 Generalized bone marrow hypoplasia has
such as azathioprine, and radiation.150,180,359,366,510 Thiacetars- been reported in eight young standard-bred horses sired by the
amide, meclofenamic acid, and quinidine have also been same stallion, suggesting an inherited etiology.262
incriminated as potential causes of aplastic anemia in dogs.521
In addition to exogenous estrogen injections, aplastic Myelophthisic Disorders
anemia can occur in dogs because of high levels of endogenous Myelophthisic disorders are characterized by the replacement
estrogens produced by Sertoli cell, interstitial cell, and granu- of normal hematopoietic cells with abnormal ones. Examples
losa cell tumors.440 Functional cystic ovaries also have the include myelogenous leukemias, lymphoid leukemias, multi-
potential of inducing myelotoxicity in dogs.69 Ferrets have ple myeloma, myelodysplastic syndromes, and myelofibrosis
induced ovulations and may remain in estrus for long periods (often associated with anemia but less often with pancytope-
of time when they are not bred. This prolonged exposure to a nia). Nonregenerative anemia with leukopenia and/or throm-
high endogenous estrogen concentration can result in aplastic bocytopenia are often recognized in cats infected with FeLV
anemia.258 and less often with cats infected with FIV.154,174 Examination
Acute parvovirus infections may cause transient marrow of bone marrow generally reveals the presence of a myelodys-
hypoplasia, but not true aplastic anemia. Parvovirus infections plastic syndrome or less often leukemia.154,512 Multiple cyto-
can cause erythroid hypoplasia, as well as myeloid hypoplasia penias may sometimes be present secondary to the extensive
in canine pups,375,399 but generally only myeloid hypoplasia in metastasis of lymphomas, carcinomas, and mast cell tumors.*
adult dogs and cats.68,272,273 Either affected animals die acutely Myelophthisic disorders do not simply “crowd out” normal
or the bone marrow returns to normal within a week. If cells, but also alter the marrow microenvironment so that
present, anemia is usually mild, unless GI hemorrhage is normal hematopoiesis is compromised. In the case of myelo-
severe, because of the long erythrocyte life span. Thrombocy- dysplastic syndromes, increased apoptosis probably accounts
topenia, if present, is generally mild unless DIC occurs as part for the ineffective hematopoiesis that is present.425
of the disease process.184,520
Although some degree of marrow hypoplasia and/or dys- Physiologic Anemia of Neonatal Animals
plasia often occurs in cats with FeLV infections,103 true aplas- HCT and hemoglobin values increase during fetal develop-
tic anemia is not a well-documented sequela,400 but it may ment, reaching values near those of adult animals upon birth
rarely occur.515 Hypocellular bone marrow has been reported (Fig. 4-102). Following birth, there is a rapid decrease in these
in experimental cats coinfected with FeLV and feline parameters during the first few weeks of life, followed by a
parvovirus.289 gradual increase to adult values by 4 months of age in most
Dogs with acute Ehrlichia canis infections may spontane- species (Fig. 4-103). Factors involved in the development of
ously recover or develop chronic disease that generally exhibits the anemia of the neonate include absorption of colostral
some degree of marrow hypoplasia. Although rare, aplastic proteins during the first day of life (increases plasma volume
anemia may develop in association with severe chronic through an osmotic effect), decreased erythrocyte production
ehrlichiosis in dogs.75,333,337 Natural cases of East Coast fever during the early neonatal period, shortened life span of eryth-
(Theileria parva infection) have been described in cattle with rocytes formed in utero, and rapid growth with hemodilution
generalized marrow hypoplasia and pancytopenia.305 resulting from total plasma volume expansion, which occurs
Aplastic anemia has been reported in five cats with chronic more rapidly than the increase in total erythrocyte mass.205
renal failure. These cats also exhibited prolonged anorexia and/ In some species, production of erythrocytes is decreased
or emaciation, and it was suggested that starvation played a because of low EPO concentrations at birth. The decreased
role in the development of marrow aplasia in these cases.515 stimulus for EPO production at birth may occur as a result of
Idiopathic aplastic anemia has been reported in dogs,497,520 a placental blood transfusion that increases erythrocyte mass
cats,515 and horses.275,317 One case of erythroid and myeloid
aplasia with normal megakaryocyte numbers has been reported
in a horse; the etiology was unknown.501 In humans, most *References 4, 22, 135, 154, 220, 279, 299, 420, 514, 519, 520.
immediately after birth, a rapid increase in PaO2 associated

with breathing air, and a decrease in hemoglobin O2 affinity ERY T H RO C Y T O S I S
due to an increase in erythrocyte 2,3DPG content after birth. ( P O LYC Y T H EM I A )
Although not involved in the early, rapid decrease in HCT,
iron availability may limit the response to anemia in some Erythrocytosis refers to an increase in HCT, hemoglobin, and
rapidly growing animals.204 RBC count above the normal reference interval. The reference
interval can sometimes vary by breed as well as by species. The
13 HCTs of hot-blooded horses (e.g., thoroughbreds, quarter
horses, and Arabians) are usually higher than those of draft
12 horses because of the larger spleens, relative to body weight,
Hemoglobin (g/dL)
11
in the hot-blooded group.254 Some sighthound breeds (grey-
hounds, whippets, Afghan hounds, and salukis) have higher
10 HCTs than other breeds.238,481 The reference interval for the
HCT in adult greyhound dogs is reported to be 48% to
9 64%.423 A reference interval of 50% to 69% was reported using
8
blood from healthy adult whippets, Afghan hounds, and
salukis.219 In addition, slightly increased HCTs are sometimes
7 measured in individuals from some non-sighthound breeds
30 20 10 0 10 20 30 40 50 60 150 (i.e., poodle, German shepherd, boxer, beagle, dachshund,
Days from birth and Chihuahua).238 These somewhat higher values may result
from splenic contraction in animals with a high normal eryth-
FIGURE 4-102
rocyte mass.
Blood hemoglobin values in prenatal and postnatal cats.
Relative Erythrocytosis
Data from Windle WF, Sweet M, Whitehead WH. Some aspects of prenatal
and postnatal development of the blood of cats. Anat Rec. 1940;78:321. Erythrocytosis is either relative (spurious) or absolute (Box
4-6). A relative erythrocytosis is one in which the HCT is
Basenji pup high but the total erythrocyte mass in the body is normal.
6.5 It is caused by splenic contraction or dehydration. Splenic
contraction results from sympathetic stimulation as occurs
PP (g/dL)
6.0
with excitement, fear, pain, or exercise. The HCT measured in
5.5 blood from peripheral veins increases because the HCT in the
5.0 spleen is considerably higher than that in the general circula-
90 tion.110,449 Splenic contraction results in higher (30% to 50%)
MCV (fL)
80
70
60
Box 4-6
Erythrocytosis in Domestic
15 Animals
Retic (%)
10 Relative Erythrocytosis
5 1. Splenic contraction: Excitement, exercise, pain (primarily in
horses, dogs, and cats)
0
2. Dehydration: Water loss, water deprivation, shock with fluid
40 shift into tissues
Absolute Erythrocytosis
PCV (%)
30
1. Primary erythrocytosis: A myeloproliferative neoplasm in
adult dogs and cats
20
2. Familial erythrocytosis in young Jersey cattle: Etiology
unknown
3. Hypoxemia with compensatory increased erythropoietin
0 20 40 60 80 100 120
production: Chronic lung disease, heart disease with
Days of age right-to-left shunting of blood, chronic methemoglobinemia
(rare in dogs and cats)
FIGURE 4-103 4. Inappropriate erythropoietin production: Renal lesions
Age-related changes in total plasma protein (PP) concentration, mean (primarily tumors), nonrenal erythropoietin secreting
cell volume (MCV), reticulocyte (Retic) count, and packed cell volume tumors (rare)
(PCV) in blood from a basenji dog.
increases in HCT in dogs, cats, and hot-blooded horses than confirmed by finding evidence of dehydration on physical
in ruminants, pigs, or draft horses because the former group examination.
have large, contractile spleens.449,485 The persistence of a moderate or marked increase in HCT
Dehydration results from increased water loss (diarrhea, suggests that an absolute erythrocytosis is present. Tests that
vomiting, excessive diuresis, or sweating) or from water depri- may help determine the cause of the absolute erythrocytosis
vation. The plasma protein concentration is also usually include arterial blood gas measurements, diagnostic imaging,
increased. The HCT may also be high when increased vascular a methemoglobin screening test, and a validated EPO test.
permeability results in water loss from the circulation into the The cytologic examination of bone marrow is not useful.
tissues, as occurs in endotoxic shock.5,111,251 When present, methemoglobinemia is easily recognized using
a simple spot test (see “Methemoglobin Determination,”
Absolute Erythrocytosis above). The presence of low PaO2 suggests that either a heart
An absolute erythrocytosis is one in which the HCT is high defect (with right-to-left shunting of blood) or chronic lung
because the total erythrocyte mass in the body is increased. disease is present. Diagnostic imaging procedures are used to
Absolute erythrocytosis may occur secondary to increased differentiate heart and lung disease and search for renal lesions
EPO production (secondary erythrocytosis) or in disorders and tumors. Plasma EPO values should be increased when
where increased erythrocyte proliferation occurs in the pres- hypoxemia, renal lesions, or EPO-secreting tumors cause
ence of normal or low blood EPO values (primary erythrocy- the erythrocytosis, but are low when primary erythrocytosis
tosis). Causes of secondary erythrocytosis include chronic is present. Unfortunately, there is considerable overlap among
hypoxemia (heart defects with right-to-left shunting of patients with primary and secondary erythrocytosis, limiting
blood,102,323,418 diffuse lung disease,42,367 persistent methemo- the diagnostic value of the EPO assay.96 A diagnosis of
globinemia203), renal disorders causing local tissue hypoxia primary erythrocytosis is reached after ruling out other poten-
(renal tumors74,253 and localized inflammation248), and tial causes of persistent erythrocytosis.
tumors that secrete EPO, EPO-like proteins, or other hor-
mones such as androgens that might enhance the effects of
EPO.104,177,256,408,546
R EF ER EN C E S
Primary erythrocytosis (polycythemia vera) is considered
to be a myeloproliferative neoplasm that results from an
1. Abkowitz JL. Retrovirus-induced feline pure red blood cell aplasia: pathogenesis and
autonomous (erythropoietin-independent) proliferation of response to suramin. Blood. 1991;77:1442-1451.
erythroid precursor cells, resulting in high numbers of mature 2. Abrams K, Yunusov MY, Slichter S, et al. Recombinant human macrophage
erythrocytes in blood.358,503 In contrast to humans with poly- colony-stimulating factor-induced thrombocytopenia in dogs. Br J Haematol.
2003;121:614-622.
cythemia vera, blood granulocyte and platelet numbers are 3. Adachi K, Makimura S. Changes in anti-erythrocyte membrane antibody level of dogs
generally not increased in animals; consequently, in veterinary experimentally infected with Babesia gibsoni. J Vet Med Sci. 1992;54:1221-1223.
medicine, the term primary erythrocytosis is more appropriate 4. Adam F, Villiers E, Watson S, et al. Clinical pathological and epidemiological assess-
ment of morphologically and immunologically confirmed canine leukaemia. Vet Comp
than polycythemia vera. A diagnosis of primary erythrocytosis Oncol. 2009;7:181-195.
is ultimately made by ruling out causes of the secondary 5. Adams T Jr, Traber DL. The effects of a prostaglandin synthetase inhibitor, ibuprofen,
erythrocytosis. on the cardiopulmonary response to endotoxin in sheep. Circ Shock.
1982;9:481-489.
Familial erythrocytosis (HCTs of 60% to 80%) has been 6. Aird B. Acute blood loss. In: Feldman BF, Zinkl JG, Jain NC, eds. Schalm’s Veterinary
described in calves from a highly inbred Jersey herd.475 The Hematology, 5th ed. Philadelphia: Lippincott Williams & Wilkins; 2000:151-153.
cause of this defect was not determined. Affected calves had 7. Al-Huniti NH, Widness JA, Schmidt RL, et al. Pharmacodynamic analysis of changes
in reticulocyte subtype distribution in phlebotomy-induced stress erythropoiesis. J
normal hemoglobin types and arterial blood gas values and Pharmacokinet Pharmacodyn. 2005;32:359-376.
lacked measurable EPO in plasma. The majority of the affected 8. Allard RL, Carlos AD, Faltin EC. Canine hematologic changes during gestation and
calves died by 6 months of age. HCTs of surviving animals lactation. Compan Anim Pract. 1989;19:3-6.
9. Alleman AR, Harvey JW. The morphologic effects of vincristine sulfate on canine
returned slowly to normal by maturity. Erythrocytosis of bone marrow cells. Vet Clin Pathol. 1993;22:36-41.
unknown etiology has also been diagnosed in an 8-month-old 10. Allen AL, Meyers SL, Searcy GP, et al. Hematology of equine fetuses with compari-
Japanese black heifer.468 A variety of familial and congenital sons to their dams. Vet Clin Pathol. 1998;27:93-100.
11. Allen BV. Relationships between the erythrocyte sedimentation rate, plasma proteins
erythrocytosis syndromes have been described in humans. and viscosity, and leucocyte counts in thoroughbred racehorses. Vet Rec.
They include altered hypoxia sensing, mutations in the 1988;122:329-332.
EPO receptor gene, high-affinity hemoglobins, and 2,3DPG 12. Allison RW, Fielder SE, Meinkoth JH. What is your diagnosis? Blood film from an
icteric cat. Vet Clin Pathol. 2010;39:125-126.
deficiency.205 13. Allison RW, Velguth KE. Appearance of granulated cells in blood films stained by
automated aqueous versus methanolic Romanowsky methods. Vet Clin Pathol.
Differential Diagnosis of Erythrocytosis 2010;39:99-104.
14. Allsopp MT, Cavalier-Smith T, De Waal DT, et al. Phylogeny and evolution of the
Splenic contraction is considered a likely cause of erythrocy- piroplasms. Parasitology. 1994;108(Pt 2):147-152.
tosis when the HCT is slightly to moderately increased in the 15. Alsaker RD, Laber J, Stevens JB, et al. A comparison of polychromasia and reticulo-
absence of evidence of dehydration. A slight to moderate cyte counts in assessing erythrocyte regenerative response in the cat. J Am Vet Med
Assoc. 1977;170:39-41.
increase in HCT with increased plasma protein concentration 16. Altman NH. Intraerythrocytic crystalloid bodies in cats and their comparison with
suggests that dehydration is present. This interpretation is hemoglobinopathies of man. Ann N Y Acad Sci. 1974;241:589-593.
17. Altman NH, Melby EC, Squire RA. Intraerythrocytic crystalloid bodies in cats. Blood. 49. Birkenheuer AJ, Correa MT, Levy MG, et al. Geographic distribution of babesiosis
1972;39:801-803. among dogs in the United States and association with dog bites: 150 cases (2000-
18. Alward A, Corriher CA, Barton MH, et al. Red maple (Acer rubrum) leaf toxicosis 2003). J Am Vet Med Assoc. 2005;227:942-947.
in horses: a retrospective study of 32 cases. J Vet Intern Med. 2006;20:1197- 50. Birkenheuer AJ, Le JA, Valenzisi AM, et al. Cytauxzoon felis infection in cats in the
1201. mid-Atlantic states: 34 cases (1998-2004). J Am Vet Med Assoc. 2006;228:568-571.
19. Ammann VJ, Fecteau G, Helie P, et al. Pancytopenia associated with bone marrow 51. Birkenheuer AJ, Neel J, Ruslander D, et al. Detection and molecular characterization
aplasia in a Holstein heifer. Can Vet J. 1996;37:493-495. of a novel large Babesia species in a dog. Vet Parasitol. 2004;124:151-160.
20. Anderson GJ, Frazer DM, McLaren GD. Iron absorption and metabolism. Curr Opin 52. Biryomumaisho S, Katunguka-Rwakishaya E. The pathogenesis of anaemia in goats
Gastroenterol. 2009;25:129-135. experimentally infected with Trypanosoma congolense or Trypanosoma brucei: use
21. Andress JL, Day TK, Day D. The effects of consecutive day propofol anesthesia on of the myeloid:erythroid ratio. Vet Parasitol. 2007;143:354-357.
feline red blood cells. Vet Surg. 1995;24:277-282. 53. Blakley BR, Fraser LM, Waldner C. Chlorate poisoning in beef cattle. Can Vet J.
22. Angel KL, Spano JS, Schumacher J, et al. Myelophthisic pancytopenia in a pony mare. 2007;48:1071-1073.
J Am Vet Med Assoc. 1991;198:1039-1042. 54. Bloom JC, Thiem PA, Sellers TS, et al. Cephalosporin-induced immune cytopenia in
23. Anosa VO, Logan-Henfrey LL, Wells CW. The haematology of Trypanosoma congo- the dog: demonstration of erythrocyte-, neutrophil-, and platelet-associated IgG fol-
lense infection in cattle I. Sequential cytomorphological changes in the blood and lowing treatment with cefazedone. Am J Hematol. 1988;28:71-78.
bone marrow of Boran cattle. Comp Haematol Int. 1997;7:14-22. 55. Blue J, Weiss L. Vascular pathways in nonsinusal red pulp—an electron microscope
24. Arese P, De Flora A. Denaturation of normal and abnormal erythrocytes II. Pathophysi- study of the cat spleen. Am J Anat. 1981;161:135-168.
ology of hemolysis in glucose-6-phosphate dehydrogenase deficiency. Semin Hematol. 56. Blue JT, French TW, Kranz JS. Non-lymphoid hematopoietic neoplasia in cats: a
1990;27:1-40. retrospective study of 60 cases. Cornell Vet. 1988;78:21-42.
25. Aroch I, Segev G, Loeb E, et al. Peripheral nucleated red blood cells as a prognostic 57. Boas FE, Forman L, Beutler E. Phosphatidylserine exposure and red cell viability in red
indicator in heatstroke in dogs. J Vet Intern Med. 2009;23:544-551. cell aging and in hemolytic anemia. Proc Natl Acad Sci U S A. 1998;95:3077-3081.
26. Aronson LR, Gregory C. Possible hemolytic uremic syndrome in three cats after renal 58. Boisvert AM, Tvedten HW, Scott MA. Artifactual effects of hypernatremia and hypo-
transplantation and cyclosporine therapy. Vet Surg. 1999;28:135-140. natremia on red cell analytes measured by the Bayer H*1 Analyzer. Vet Clin Pathol.
27. Aslani MR, Mohri M, Movassaghi AR. Heinz body anaemia associated with onion 1999;28:91-96.
(Allium cepa) toxicosis in a flock of sheep. Comp Clin Path. 2005;14:118-120. 59. Bonfanti U, Comazzi S, Paltrinieri S, et al. Stomatocytosis in 7 related Standard
28. Atwell RB, Johnstone I, Read R, et al. Haemolytic anaemia in two dogs suspected to Schnauzers. Vet Clin Pathol. 2004;33:234-239.
have been induced by levamisole. Aust Vet J. 1979;55:292-294. 60. Bottomley SS. Sideroblastic anemias. In: Greer JP, Foerster J, Lukens JN, et al, eds.
29. Aucoin DP, Peterson ME, Hurvitz AI, et al. Propylthiouracil-induced immune-mediated Wintrobe’s Clinical Hematology. 11th ed. Philadelphia: Lippincott Williams & Wil-
disease in the cat. J Pharmacol Exp Ther. 1985;234:13-18. liams; 2004:1011-1033.
30. Azwai SM, Abdouslam OE, Al-Bassam LS, et al. Morphologic characteristics of blood 61. Boyle AG, Magdesian KG, Ruby RE. Neonatal isoerythrolysis in horse foals and a mule
cells in clinically normal adult llamas (Lama glama). Veterinarski Arhiv. foal: 18 cases (1988-2003). J Am Vet Med Assoc. 2005;227:1276-1283.
2007;77:69-79. 62. Brace RA, Langendorfer C, Song TB, et al. Red blood cell life span in the ovine fetus.
31. Backman L, Jonasson JB, Horstedt P. Phosphoinositide metabolism and shape control Am J Physiol Regul Integr Comp Physiol. 2000;279:R1196-R1204.
in sheep red blood cells. Mol Membr Biol. 1998;15:27-32. 63. Breitschwerdt EB. Feline bartonellosis and cat scratch disease. Vet Immunol Immu-
32. Badylak SF, Van Vleet JF, Herman EH, et al. Poikilocytosis in dogs with chronic nopathol. 2008;123:167-171.
doxorubicin toxicosis. Am J Vet Res. 1985;46:505-508. 64. Breitschwerdt EB, Armstrong PJ, Robinette CL, et al. Three cases of acute zinc toxi-
33. Balch A, Mackin A. Canine immune-mediated hemolytic anemia: pathophysiology, cosis in dogs. Vet Hum Toxicol. 1986;28:109-117.
clinical signs, and diagnosis. Compend Contin Educ Vet. 2007;29:217-225. 65. Bremner KC. The reticulocyte response in calves made anaemic by phlebotomy. Aust
34. Ban A, Ogata Y, Kato T, et al. Erythrocyte morphology and the frequency of sphero- J Exp Biol Med Sci. 1966;44:251-258.
cytes in hereditary erythrocyte membrane protein disorder in Japanese Black cattle. 66. Breznock EM, Strack D. Effects of the spleen, epinephrine, and splenectomy on
Bull Nippon Vet Anim Sci Univ. 1995;44:21-27. determination of blood volume in cats. Am J Vet Res. 1982;43:2062-2066.
35. Barker EN, Tasker S, Day MJ, et al. Development and use of real-time PCR to detect 67. Brinson JJ, Messick JB. Use of polymerase chain reaction assay for detection of
and quantify Mycoplasma haemocanis and “Candidatus Mycoplasma haematopar- Haemobartonella canis in a dog. J Am Vet Med Assoc. 2001;218:1943-1945.
vum” in dogs. Vet Microbiol. 2010;140:167-170. 68. Brock KV, Jones JB, Shull RM, et al. Effect of canine parvovirus on erythroid progeni-
36. Barrows PL, Prestwood AK, Green CE. Experimental Sarcocystis suicanis infections: tors in phenylhydrazine-induced regenerative hemolytic anemia in dogs. Am J Vet
disease in growing pigs. Am J Vet Res. 1982;43:1409-1412. Res. 1989;50:965-969.
37. Baskurt OK, Farley RA, Meiselman HJ. Erythrocyte aggregation tendency and cellular 69. Brockus CW. Endogenous estrogen myelotoxicity associated with functional cystic
properties in horse, human, and rat: a comparative study. Am J Physiol. ovaries in a dog. Vet Clin Pathol. 1998;27:55-56.
1997;273:H2604-H2612. 70. Brown DE, Meyer DJ, Wingfield WE, et al. Echinocytosis associated with rattlesnake
38. Beaudoin S, Lanevschi A, Dunn M, et al. Peripheral blood smear from a dog [hemo- envenomation in dogs. Vet Pathol. 1994;31:654-657.
globin crystals]. Vet Clin Pathol. 2002;31:33-35. 71. Brown DE, Weiser MG, Thrall MA, et al. Erythrocyte indices and volume distribution
39. Belford CJ, Raven CR, Black H. Chronic copper poisoning in Angora kids. New Zealand in a dog with stomatocytosis. Vet Pathol. 1994;31:247-250.
Veterinary Journal. 1989;37:152-154. 72. Brown HM, Berghaus RD, Latimer KS, et al. Genetic variability of Cytauxzoon felis
40. Berckmans RJ, Neiuwland R, Boing AN, et al. Cell-derived microparticles circulate in from 88 infected domestic cats in Arkansas and Georgia. J Vet Diagn Invest.
healthy humans and support low grade thrombin generation. Thromb Haemost. 2009;21:59-63.
2001;85:639-646. 73. Brown KM, Arthur JR. Selenium, selenoproteins and human health: a review. Public
41. Berman E. Hemograms of the cat during pregnancy and lactation and after lactation. Health Nutr. 2001;4:593-599.
Am J Vet Res. 1974;35:457-460. 74. Bryan JN, Henry CJ, Turnquist SE, et al. Primary renal neoplasia of dogs. J Vet Intern
42. Bertazzolo W, Zuliani D, Pogliani E, et al. Diffuse bronchiolo-alveolar carcinoma in a Med. 2006;20:1155-1160.
dog. J Small Anim Pract. 2002;43:265-268. 75. Buhles WC Jr, Huxsoll DL, Hildebrandt PK. Tropical canine pancytopenia: role of
43. Bessis M. Living Blood Cells and Their Ultrastructure. New York: Springer-Verlag; aplastic anaemia in the pathogenesis of severe disease. J Comp Pathol.
1973. 1975;85:511-521.
44. Bessis M, Delpech G. Sickle cell shape and structure: images and concepts (1840- 76. Bull BS. Morphology of the erythron. In: Lichtman MA, Beutler E, Kipps TJ, et al, eds.
1980). Blood Cells. 1982;8:359-435. Williams Hematology. 7th ed. New York: McGraw-Hill; 2006:369-385.
45. Beutler E. Disorders of iron metabolism. In: Lichtman MA, Beutler E, Kipps TJ, 77. Caldin M, Carli E, Furlanello T, et al. A retrospective study of 60 cases of eccentro-
et al, eds. Williams Hematology. 7th ed. New York: McGraw-Hill; 2006:511- cytosis in the dog. Vet Clin Pathol. 2005;34:224-231.
553. 78. Camacho AT, Guitian FJ, Pallas E, et al. Theileria (Babesia) equi and Babesia caballi
46. Bexfield N, Archer J, Herrtage M. Heinz body haemolytic anaemia in a dog secondary infections in horses in Galicia, Spain. Trop Anim Health Prod. 2005;37:293-302.
to ingestion of a zinc toy: A case report. Vet J. 2007;174:414-417. 79. Camacho AT, Pallas E, Gestal JJ, et al. Infection of dogs in north-west Spain with a
47. Bhasin JL, Freeman MJ, Morter RL. Properties of a cold hemagglutin associated with Babesia microti-like agent. Vet Rec. 2001;149:552-555.
leptospiral hemolytic anemia of sheep. Infect Immun. 1971;3:398-404. 80. Canfield PJ, Watson ADJ. Investigations of bone marrow dyscrasia in a poodle with
48. Birkenheuer AJ, Breitschwerdt EB, Alleman AR, et al. Differentiation of Haemobar- macrocytosis. J Comp Pathol. 1989;101:269-278.
tonella canis and Mycoplasma haemofelis on the basis of comparative analysis of 81. Canfield PJ, Watson ADJ, Ratcliffe RCC. Dyserythropoiesis, sideroblasts/siderocytes
gene sequences. Am J Vet Res. 2002;63:1385-1388. and hemoglobin crystallization in a dog. Vet Clin Pathol. 1987;16(1):21-28.

Hematology Vet-1-120

Uploaded by

Copyright:

Available Formats

Hematology Vet-1-120

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Hematology Vet-1-120

Uploaded by

Copyright:

Available Formats

What is the book about?

What is the book about?

What are some common hematological disorders mentioned?

What are some common hematological disorders mentioned?

Veterinary Hematology

A Diagnostic Guide and Color Atlas

John W. Harvey, DVM, PhD, DACVP

with 836 illustrations

Veterinary Hematology: A Diagnostic Guide and Color Atlas ISBN: 978-1-4377-0173-9

Library of Congress Cataloging-in-Publication Data

Vice President and Publisher: Linda Duncan

Laboratory tests are done for a variety of reasons. Screening I N T ER N A L V ER S U S EX T ER N A L

animals of both sexes and various breeds. Monogastric animals

nancy, emotional state, and activity level. The types of animals

FIGURE 1-2 FIGURE 1 -3

(Tables 1-1 and 1-2).7

above and 2.5% of healthy animals having values below the

Box 1-1 Table 1-2

FIGURE 1-5 FIGURE 1 -6

FIGURE 1-7 FIGURE 1 -8

Coefficient of variation (%)

count for a dog is 4600/µL before treatment and 5800/µL

Plasma proteins Alpha 14%

Inorganic salts, lipids, hormones,

with a central iron molecule that must be maintained in the

hyperbilirubinemia) in horses owing to reduced removal of

A, Courtesy of Heather Wamsley.

a refractometer for plasma protein determination. Plasma B L O O D C ELL CO U N T I N G

purple colors, and eosin stains bases, resulting in red color-

reticulocytes. The use of a Miller’s disc in one of the microscope

FIGURE 2-18 FIGURE 2 -19

In normal cats as well as in cats with a regenerative anemia,

FIGURE 2-23 FIGURE 2 -25

FIGURE 2-24 FIGURE 2 -26

in which the total leukocyte counts have been accurately

Moderately basophilic with Döhle bodies 2+

Hypochromasia and Shapes

Miscellaneous Cells and Parasites in Blood

examination of blood buffy coat smears. Those reaching the Microfilaria

FIGURE 2-36 FIGURE 2 -38

OV ERV I EW Organization of Bone Marrow

From Weiss L. The hematopoietic microenvironment of the bone marrow: an

Hematopoietic Growth Factors ERY T H RO P O I E S I S

Marrow macrophages are important components of the

Nutrients Needed for Erythropoiesis

Eosinophilic Eosinophilic Eosinophilic

Neutrophilic Neutrophilic Neutrophilic

Basophilic Basophilic Basophilic

Drawing by Perry Bain.

FIGURE 3-11 Blood

BFU-E CFU-E NRBC Myelocytes

Oxygen sensor Bands

FIGURE 3-14 Neutrophils

individual platelets within the sinuses and general circula-

Each hemoglobin tetramer is capable of binding four mol-

FIGURE 4-1 FIGURE 4 -3

the outer layer of the membrane, with carbohydrate groups

FIGURE 4-5 FIGURE 4 -7

FIGURE 4-6 FIGURE 4 -8

O2 O2 2H

dimers “Free” hemoglobin