S e r u m Erythropoietin Concentrations M e a s u r e d by
Radioimmunoassay in Normal, P o l y c y t h e m i c , and Anemic
Dogs and C a t s
Suzanne M. Cook and Clinton D. Lothrop, Jr.
Serum erythropoietin (Epo) concentrations were measured
by radioimmunoassay (RIA) in normal, polycythemic, and
anemic dogs and cats. The serum Epo concentration in nor-
mal dogs (n = 25) ranged from 7 to 37 mU/mL (median, 20
mU/mL); and in normal cats (n = 11) ranged from 9 to 38
mU/mL (median, 18 mU/mL). Polycythemic animals (PCV
> 55% in dogs, >45% in cats) were classified as those with
primary (polycythemia Vera), secondary, or polycythemia
of uncertain etiology. Dogs with polycythemia Vera (PV, n
= 8) had a median serum Epo concentration in the normal
range (17 mU/mL); cats with PV (n = 7) also had a median
serum Epo concentration that was within the normal range
(10 mU/mL). In the category of secondary polycythemias,
dogs (n = 7) (median, 30.7 mU/mL) and cats (n = 2 ) had
normal Epo concentrations. The median serum Epo
E rythropoietin (Epo), a glycoprotein hormone produced
by the kidney, is the principal regulator of red blood cell
(RBC) production.14 Epo is unique among hematopoietic
growth factors in that it is made primarily by a single organ
(the kidney) and acts specifically on erythroid precursors,
having little effect on other cell lineages.' In addition, Epo is
part of a classic feedback control mechanism: hypoxia acts to
increase erythropoietin production, whereas hyperoxia de-
creases its formation.' The RBC mass is thereby maintained
within the range that is most optimal for oxygen transport.'
Serum Epo determinations may be useful in distinguish-
ing primary and secondary polycythemias, in assessing the
degree to which renal impairment may be contributing to
anemia, and in identifying patients that may benefit from
replacement therapy with recombinant human Epo.
Until recently, Epo was measured in serum or plasma
with in vivo or in vitro bioassays. Bioassays are cumber-
some, nonspecific, and relatively insensitive, generally de-
tecting only increased Epo a~tivities.~ Despite these disad-
vantages, the in vivo bioassay (polycythemic mouse assay)
detects biologically active Epo in all mammalian species and
is the standard with which other assays are ~ o m p a r e dBy .~
contrast, in vitro bioassays yield conflicting results in nor-
From the Department of Environmental Practice, University of
Tennessee, College of Veterinary Medicine, Knoxville, TN.
Accepted June 9, 1992.
This work was supported in part by a generous grant from the G.
Harold and Leila Y. Mathers Charitable Foundation and NIH
Grant HL-15647.
Dr Cook's current address is PO Box 22994, Carmel, CA 93922.
Reprint requests: Dr Clinton D. Lothrop, Scott-Ritchey Research
Center, College of Veterinary Medicine, Auburn University, Au-
burn, A L 36849.
Copyright 01994 by the American College of Veterinary Internal
Medicine
0891-6640/94/0801-0002$3.00/0
18 Journal of VeterinaryInternal Medicine, Vol8, No 1 (January-February),1994:p p 18-25
concentration was significantly decreased (P< .05) in dogs
with PV compared with dogs with secondary polycythe-
mias. The median serum Epo concentrations in dogs (n =
13) and cats (n = 5) with anemias not due to chronic renal
disease were significantly increased (P < .05) compared
with normal dogs and cats. In cats with anemias due to
chronic renal disease (n = 5) the median serum Epo con-
centration was not significantly different from normal cats.
The measurement of the serum EPO concentration may be
useful in assessment of anemia or polycythemia but the
overlap of values with the normal range in all groups evalu-
ated limit its diagnostic use.
J Vet Intern Med 1994;8: 18-25. Copyright 0 1994 by the
American College of VeterinaryInternal Medicine.
ma1 dogs, presumably due to effects of substances in the se-
rum other than Epo.'-'
Recently, the cloning and expression of the human Epo
gene8.9has allowed highly purified recombinant human Epo
(rHuEpo) to become widely available. Radioimmunoassays
for Epo based on reagents derived from rHuEpo have been
shown to be specific and sensitive in humans." The RIAs
are relatively easy to perform and are widely available for
clinical use in human patient^.^ Variable immunologic
cross-reactivity among mammalian species renders many of
these assays unreliable for the measurement of Epo in dogs
and cats5 However, several RIAs have been shown to mea-
sure canine and feline Epo, including an assay based on
polyclonal antibodies against human E ~ o . ~ ~ " ~ ' '
The purpose of this study was to investigate the use of a
commercially available rHuEpo-derived RIA for serum Epo
determinations in normal dogs and cats, and in dogs and
cats with polycythemia or anemia.
Materials and Methods
Animals
Serum samples from 43 dogs and 18 cats submitted to the Uni-
versity of Tennessee College of Veterinary Medicine Clinical Endo-
crinology Laboratory for Epo determinations over a 3-year period
(1 988- 1991) were used in this study. Pertinent clinical information
(signalment, history, clinical signs, laboratory and ancillary diag-
nostic findings, final diagnosis, treatment, and outcome) were pro-
vided by the submitting veterinarian. Serum Epo determination
from 25 dogs and 1 1 cats that were clinically normal on the basis of
history and physical examination served as controls. Two normal
dogs in which anemia was chemically induced were cared for ac-
cording to the principles outlined in the National Institutes of
Health Guide for the Care and Use of Laboratory Animals.
Classification
Dogs and cats were classified according to packed cell volume
(PCV) as polycythemic (PCV > 55% in dogs, > 45% in cats) or ane-
 SERUM ERYTHROPOIETIN IN DOGS AND CATS
mic (PCV < 37% for dogs, < 24% for cats).I3 Polycythemic animals
were further classified as primary (polycythemia Vera), secondary,
or polycythermia of uncertain etiology. The diagnosis of polycythe-
mia Vera (PV) in cats (n = 7) and dogs (n = 8) was based on the
exclusion of common causes of secondary polycythemia in the di-
agnostic workup (eg, normal cardiopulmonary examination and
lack of evidence for renal or other neoplasia). Secondary polycythe-
mias (cats n = 2, dogs n = 8) were diagnosed based on the presence
ofa low arterial Po, and/or diagnostic evidence of cardiopulmonary
disease or histopathologic confirmation of renal neoplasia. A third
category, polycythemias of uncertain origin, was defined in dogs ( n
= 15) based on the presence ofa high PCV that was either associated
with diseases of noncardiopulmonary or nonrenal etiology, or in
which a diagnosis was not established. Dogs with spontaneous reso-
lution of unexplained polycythemia were also included in this cate-
gory.
Anemic animals were subdivided into those with anemia of non-
renal failure origin and anemia caused by chronic renal disease.
This classification was based on the presence or absence of bio-
chemical evidence (serum blood urea nitrogen [BUN] and creati-
nine concentrations, urine specific gravity) of impaired renal func-
tion. Causes of nonrenal failure anemias in cats included hepatic
lipidosis (n = I ) , gastrointestinal hemorrhage (n = l), feline leuke-
mia virus (FeLV) infection ( n = 1), gastrointestinal lymphosarcoma
(n = I), and coagulopathy ( n = 1). Causes of nonrenal failure ane-
mias in dogs included pure red cell aplasia ( n = 4), immune-medi-
ated hemolytic anemia ( n = 3), pyruvate kinase deficiency ( n = 2),
chemically-induced anemia (n = 2), myelofibrosis ( n = I), and id-
iopathic nonregenerative anemia (rz = 1). Five cats with anemia
caused by chronic renal disease were also evaluated.
Chemically Induced Anemias
To induce reticulocytosis and an increased serum Epo concentra-
tion, 2 normal, adult female mixed-breed dogs were given phenyl-
hydrazine (Sigma, St Louis, MO) (2.5 mg/kg 1V SID for 4 days).I4
Serum samples were collected for erythropoietin determination,
and the PCV was measured on days 1 through 4,7,9, and 1 1, with
day 1 being the first day of phenylhydrazine treatment.
Erythropoietin Radioimmunoassay
All serum samples were frozen and shipped by overnight carrier
on dry ice and stored at -70°C until assayed. Serum Epo concen-
trations were determined by RIA using a commercially available kit
that has been previously validated for people (EPO-Trac '"I RIA
kit, Catalog No 23200, Incstar Corporation, Stillwater, MN). The
standard curve ranged from 10 to 298 mU/mL. The intra-assay co-
efficient of variation (precision) determined on two separate occa-
sions for pooled canine serum was 3.2% (24.9 k 0.8 m U / m L five
replicates) and 3.6% (10.9 t 0.4 mU/mL; five replicates), and for
pooled feline serum it was 16.7% (1 1.8 2 2.0 mU/mL; five repli-
cates)and 5.7% (28.2 ? 1.6 mU/mL; five replicates). The inter-assay
coefficientsof variation (reproducibility) were 25.4% and 15.2%for
two separate canine serum pools (three replicates), and 6.6% and
7.8% for two separate feline serum pools (three replicates). The in-
ter-assay coefficient of variation for a high (74.7 ? 0.5 mU/mL;
*
three replicates) and low (2 I .2 1.5 mU/mL; three replicates) pool
of recombinant human Epo added to saline were 0.7% and 7.3%,
respectively. Canine and feline serum pools spiked with the highest
standard diluted linearly when mixed with Epo-free diluent. Recov-
ery experiments with purified canine and feline Epo were not done
because purified canine and feline Epo was not available.
19
Data Analysis
All data are expressed as ranges and medians. Statistical compar-
isons of groups were made by the use of the Kruskal-Wallis test
(SAS' System for Elementary Statistical Analysis Computer Soft-
ware, Cary, NC)." Correlation coefficients (r values) were calcu-
lated to assess the relationship between PCV and Epo concentra-
tions.16
Results
Serum Erythropoietin Concentrations
in Normal Dogs and Cats
The serum Epo concentrations in normal dogs ( n = 25)
ranged from 7 to 37 mU/mL (median 20 mU/mL). The se-
rum Epo concentrations in normal cats ( n = l l ) ranged
from 9 to 38 mU/mL (median 16 mU/mL).
Serum Erythropoietin Concentrations
in Polycythemia
Dogs. The clinical information and serum Epo concen-
trations for dogs with polycythemia are presented in Table
1. The PCV in dogs with PV ( n = 8) ranged from 64.3% to
85.0%(median PCV, 72.9%), and the serum Epo concentra-
tion ranged from 10.0 to 57.3 mU/mL (median Epo 17.4
mU/mL). The PCV in dogs with secondary polycythemias
( n = 7) ranged from 57.0% to 83.0% (median PCV 65.0%),
and the serum Epo concentration ranged from 13.3to 102.9
mU/mL (median Epo 30.7 mU/mL). In dogs with polycy-
themias of uncertain origin ( n = 15) the PCV ranged from
53.0% to 78.0% (median 65.2%), and the serum Epo con-
centration ranged from 8.1 to 131.O mU/mL (median Epo
13.3 mU/mL). There were no statistically significant differ-
ences in the rank sum of serum Epo concentrations of each
group (PV, secondary polycythemia, and polycythemias of
uncertain origin) compared with normal dogs. However, the
rank sum of serum Epo concentration was significantly
lower (P< .05) in dogs with PV compared with dogs with
secondary polycythemias.
Cars. The clinical information and serum Epo concen-
trations for cats with polycythemia is presented in Table 2.
The PCV in cats with PV ( n = 7) ranged from 70% to 8 1.4%
(median 76.5%), and the serum Epo concentration ranged
from 9 to 92 mU/mL (median 10 mU/mL). There were 2
cats with secondary polycythemia; both had renal carcino-
mas. There were no significant differences in the rank sum
of serum Epo concentrations of either group when com-
pared with normal cats. The rank sum of serum Epo con-
centration was not significantly different in cats with PV
compared with cats with secondary polycythemias.
Serum Erythropoietin Concentrations in Anemia
Dogs. The clinical information and serum Epo concentra-
tions for anemic dogs is presented in Table 3. The K V in dogs
with anemias of nonrenal failure origin (n = 13) ranged from
8% to 27% (median 14%) and the serum Epo concentrations
ranged from 22 to 320 mU/mL (median 50 mU/mL). The
rank sum of serum Epo concentration was significantly in-
creased (P< .05) in dogs with nonrenal failure anemias when
20 COOK AND LOTHROP

Table 1. Summary of Clinical Findings in 30 Dogs With Polycythemia

Serum
Dog Age PCV Epo Po2
No. Breed (years) Sex (%) (mU/mL) (mmHg) Diagnosis Treatment Outcome
1 Bernese Mountain 3 MC 76.0 34.8 NA PV Phlebotomy Controlled with
Dog Hydroxyurea therapy
2 Cairn Terrier 11 FS 71.8 13.2 99 PV Phlebotomy Controlled with
Hydroxyurea therapy
3 Springer Spaniel 9 M 72.8 13.4 NA PV Phlebotomy Died
4 Scottish Terrier 4 MC 64.3 c10.0 80 PV Phlebotomy Controlled
Hydroxyurea
5 Beagle 2 F 73.0 21.4 75 PV Hydroxyurea Controlled
6 German shepherd 10 MC 85.0 57.3 NA PV Cytopheresis Controlled
dog Hydroxyurea
7 Basset 9 F NA 13.5 NA PV NA NA
8 Akita-mix 4 FS NA 21.9 NA PV NA NA
9 Shetland 12 FS 65.0 39.2 62 Idiopathic pleural Symptomatic NA
Sheepdog effusion
10 Irish Setter 6.5 FS 83.0 30.7 31 Pulmonary Phlebotomy Died
arteriosclerosis Prednisone
11 Cocker 14 MC 68.0 13.3 74 COPD None NA
12 Miniature 0.25 M 57.0 23.7 NA R to L shunt; VSD Propranolol Signs controlled with
Schnauzer pulmonic stenosis medication
13 Cocker Spaniel 1 M 75.2 19.4 28 Tetralogy of Fallot None Dead
14 Doberman 5 F 58.0 102.9 NA Bilateral renal LSA Doxorubicin Epo decreased to
40.9 with therapy
15 Brittany Spaniel 4 MC 63.2 35.9 109 Histiocytic LSA- Nephrectomy Euthanatized
renal, nasal, skin Chemotherapy
16 Mix 11 M 70.0 35.9 86 Testicular neoplasia Castration Died (unrelated)
(seminoma) Phlebotomy
17 Dachshund 7 FS 65.0 21.3 88 Allergic rhinitis None NA
18 Poodle 12 FS 62.0 45.7 90 Nasal Rhinotomy, NA
hemangiosarcoma interstitial
brachytherapy
19 Dalmatian 1.5 MC 72.0 19.0 NA GI foreign body Enterotomy Resolved
20 Collie 13 FS 66.0 11.7 NA Chronic nasal None NA
discharge
21 Siberian Husky 12 F 78.0 131.7 90 Diabetes mellitus Phlebotomy Poor response
(poorly regulated) Hydroxyurea
22 Collie 3 F 53.0 9.9 NA Uncertain Phlebotomy NA
23 Staffordshire 9 FS 72.0 <10.0 97 Mast cell tumors None Resolved
Terrier
24 Dachshund 9 FS 65.0 16.6 NA Historicalhemolytic None Resolved
anemia
25 Dachshund 3 FS 60.0 8.1 NA Undetermined None Resolved
26 Scottish Terrier 9 FS 62.0 13.3 89 UNC None Resolved
27 Dachshund 4 F 75.0 <12.0 NA UNC None Resolved
28 Boxer 5 MC 65.2 12.2 NA UNC None Resolved
29 Schnauzer 3 F 64.4 <10.0 99 UNC None Resolved
30 Schnauzer 5 FS 73.8
. ~ 22.9 NA UNC
~.- Phlebotomy Resolved
Abbreviations: M, male; F, female; MC, castrated male; FS,spayed female; Epo, erythropoietin; PV, polycythemia Vera; NA, not available; LSA,
lymphosarcoma; COPD, chronic obstructive pulmonary disease; UNC, uncertain; VSD, ventricular septa1 defect; GI, gastrointestinal.

compared with normal dogs. Serum samples from dogs with
anemias caused by chronic renal failure were not evaluated.
In the 2 normal dogs with chemically induced anemia,
the serum Epo concentration increased as the PCV de-
creased, with the Epo peak occumng at the PCV nadir (Fig
1). The correlation coefficient (r value) for the PCV and se-
rum Epo concentrations in these 2 dogs was 0.83.
Cuts. The clinical information and serum Epo con-
centrations for anemic cats are presented in Table 4.
Cats with anemia of nonrenal failure origin ( n = 5) had
PCVs ranging from 14.0% to 19.0% (median 16.5%), and
serum Epo concentrations ranging from 17.4 to 207.3
mU/mL (median 38.7 mU/mL). The rank sum of serum
Epo concentration in cats with nonrenal failure anemias
was significantly increased (P< .05) when compared to
normal cats. Cats with anemias caused by chronic renal
disease ( n = 5) had PCVs ranging from 20% to 23.2%
(median 2 1%) and serum Epo concentrations ranging
from 9 to 42 mU/mL (median 28 mU/mL). The rank
sum of serum Epo concentration in cats with anemias
 SERUM ERYTHROPOIETIN IN DOGS AND CATS 21

Table 2. Summary of Clinical Findings in 10 Cats With Polycthemia

Cat Age PCV SerumEpo Po2
No. Breed (years) Sex (%) (mU/mL) (mm Hg) Diagnosis Treatment Outcome
1 DSH 5 MC 78 27.3 NA PV Phlebotomy Controlled
Hydroxyurea
2 DSH 7 MC 78 91.1 NA PV Phlebotomy Controlled
Hydroxyurea
3 DSH 3 MC 81 20.7 NA PV Phlebotomy Controlled
4 DSH 15 FS 75.6 9.z NA PV Phlebotomy Controlled
5 DLH 12 MC NA <9 NA PV NA NA
6 DSH 10 MC 70 9.6 NA PV Phlebotomy NA
7 DSH 5 FS 75 9.7 75 PV Phlebotomy DeveIoped
coagulopathy
8 DSH 11 MC 65 10.9 95 Renal carcinoma Nephrectomy Resolved
9 DSH 9 MC 65 26.5 NA Renal carcinoma Nephrectomy Resolved
10 DSH 11 MC 54 17.7 NA Allergic bronchitis NA NA
Abbreviations: DSH, Domestic Shorthair; DLH, Domestic Longhair; MC, castrated male; FS, spayed female; Epo, erythropoietin;PV, polycythemia
Vera; NA, not available

caused by chronic renal disease were not significantly
different from those of normal cats.
Discussion
The measurement of serum Epo concentrations by RIA
may be a valuable and practical diagnostic tool in the eval-
uation of polycythemias and anemias in both dogs and
' , I 2 Overlap in serum Epo determinations between
dogs and cats with PV and secondary polycythemias, and
normal dogs and cats indicates that judicious interpretation
of Epo values is necessary. Serum Epo concentrations were
significantlydecreased in dogs with PV compared with dogs
with secondary polycythemias. In cats, the results demon-
strated a trend toward low serum Epo concentrations in PV,
although statistical significance was not shown. A low serum
Epo concentration is the expected finding in PV because it
is a myeloproliferative disorder of erythropoiesis that is in-
dependent of Epo p r o d ~ c t i o n . However,
'~ 2 of the 8 dogs
with PV (Dogs 1 and 6) and 3 of the 7 cats with PV (cats
1, 2, and 3) did not have low serum Epo concentrations as
expected. One possible explanation for this is misdiagnosis,
although we believe this unlikely. A second explanation is
that serum Epo concentrations may fluctuate in dogs and
cats. A circadian rhythm of Epo secretion has been docu-
mented in healthy human beings," and it has been pro-
posed that Epo secretion may be intermittent." Lastly, it is
possible that the serum samples submitted for Epo determi-
nation were obtained after phlebotomy (unfortunately in-
formation regarding timing of serum samples was not avail-
able), which may stimulate Epo production, thereby
resulting in a misleadingly increased serum Epo concentra-
tion.20Therefore, it is important to obtain serum samples
for Epo determinations in polycythemic animals prior to
phlebotomy. Additionally, when PV is suspected based on
clinical findings but is not supported by a low serum Epo
concentration, it may be useful to repeat serum Epo mea-
surements.
Our results are in agreement with preliminary findings us-
ing RIAs in dogs, in which 6 dogs with PV were found to
have low or low-normal (10 mU/mL) serum Epo concen-
trations." Before the increased availability of RIAs for Epo
determinations, there was little known regarding Epo con-
centrations in dogs or cats with polycythemia Vera. Epo con-
centrations were not detectable by an in vivo bioassay in
3 dogs and 1 cat with polycythemia vera.2'*22Our results
pertaining to PV are in agreement with findings in human
beings with PV.23-25Polycythemia Vera is a rather well-
defined syndrome in human beings, and certain criteria put
forth by the National Polycythemia Vera Study Group must
be fulfilled before a diagnosis of PV is made.26 However,
serum Epo determinations have proved useful in the differ-
ential diagnosis of some people with p~lycythernia.~~-'~ Be-
cause most of the criteria for diagnosis of PV in human be-
ings have not been recognized in dogs and cats (eg,
hepatomegaly, splenomegaly, leukocytosis, and thrombo-
cytosis), serum Epo determinations could have increased
value in the differential diagnosis of dogs and cats with poly-
cythemia.2'
In the present study, 5 dogs had secondary polycythemia
caused by systemic hypoxia. Serum Epo concentrations are
expected to be increased in such cases, representing an ap-
propriate physiological response. However, serum Epo con-
centrations were normal to only moderately increased in all
5 of these dogs. As discussed for the dogs with PV, daily
fluctuations in Epo secretion and obtaining serum samples
for Epo determinations after phlebotomy are possible expla-
nations for these results. Additionally, variability in the se-
rum Epo concentrations because of the duration and sever-
ity of the hypoxemia may be present. The value of serum
Epo concentrations appears to be diagnostically limited in
such dogs because hypoxemia provides the clinician with an
explanation for the polycythemia.
The remaining animals in the category of secondary poly-
cythemias were 2 dogs and 2 cats with renal tumors. The 2
dogs had renal lymphosarcoma. One of these dogs had an
increased serum Epo (102.9 mU/mL), which decreased to
 Table 3. Summarv of Clinical Findings in 13 Dogs With Anemia

Treatment 0
~ ~ ~

Dachshund 4 MC 11 .o NA 46.5 NA Pure red cell aplasia Prednisone NA

Doberman cross 6 FS 8.8 0.0 320.0 NA Pure red cell aplasia Transfusion, prednisone, Eu
cyclophosphamide
Mix 6 M 11.0 1.6 21.5 NA Pure red cell aplasia Transfusion, prednisone, Re
nandrolone
abrador 9 FS 15.0 0.0 46.7 NA Myelofibrosis Prednisone NA
Dalmatian 10 FS NA NA 42.7 NA AlHA Prednisone Eu
Dachshund 7 MC 11.4 0.3 220.0 NA Pure red cell aplasia Azathiopnne Di
cyclophosphamide
German shepherd 6 FS 8.0 0.1 30.8 NA idiopathic None Re
Dog nonregenerative
anemia
Miniature Poodle 6 M 12.9 12.6 58.6 NA AIHA, pulmonary NA Di
thromboembolism
Chow 2 F 15.0 NA 42.8 NA AlHA Prednisone Re
cyclophosphamide
Basenji 3 M 21 .o 19.7 50.0 NA Pyruvate kinase def. None Al
Basenji 3 M 17.0 30.0 50.0 NA Pyruvate kinase def. None Di
Mix 1 F 27.0 18.0 53.6 NA Phenylhydrazine None Re
induced anemia
Mix 1 F 26.0 20.0 55.0 NA Phenylhydrazine None Re
induced anemia

ns: M, male; F, female; MC, castrated male; FS, spayed female; Epo, erythropoietin; NA, not available; AIHA, autoimmune hemolytic anemia; % Ret, reticulocyte %.
 SERUM ERYTHROPOIETIN IN DOGS AND CATS 23

Po mal. A possible explanation is the production of another

hormone such as hematopoeitic growth factor not detected
by the Epo RIA. Inappropriate Epo production in associa-
tion with renal tumors in and tumor-associated
erythrocytosis in a dog with a nasal fibro~arcoma~' have
been previously reported. In human patients with renal neo-
plasia, an inappropriate secretion of Epo often is present,
which generally resolves after surgical resection of the renal
tumor if metastases are not present.32
In the category of polycythemias of uncertain origin, the
0
0 1
. 2 , 3 .4 .5 6. 7 . 8 ,9 1
.
-m- €Po
,
0 1
,
1
.
1 2
0
results of serum Epo determinations were variable and did
not allow differentiation of primary and secondary polycy-
-I Time (Days) themia. In addition, many of these dogs may have had rela-
Phenylhydrazine
Treatment
tive rather than absolute polycythemias. Red blood cell
(RBC) mass determination^^^ would have better defined the
Fig 1. Serial serum erythropoietin concentrations (mU/mL) and classification of polycythemic animals in this study but were
packed-cell volumes (%) in 2 adult female dogs given phenylhydra- not done because of the study's retrospective nature. Mea-
zine (2.5 mg/kg IV SID for 4 days) to induce anemia and reticulo-
cytosis. Open circle, closed circle-packed cell volume; open square,
surement of RBC mass can differentiate relative (normal
closed square-erythropoietin. RBC mass) from absolute (increased RBC mass) polycythe-
mia, although these two conditions often may be distin-
guished from one another clinically by the presence or ab-
sence of signs of dehydrati~n.~ Because normal serum Epo
40.9 mU/mL after 10 days of chemotherapy and resultant values were detected in dogs with PV, it is possible that some
clinical remission. The other dog had normal Epo concen- of the dogs included in the uncertain etiology category may
tration (35.9 mU/mL), which increased after nephrectomy have had PV or secondary polycythemia, although it is not
(48.7 mU/mL). The lymphosarcoma in this dog involved likely that the dogs with spontaneous resolution of their
multiple organ systems (kidney, nasal cavity, skin), and polycythemia (dogs 23-30) had PV or secondary polycythe-
there was progression of the disease despite nephrectomy mia.
and chemotherapy. The serum Epo concentrations in 2 In the absence of renal disease in human beings, erythro-
polycythemic cats with renal adenocarcinomas were nor- poietin production is related primarily to the h e m a t ~ c r i t . ~ ~

Table 4. Summary of Clinical Findings in 10 Cats With Anemia

Cat Age PCV % SerumEpo Po,
No. Breed (years) Sex (%) Ret (mU/mL) (mmHg) Diagnosis Treatment Outcome
11 Siamese 12 MC 18 0.2 207.3 NA GI LSA rHuEpo No response
12 DLH 8 MC 14 0.0 38.9 NA FeLV Supportive Controlled
13 DSH 9 FS 15 NA 38.5 NA GI bleed Resection of FeLV positive
bleeding pancytopenia
ulcer
14 DLH 7 MC 8 1.5 22.2 NA Hepatic lipidosis Transfusion Euthanatized
15 DSH 5 FS 19 NA 17.4 NA Vitamin K- Vitamin K NA
responsive
coagulopathy
16 DSH 16.5 FS 21 0.6 28.1 NA CRF rHuEpo Hematocrit
increased to
normal
17 DSH 19.5 MC 22 0.1 31.2 NA CRF Nandrolone, B Euthanatized
vitamins
18 DSH 20.2 MC 20 0.9 17.9 NA CRF, LSA, diabetes Chemotherapy Euthanatized
supportive
care
19 DSH 12.0 MC 21 0.3 42.0 NA CRF Symptomatic Euthanatized
care
20 DSH 7 MC 23.2 NA 19 NA CRF Low protein Control
diet
Abbreviations: MC. castrated male; FS, spayed female; Epo, erythropoietin; DLH, Domestic Longhair; DSH, Domestic Shorthair; GI, gastrointestinal;
LSA, lymphosarcoma; rHuEpo, recombinant human erythropoietin; FeLV. feline leukemia virus; NA, not available; CRF. chronic renal failure; % Ret,
reticulocyte %.
24
Our findings tend to support this concept in anemic dogs
and cats. In the 2 dogs with chemically induced anemia, the
serum Epo concentration increased as the PCV decreased (r
value of0.83), with the Epo peak occumng at the PCV nadir
(Fig 1). Additionally, 3 of the 4 dogs with pure red cell
aplasia (diagnosis based on cytologic examination of bone
marrow aspirates) had increased serum Epo concentrations
(46.5, 220, and 320 mU/mL). Similar findings have been
previously r e p ~ r t e d The
. ~ remaining dog had a serum Epo
value that was normal. Similar to dogs with pure red cell
aplasia, it has been previously reported that cats with feline
leukemia virus-induced red cell aplasia had high serum Epo
concentration^.'^ The 5 cats with nonrenal failure anemias
in our study had serum Epo concentrations that varied from
normal to increased.
The finding of increased Epo concentrations in animals
with erythroid aplasia suggests that the bone marrow is un-
able to respond and the patient would not likely benefit
from rHuEpo replacement therapy. Cat 11 with anemia as-
sociated with GI lymphosarcoma had an increased serum
Epo concentration and did not improve after treatment with
rHuEpo. By contrast, cat 16 with an anemia secondary to
chronic renal failure and a normal serum Epo concentration
responded to treatment with rHuEpo. The hematocrit re-
turned to the normal range.
Four of the 5 cats with anemia caused by chronic renal
failure evaluated in this study had normal serum Epo val-
ues. One possible explanation is that a relative Epo defi-
ciency is present (ie, an inappropriate Epo response for the
degree of anemia present).36Similarly, human patients with
chronic renal failure may have normal to increased serum
Epo concentration^.^^ The response of cat 16 to rHuEpo
supports this concept.
In summary, the measurement of serum Epo concentra-
tion by RIA yields variable results in dogs and cats with
polycythemias and anemias. It is important to obtain serum
samples for Epo determination before therapy of these dis-
orders, especially phlebotomy or transfusion. It is also im-
portant that the determinations are done using an RIA that
is known to detect canine and feline Epo. Further studies
using this particular RIA to evaluate serum Epo concentra-
tions in larger numbers of animals with these conditions is
needed to define its diagnostic value.
In polycythemic animals, serum Epo determinations
compliment a thorough diagnostic work-up in the differen-
tial diagnosis of primary and secondary polycythemias. An
increased serum Epo concentration in an animal with unex-
plained polycythemia should prompt a search for neoplastic
disease, especially renal. Finally, the evaluation of serum
Epo concentrations in anemic animals may help to assess
the contribution of renal disease to the anemia as well as
to screen for those animals that may benefit from rHuEpo
replacement therapy.
References
1. Krantz SB. Erythropoietin. Blood 1991;77:419-434.
2. Erslev AJ. Erythropoietin. Leuk Res 1991;14:683-688.
COOK AND LOTHROP
3. Abels R. Review of hematological effects of erythropoietin.
Semin Nephrol 1990;1O(suppl): 1-10,
4. Graber SE, Krantz SB. Erythropoietin: Biology and clinical
use. Hematol Oncol Clin North Am 1989;3:369-400.
5. Giger U. Erythropoietin and its clinical use. Comp Cont Ed
Pract Vet 1992;14:25-34.
6. Ikeda T, Inaba M, Maede Y. Serum erythropoietin level in
normal dogs. Jpn J Vet Sci 1990;52:877-878.
7. Dunn CDR, Jones JB, Jolly JD, Lange RD. Progenitor cells in
canine cyclic hematopoiesis. Blood 1977;50:11 11-1 120.
8. Jacobs K, Shoemaker C, Rudersdorf R, et al. Isolation and
characterization of genomic and cDNA clones of human erythro-
poietin. Nature 1985;313:806-8 10.
9. Lin FK, SUES S, Lin CH, et al. Cloning and expression of
the human erythropoietin gene. Proc Natl Acad Sci USA 1985;82:
7580-7584.
10. Egrie JC, Cotes PM, Lane J, et al. Development of radioim-
munoassay for human erythropoietin using recombinant erythro-
poietin as tracer and immunogen. J Immunol Methods 1987;99:
235-241.
1 1. Giger U. Serum erythropoietin concentrations in polycythe-
mic and anemic dogs. Proc 9th ACVIM Forum, New Orleans, LA,
199I: 143-145.
12. Petrites-Murphy FMB, Pierce KR, et al. Effect of incorpora-
tion of serum from dogs with renal impairment on canine erythroid
bone marrowcultures. Am J Vet Res 1989;50:1537-1543.
13. Jain NC. Schalm’s Veterinary Hematology. 4th ed. Philadel-
phia, PA: Lea and Febiger 1986.
14. Brock KV, Jones JB, Shull RM, et al. Effect of canine parvov-
irus on erythroid progenitors in phenylhydrazine-induced regenera-
tive hemolytic anemia in dogs. Am J Vet Res 1989;50965-969.
15. Daniel WW. Biostatistics. A foundation for analysis in the
health sciences. New York, NY: Wiley, 1974:346-351.
16. Daniel WW. Biostatistics. A foundation for analysis in the
health sciences. New York, NY: Wiley, 1974:253-262.
17. Adamson J, Fialkow PJ, Murphy S, et al. Polycythemia Vera:
Stem-cell and probably clonal origin of the disease. N Engl J Med
1976;295:913-916.
18. Wide L, Bengtsson C, Birgegard G. Circadian rhythm of
erythropoietin in human serum. Br J Hematol 1989;75:85-90.
19. Cote PM, Dore CJ, Liu Yin JA, et al. Determination of se-
rum immunoreactive erythropoietin in the investigation of eryth-
rocytosis. N Engl J Med 1986;315:283-287.
20. Adamson JW. The erythropoietin/hematocrit relationship
in normal and polycythemic man: Implications of marrow regula-
tion. Blood 1968;32:597-609.
21. Peterson ME, Randolph JF. Diagnosis of canine primary
polycythemia and management with hydroxyurea. J Am Vet Med
Assoc 1982;1804 15-4 18.
22. Foster ES, Lothrop CD. Polycythemia Vera in a cat with
cardiac hypertrophy. J Am Vet Med Assoc 1988;192:1736-1738.
23. Erslev AJ, Caro J, Kansu E, et al. Plasma erythropoietin in
polycythemia. Am J Med 1979;66:243-247.
24. Garcia JF, Ebbe SN, Hollander L, et al. Radioimmunoassay
of erythropoietin: Circulation levels in normal and polycythemic
human beings. J Lab Clin Med 1982;99:624-635.
25. Koeffler HP, Goldwasser E. Erythropoietin radioimmunoas-
say in evaluating patients with polycythemia. Ann Intern Med
1981;94:44-47.
26. Murphy S. Polycythemia Vera. In: Williams WJ, Beutler E,
Erslev AJ, et al, eds. Hematology. 3rd ed. New York, NY: McGraw
Hill, 1983:185-196.
27. Scott RC, Patnaik AK. Renal carcinoma with secondary
polycythemia in the dog. J Am Anim Hosp Assoc 1972;8:275-283.
SERUM ERYTHROPOIETIN IN DOGS AND CATS
28. Peterson ME, Zanjani ED. Inappropriate erythropoietin pro-
duction from a renal carcinoma in a dog with polycythemia. J Am
Vet Med Assoc 1981;179:995-996.
29. Nelson RW, Hager D, Zanjani ED. Renal lymphosarcoma
with inappropriate erythropoietin concentration in a dog. J Am Vet
Med Assoc 1983;182:1396-1397.
30. Waters DJ, Preuter JC. Secondary polycythemia associated
with renal disease in the dog: Two case reports and review of litera-
ture. J Am Anim Hosp Assoc 1988;24:109-1 14.
31. Couto CG, Boudrieau RJ, Zanjani ED. Tumor-associated
erythrocytosis in a dog with nasal fibrosarcoma. J Vet Intern Med
1989;3:183-185.
25
32. Altaffer LF, Chenault OW. Paraneoplastic endocrinopathies
associated with renal tumors. J Urol 1979;122573-577.
33. Alazraki NP, Mishkin FS,eds. Fundamentals of Nuclear Med-
icine. 2nd ed. New York, NY: Society of Nuclear Medicine, Inc 1988.
34. Erslev AJ, Wilson J, Car0 J. Erythropoietin titers in anemic,
nonuremic patients. J Lab Clin Med 1987;109:429-433.
35. Kociba GJ, Lange RD, Dunn CDR, et al. Serum erythropoi-
etin changes in cats with feline leukemia virus-induced erythroid
aplasia. Vet Pathol 1983;20:548-552.
36. Eschbach JW. The anemia of chronic renal failure: Patho-
physiology and the effects of recombinant erythropoietin. Kid Int
1989;35:134-148.