The Diagnosis of Feline Heartworm Infection
Clarke Atkins, DVM, Introduction
Diplomate ACVIM
(Internal Medicine & Cardiology)
From the Department of Companion Animal
and Special Species Medicine,
College of Veterinary Medicine,
North Carolina State University,
4700 Hillsborough Street,
Raleigh, North Carolina 27606.
JOURNAL of the American Animal Hospital Association
The diagnosis of heartworm infection (HWI) or disease (HWD) in cats is
more difficult to determine than in dogs and, hence, a more aggressive
diagnostic approach is necessary. The algorithms below demonstrate the
author’s approach to routine screening and when HWI is suspected.
Clinical signs in cats are most often respiratory in nature (cough,
dyspnea), but vomiting, neurological dysfunction, and malaise are also
reported.1–4 Cats differ from dogs in that sudden death may be the first
sign of HWD and heart failure is unusual.3 The finding of dyspnea,
chronic and unexplained vomiting, and especially cough in cats that are in
heartworm (HW) endemic areas, should prompt the suspicion of HWD.
Screening
Testing prior to institution of preventative medication in cats is not
imperative because most are amicrofilaremic and, if microfilaria are
present, numbers are small. Because of this, and because macrolides are
used as preventatives, the chance of adverse reactions is very unlikely.
Antigen testing is insensitive in cats, and false negative results may later
give the impression of prophylactic failure in cats actually infected prior
to preventative but which tested antigen-negative. On the other hand,
antibody testing (see first algorithm) provides information as to past
exposure. If the cat is antibody-negative, the risk for current HWI is very
low. If positive, there is risk of current (and likely future) HWI, and the
cat should definitely receive preventative. This author advises antibody
testing prior to administration of preventative, but does not require it.
Diagnosis in Suspected Cases
Immunodiagnostic methods are imperfect in this species because of the
low worm burdens (one to 12, mean = three)2 and hence, light antigenic
load in cats.5 Enzyme-linked immunosorbent assay (ELISA) for heart-
worm antigen were positive on sera from 36% to 93% of 31 cats harboring
one to seven female HW, with sensitivity increasing as female worm
burden increased. 5 Cats with male worm(s) were not detected as positive.
Therefore, false negative tests occur frequently, depending on the test
used, maturity and gender of worms, and worm burden. All tests were
100% specific. Data on sensitivity in natural infections is scarce, but
antigen tests probably identify only 25% to 40% of actual infections.6–8
Antigen tests are, nevertheless, useful because they allow a definitive
diagnosis of HWI and should be used whenever the disease is suspected
(see second algorithm).
Though less specific, heartworm antibody tests (ELISA) are useful in
detecting exposure and larval development to the fourth larval (L4)
185
186 JOURNAL of the American Animal Hospital Association May/June 1999, Vol. 35

Routine Screening of Cats for Heartworm Infection

Antibody Test

➔
Positive
Cat exposed to heartworms.
Consider antigen test.
Can start macrolide* preventative.
Negative
Cat has no recent exposure.
Start macrolide preventative.

Antigen Test
➔

Positive
Cat has heartworm infection.
Start macrolide preventative
and symptomatic therapy, as
needed.
➔ Negative
Cat may have heartworm infection.
Perform CBC,† thoracic radiography,
echocardiography ± microfilaria test;
see next algorithm. Start macrolide
preventative and symptomatic therapy, as needed.

Diagnostic Procedure in Cats Suspected of Heartworm Infection

Antibody Test
➔

Positive
Cat exposed to heartworms.
Perform antigen test, microfilarial
test, CBC, thoracic radiographs,
and echocardiography.
➔
Negative
Cat probably does not have heartworm
infection (small percent false negative).
Start macrolide preventative.

Antigen, Echo,‡ or Antigen, Echo, and

Microfilaria-Positive
Cat has heartworm infection.
Start macrolide preventative and
symptomatic therapy, as needed.
Microfilaria-Negative
Cat unlikely has heartworm infection.
However, if CBC, radiographs, and clinical signs
are suggestive, a presumptive diagnosis of
heartworm infection can be made. Heartworm infection
is difficult to rule out definitively, especially in an
antibody-positive cat. Start macrolide preventative.

Algorithms adapted with permission from Atkins CE. Heartworm disease: an update on testing and prevention in dogs
and cats. The Veterinary CE Advisor, Vet.Med. (Suppl.), 1998;93(12):2–18.
* Macrolide=ivermectin or milbemycin
†
CBC=complete blood count
‡
Echo=echocardiography

stage. 7,9,10 In addition to feline heartworm antibody tests
performed by commercial laboratories, there are now
two in-house antibody test kits. a,b The antibody tests are
excellent for screening prior to placing cats on preventa-
tive; for epidemiological screening of risk for heartworm
infection (tests demonstrate exposure, hence risk); and
for adding credence to clinical suspicion of HWD in
cats. It is important to realize that up to 90% of antibody-
positive cats ultimately resist natural infection and that,
of the serological tests, only antigen tests allow a defini-
tive diagnosis of active infection. It is generally accepted
that an antibody-negative cat does not have HWI. Never-
theless, studies at North Carolina State University and
elsewhere suggest that false negative antibody tests do
occur and may exceed 10%. 3,11 This means that in cats
with signs compatible with HWD, a negative antibody
May/June 1999, Vol. 35 Feline Heartworm Infection 187

test should not always deter further investigation (see 5. McTier TL, Supakorndej N, McCall JW, Dzimianski MT. Evaluation of
ELISA-based adult heartworm antigen test kits using well-defined sera
second algorithm). from experimentally and naturally infected cats. Proc Am Assoc Vet Parasit
Eosinophilia (particularly greater than 2,300/µl) is (abst 45) 1993;38:37.

supportive of HWI in cats. 3 Thoracic radiographs are
quite useful, representing a reasonable screening test for
feline HWD (i.e., in symptomatic patients). The findings
may be subtle; enlarged caudal pulmonary arteries
(greater than 1.6 times the width of the ninth rib at the
ninth intercostal space), often with ill-defined margins,
represent the most common finding. Pulmonary paren-
chymal changes include focal or diffuse infiltrates,
perivascular density, and occasionally, atelectasis and
hyperinflation.3,10,12,13 While unable to provide a defini-
tive diagnosis alone, when coupled with appropriate clini-
cal signs, a positive antibody test, or both, HWD can be
diagnosed radiographically. Pulmonary angiography has
also been utilized to demonstrate radiolucent linear in-
travascular “foreign bodies” as well as enlarged, tortu-
ous, and blunted pulmonary arteries.
Echocardiography, in the author’s experience, is more
sensitive in cats than in dogs, with detection of up to
78% of cases with prospective evaluation. 10 A sensitivity
approaching 100% has been attained in hyperendemic
areas, 8 but lesser sensitivity has been described for retro-
6. Guerrero J, McCall JW, Dzimianski MT, et al. Prevalence of dirofilaria
immitis infection in cats from the southeastern United States. In: Soll MD,
ed. Proc Am Heartworm Sym ’92. Batavia, IL: Am Heartworm Society,
1992:92–5.
7. McCall JW, Nonglak S, Ryan W, Soll MD. Utility of ELISA-based
antibody test for detection of heartworm infection in cats. In: Soll MD,
Knight DH, eds. Proc Am Heartworm Sym ’95. Batavia, IL: Am
Heartworm Society, 1995:127–33.
8. Genchi C, Kramer L, Venco L, et al. Comparison of antibody and antigen
testing with echocardiography for detection of heartworm (dirofilaria
immitis) in cats. In: Soll MD, Knight DH, eds. Proc Am Heartworm Sym
’98. Batavia, IL: Am Heartworm Society, 1998:in press.
9. Stewart VA, Hepler DI, Grieve RB. Efficacy of milbemycin oxime in
chemoprophylaxis of dirofilariasis in cats. Am J Vet Res 1992;53:2274.
10. Atkins CE, DeFrancesco TD, Miller MW, et al. Prevalence of heartworm
infection in cats with signs of cardiorespiratory abnormalities. J Vet Med
Assoc 1997;212:517–20.
11. Dillon AR, Brawner WR, Robertson CW, Guerrero J. Feline heartworm
disease: correlations of clinical signs, serology, and other diagnostics:
results of a multicenter trial. In: Soll MD, Knight DH, eds. Proc Am
Heartworm Sym ’98. Batavia, IL: Am Heartworm Society, 1998:in press.
12. Schafer M, Berry CR. Cardiac and pulmonary artery mensuration in feline
heartworm disease. Vet Radiol & Ultrasound 1995;36:499–505.
13. Selcer BA, Newell SM, Mansour MS, McCall JW. Radiographic and 2-D
echocardiographic findings in eighteen cats experimentally exposed to D.
immitis via mosquito bites. Vet Radiol & Ultrasound 1996;37:37–44.
14. DeFrancesco TD, Atkins CE, Miller MW, et al. Diagnostic utility of
echocardiography in feline heartworm disease. J Vet Med Assoc, in press.

spective (approximately 50%)14 and prospective studies

(78%)10,13 in the United States. It is clear that careful
examination is necessary, that the pulmonary arteries
must be interrogated, and that a high index of suspicion
is helpful.14 Typically, a “double-lined echodensity” is
evident in the main pulmonary artery or one of its
branches (71%); the right ventricle (41%) or, occasion-
ally, the right atrioventricular junction; or the caudal
vena cava (6%).14 Because of its high specificity, a posi-
tive echocardiographic finding is considered to provide a
definitive diagnosis.
Ultimately, the diagnosis of HWI can be made defini-
tively with a positive result for microfilaria test, antigen
test, or echocardiographic detection. A presumptive di-
agnosis can be made with some combination of sugges-
tive clinical signs, hemogram changes, radiographic
findings, and a positive antibody test.

a
ASSURE/FH; Synbiotics Corporation, San Diego, CA
b
Solo Step PH; HESKA, Fort Collins, CO

References
1. Henry CJ, Dillon R. Heartworm disease in dogs. J Am Vet Med Assoc
1994:1148.
2. Harpster NK. The cardiovascular system. In: Holzworth J, ed. Diseases of
the cat. Vol 1. Philadelphia: WB Saunders, 1987.
3. Atkins CE, DeFrancesco TD, Coats J, Keene BW. Feline heartworm
disease: the North Carolina experience. In: Soll MD, Knight DH, eds. Proc
Am Heartworm Sym ’98. Batavia, IL: Am Heartworm Society, 1998: in
press.
4. Holmes RA. Feline heartworm disease. Compend Contin Ed Pract Vet
1993;15:687.
 Risk Factors for Odontoclastic
Resorptive Lesions in Cats
A cross-sectional study evaluating potential risk factors for odontoclastic resorptive lesions
(ORL) in feline teeth was conducted. Owners of 32 cats with ORL and 27 cats without ORL
were interviewed regarding their respective cat’s demographic characteristics, diet, and medical
and dental histories. Four factors were identified as significantly associated with ORL using
unconditional logistic regression. A history of dental disease (gingivitis, calculus, or periodontal
disease; odds ratio [OR], 4.5); city residence (OR, 4.4); and being an exclusively indoor cat
(OR, 4.5) were associated with an increased risk for ORL. Consumption of commercial treats
(OR, 0.3) appeared protective for ORL. J Am Anim Hosp Assoc 1999;35:188–92.

Janet M. Scarlett, DVM, PhD Introduction

John Saidla, DVM Tooth destruction from odontoclastic resorptive lesions (ORL) results in a
painful dental condition of cats. Odontoclasts attack and begin to resorb
Jennifer Hess, BS teeth, either externally or internally.1,2 Externally they attack at the neck
or cervical area of the tooth, often extending into the tooth root and
surrounding alveolar area. Internally, they attack the tooth through the
RS apical foramen and can subsequently enter into the root and dentin. In the
early stages, cats are asymptomatic; but as the disease progresses and
dentinal tubules and/or pulp are exposed, the lesions become increasingly
sensitive and painful. Cats may present with a history of poor appetite or
anorexia, eating on one side of the mouth, weight loss, and depression.3–5
External lesions may be difficult to identify if they remain hidden beneath
the gingiva or calculus. As the lesions progress, they are often associated
with gingival inflammation, bleeding or hyperplasia, and increasing loss
of tooth substance. Ultimately, part or all of the affected teeth are lost.
Characterizing the full extent of root and alveolar bone involvement
requires dental radiographs.4,6
Odontoclastic resorptive lesions are common, affecting 20% to 67% of
cats receiving dental care,7 and the prevalence varies with the characteris-
tics of the population studied. In a recent study of cats undergoing a wide
range of procedures requiring anesthesia in a private veterinary practice
in Wisconsin, 48% of cats examined had one or more ORL;8 this com-
pares closely with an estimate of 52% for a similar sample in Australia.9
The frequency is probably higher, since many lesions are subgingival and
difficult to identify, and missing teeth may have been lost because of
ORL.7 Despite the variability in prevalence estimates, the frequency of
ORL appears to have increased over time. A recent reexamination of the
teeth in 80 cat skulls examined originally by Cloyer in 1936 found one
skull (1.3%) with ORL;7 a pre-1950 skull series had a 0.4% prevalence,
while a post-1970 skull series had 26.5% of skulls with lesions. 10 This
apparent increase in the prevalence, however, may reflect changes in the
age distribution of the cat populations examined.10
From the Department of Population Despite a high prevalence, the cause(s) of ORL is unknown. The
Medicine and Diagnostic Sciences
lesions were originally confused with caries11 which result from deminer-
(Scarlett, Saidla),
College of Veterinary Medicine (Hess), alization of the tooth enamel by toxins produced by carbohydrate-fer-
Cornell University, menting bacteria. Unlike caries, however, ORL occur when activated
Ithaca, New York 14853. odontoclasts resorb tooth and bone tissue. In response, bone and cemen-

188 JOURNAL of the American Animal Hospital Association

May/June 1999, Vol. 35
tum tissue initiate remodeling, odontoclastic activity con-
tinues, and the process repeats itself until ankylosis (i.e.,
fixation of the root in the alveolus through destruction of
the periodontal ligament and lamina dura) occurs, the
tooth is permanently damaged, and the root is fixed into
the alveolar bone.12
Odontoclasts are normally active only in young ani-
mals in resorption of the roots of deciduous teeth, facili-
tating loss of these teeth and making room for the
permanent dentition. In cats with ORL, these cells are
inappropriately stimulated to differentiate and become
active for reasons that are not well understood. The
apparent increase in the prevalence suggests some
change(s) in the environment of cats that influences
mechanisms that affect the odontoclasts, and the search
for some environmental risk factor(s) important in the
etiology of this disease has begun.
Numerous hypotheses regarding the etiology have
been proposed; they include chronic regurgitation due to
hair balls;10,13 a natural progression of periodontal dis-
ease;4,13 associations with stomatitis potentially caused
by feline calici, feline leukemia, and feline immunodefi-
ciency virus infections;14 nutritional factors such as acidi-
fication of dry cat foods; 15 hypervitaminosis A from
consumption of raw liver;5,16 nutritional hyperparathy-
roidism;13 inadequate dietary calcium (e.g., homemade
diets); 15 other dietary factors (e.g., feeding certain brands
of nongrocery cat foods or certain table foods);17 and
more recently, low magnesium diets.8
This cross-sectional study was initiated to examine
risk factors associated with ORL.
Materials and Methods
All cats presented for dental prophylaxis to the dental
section of the Community Practice Service of the Com-
panion Animal Hospital at Cornell University between
February and July 1994 were identified. Cats with ORL
were eligible for inclusion in the case group, and those
without lesions were randomly sampled to serve as con-
trols. Cats were evaluated primarily by Dr. Saidla. Cats
were anesthetized for dental prophylaxis. The entire
mouth was thoroughly examined for dental pathology,
and the data was recorded in the medical record by an
experienced examiner. Each tooth was probed with a
color-coded probe, looking for subgingival lesions, gin-
gival recession, extrusion of teeth, very small enamel
lesions or pits, loss of tooth crowns with retained roots,
and hyperplasia of the gingiva into enamel defects. The
degrees of gingivitis and periodontal disease were also
noted. The following data was retrieved from each cat’s
medical record: the date of diagnosis of ORL; history of
dental disease; the number and location of lesions in the
mouth; prior extractions; types of other oral pathology;
prescriptions for antibiotics and other medications; and
disease diagnoses prior to the diagnosis of ORL.
Odontoclastic Resorptive Lesions 189
Owners of both case group cats and control cats were
asked to participate in a 15- to 20-minute telephone
interview during which they were asked to describe their
cat’s demographic characteristics (e.g., age, sex), diet
(e.g., type of food, frequency of feeding, table food),
number of cats owned, time spent outdoors, previous use
of medications, and medical history at and prior to the
date of diagnosis of ORL. All owners were told the study
was an assessment of factors affecting oral health. Inter-
viewers were blind to whether clients owned cats with
ORL or not, until late in the interview.
Associations between proposed risk factors and ORL
were evaluated using the chi-square test of independence
or Fisher’s exact test (where expected cell values were
less than 5) for categorical variables, and Student’s t-test
for continuous data with a Gaussian distribution.18 The
effects of variables significant at a p value of 0.20 or less
were evaluated further to control for potentially con-
founding variables using unconditional logistic regres-
sion where the dependent variable was the presence or
absence of ORL. 19 The model parameters were obtained
by maximum likelihood estimation using the computer
program EGRET. a The models were constructed using a
forward stepwise approach. The assessment of interac-
tions was attempted, but the small sample size resulted in
failure to converge in most models. The significance of
variables in the models was determined by evaluating
the likelihood ratio chi-square statistic in each step of the
fitting process. In light of the small numbers, variables
significant at a p value of 0.10 or less were retained in
the final model. The regression coefficients were
exponentiated to obtain the adjusted odds ratio (OR) for
each parameter.
Results
Thirty-five cats with ORL and 32 without were identi-
fied, of which 33 cases (94%) and 27 controls (84%)
were interviewed respectively. Reasons for nonparti-
cipation included being too busy, death of a cat, and
travel. There were more females in the case group
(51.5%) than in the control (40.7%) group, but the differ-
ence was not significant. All cats were neutered, and
most were of mixed breed (94%). Cases were, on aver-
age, a year older than controls, but both groups had been
owned an average of 7.5 years and had a mean weight
between 11 and 12 pounds [Table 1].
When individual variables were screened for their
association with ORL, seven were significant (p value of
0.20 or less). A higher proportion of cats with ORL
received two or more brands of canned foods, were fed
liver and other organ meats from the table, and had a
history of dental disease (including calculus, gingivitis,
and periodontal disease). Control cats spent more time
outdoors daily (particularly in the summer), lived in
rural areas more often, were fed commercial treats more
190 JOURNAL of the American Animal Hospital Association May/June 1999, Vol. 35

Gender, history of treatment with antibiotics, urinary

Table 1 acidifiers or other medications, source of water supply,
Characteristics of Cats With Feline number of other cats in the household, frequency of daily
Odontoclastic Resorptive Lesions feeding, prevalence of free-choice feeding, and the pro-
(Cases) and Controls portion of cats fed canned, dry, or semimoist pet foods
were not associated with ORL in this study. Also, con-
sumption of various table foods did not differ signifi-
Cases Controls
(n=33) (n=27) cantly between cats with ORL and without ORL.
Age at diagnosis (yrs) Mean 7.7 6.8 Discussion
Range 2–16 2–18
Builder11 first mentioned ORL in 1950, but speculated
Years owned Mean 7.5 7.6
Range 2–16 2–13 that the lesions were a form of caries. It wasn’t until
Weight (lbs) Mean 11.5 11.6 1976 that Schneck first distinguished ORL from dental
Range 7–19 9–16.5 caries and coined the term “neck lesions.”20 Since that
time, the importance of ORL as a leading cause of dental
No. % No. % disease in cats has been appreciated. Factors that may
predispose cats to the disease are beginning to emerge.
Gender Male 16 48.5 16 59.3
Female 17 51.5 11 40.7 If the apparent increase in ORL prevalence is not just
a result of cats living to older ages, then the change
Purebred Yes 2 6.1 1 3.7
No 31 93.9 26 96.3 suggests the introduction of some environmental factor(s)
or the removal of protective ones, or both. The primary
objective of this study was to compare exposures to
various environmental factors between cats with ORL
often, and received coat supplements. Control cats were and without ORL.
also more likely to have been treated recently for upper Previous dental disease (as recorded in the medical
respiratory infections (URI), episodes of diarrhea, and to records) was the strongest risk factor identified, with an
have received antibiotics. When the owners were asked odds ratio of between four and five (depending on the
about their respective cat’s medical history, however, there model), suggesting that cats with previous dental disease
were no differences in the frequency of reported lifetime had approximately four to five times higher risk for ORL
episodes of URI, diarrhea, or other medical conditions. than those without previous dental disease. The prior
The seven factors that were significantly associated dental disease was predominantly calculus accumula-
with ORL using a liberal p value of 0.20 or less, were tion, gingivitis, and periodontal disease. One theory of
examined simultaneously in the multivariate models. etiopathogenesis suggests that plaque bacteria induce an
Only the four factors that remained significant (p value inflammatory response in the gingival tissue which leads
of 0.10 or less) are discussed. In the multivariate models, to attraction of circulating stem cells that become
four factors (i.e., time spent outdoors, a history of dental odontoclastic cells. These cells then initiate damage and
disease, urban-rural residence, and consumption of com- lead to granulation tissue with an enhanced blood supply
mercial cat treats) remained associated with ORL. Since which facilitates remodeling of bone-cementum tissue,
time spent outdoors was highly correlated with area of all of which contribute to ORL.21 The inflammation,
residence (i.e., city, village, or rural), both residence and however, may be secondary to the resorptive lesion.10,20
time spent outdoors were not examined in the same At least two previous studies have found an association
multivariate model. Therefore, two models are presented between the occurrence of ORL and periodontitis.7,22
[Table 3]. Living in an urban area (OR, 4.4) and a history Cats with ORL spent more time indoors, and being
of dental disease (OR, 4.5) increased risk more than indoors most of the time (18 hours or more) increased
fourfold compared to cats living on a farm and with no risk more than fourfold. The converse of this, time out-
previous dental disease, respectively. Receiving com- doors, particularly in the summer, appeared to lower
mercial treats reduced risk approximately threefold. risk. Cats with ORL also spent less time outside during
If time spent outdoors was evaluated in lieu of resi- the winter, but the results were not statistically signifi-
dence, strictly indoor cats had a 4.5 times higher risk of cant, probably reflecting the small number of cats spend-
ORL, and cats spending some time (but no more than six ing prolonged periods outdoors in central New York.
hours) outside daily had a 4.3 times higher risk of ORL The authors speculate that the significance of time spent
compared to cats spending seven or more hours outdoors outdoors may be related to hunting activities and that
daily in the summer. The risks associated with a history consuming prey may help to clean teeth and prevent oral
of dental disease and commercial treats were essentially disease leading to ORL. Many owners were unsure
unchanged in this model. whether their cat “hunted,” and, therefore, the quality of
May/June 1999, Vol. 35 Odontoclastic Resorptive Lesions 191

Table 2
Potential Risk Factors Associated (p≤0.20) With Feline Odontoclastic Resorptive Lesions

Cases Controls P Value

Variable Description No. % No. %
Location Rural/farm 10 30.3 12 44.4 0.02*
City of <2,500 people 4 12.1 9 33.3
City of >2,500 people 19 57.6 6 22.2
Hours outdoors daily Not out 16 48.5 10 37.0 0.15
During summer 1 to 6 10 30.3 5 18.5
7 or more 7 21.2 12 44.4
Previous dental disease Yes 12 36.4 5 18.5 0.13
No 21 63.6 22 81.5
Fed organ meats from table Yes 5 15.2 0 0 0.06
No 28 84.8 27 100
Fed commercial cat treats Yes 7 21.2 11 40.7 0.10
No 26 78.8 16 59.3
Fed coat supplement Yes 1 3 4 14.8 0.16
No 32 97 23 85.2
Number of canned brands fed None 11 33.3 7 25.9 0.08
1 7 21.2 13 48.2
>1 15 45.5 7 25.9

* P value comparing variable distribution between cases and controls

data solicited with questions regarding hunting is sus-

Table 3 pect. Lund, et al., 8 found no association with access to
the outdoors, but no inquiries were made about time
Risk Factors Significantly Associated With spent outdoors. Access or nonaccess to the outdoors was
Risk of Feline Odontoclastic Resorptive also not significant in this or a previous study.17 It was
Lesions in Multivariate Models only after inquiries were made about time spent outdoors
that differences emerged. There are reports of stray and
90% feral cat populations having a lower prevalence of le-
Odds Confidence
sions, 10 but more studies are needed. The demonstration
Risk Factors Ratio Interval
of ORL in wild felidae 23 does not necessarily contradict
Model I this hypothesis, as the actual prevalence of wild animals
Location with lesions has not been reported. The finding that some
Rural 1.0 —
breeds may have a higher prevalence of ORL10 may be a
Village (<2,500 people) 0.5 0.1–1.8
City (≥2,500 people) 4.4 1.4–13.6 reflection, at least partially, of their lack of access to the
History of dental disease outdoors because owners do not want to risk losing
No 1.0 — valuable animals. There were insufficient purebred cats
Yes 4.5 1.4–15.2 in this study to evaluate breed as a risk factor.
Commercial treat The increased risk associated with urban residence or,
No 1.0 — conversely, the protective effect of rural residence (which
Yes 0.3 0.1–0.9 appeared to decrease risk more than four times compared
Model II to city residence) may also be associated with hunting
Time outdoors (summer) activities. Time spent outdoors and rural residence were
≥7 hours 1.0 — significantly associated. It is also possible that some
1 to 6 hours 4.3 1.1–15.9
other factor(s) associated with rural residence (such as
Not out 4.5 1.3–15.2
water source) may be of etiological significance. Lend-
History of dental disease
No 1.0 — ing support to this hypothesis is Lund, et al.’s, observa-
Yes 5.3 1.5–18.0 tion that a higher proportion of cats with ORL drank well
Commercial treat water (according to their univariate analysis).8 Similarly,
No 1.0 — strictly indoor, city-dwelling cats may be exposed to
Yes 0.3 0.1–0.9 unknown factor(s) that increase their risk.
192 JOURNAL of the American Animal Hospital Association
The significance of the apparent protective effect of
commercial treats is unknown. Approximately 41% of
cats with no lesions and 21% of those with lesions re-
ceived treats at least once weekly. Closer examination of
this data revealed that most of the cats with and without
ORL which were fed treats received one brand, but the
significance of this observation is unclear. It may be that
some other aspect of the care or management of cats fed
treats is related to risk for ORL, or the finding occurred
by chance.
The observation that cats without ORL were more
likely to have been treated recently for URI and diarrhea
most likely reflects a bias in control selection; that is,
when these cats were presented for treatment of URI or
another presenting health concern, dental cleanings were
recommended. When owners were asked about their re-
spective cat’s history of respiratory, intestinal, and other
illnesses, there were no differences between cases and
controls. No associations were found with gender, num-
ber of cats owned, a history of hair ball regurgitation,
prior URI or other diseases, types of diet (e.g., dry or
canned), or consumption of various table foods.
While the authors attempted to collect information
about potential causes preceding the diagnosis of ORL,
the critical period during which the lesions may be in-
duced is unknown. Therefore, data regarding diet, resi-
dence, time spent outdoors, etc., was collected primarily
from the time of diagnosis of ORL or the time of recom-
mended dental procedures in controls. Since much of the
exposure data was cross-sectional in nature (i.e., col-
lected at the time of diagnosis), the design is considered
cross-sectional despite sampling cases and controls.
This design enabled the study of multiple possible
risk factors but was limited by its cross-sectional nature.
The study was also limited by incomplete records, where
specific information was either incomplete or entirely
missing. Identification of the nature and extent of previ-
ous dental disease, for example, was not always com-
plete. Similarly, although every effort was made to
remove cats with previous ORL from the control series,
a few may have had lesions themselves as not all cats were
radiographed. The relatively small sample size also ham-
pered efforts to examine multiple variables simultaneously.
Conclusion
Data from previous studies suggests that the prevalence
of feline ORL has increased, possibly due to a change(s)
in the environment of cats. Numerous hypotheses have
been advanced, but the etiology of these lesions remains
elusive. Results from this study suggest that cats spend-
ing more time outdoors or living in a rural residence, or
both, may have reduced risk, perhaps because of access
to supplementation of their diet with natural prey. A
decreasing number of cats with outdoor access may
explain the increase in frequency of this disease. The
etiology is undoubtedly complex, resulting from the in-
May/June 1999, Vol. 35
teraction of multiple factors. In light of its high fre-
quency and serious consequences, further research is
warranted.
a
Egret Statistical Package User’s Manual, 1987; Statistics and Epidemiology
Research Corporation (SERC) Software Division, Seattle, WA
References
1. Ohba S, Kiba H, Kuwabara M, et al. A histopathological study of neck
lesions in feline teeth. J Am Anim Hosp Assoc 1993;29:216–20.
2. Harvey CD. Feline resorptive lesions. Sem Vet Med Surg 1993;8:187–96.
3. Frost P, Williams CA. Feline dental disease. Vet Clin N Am Sm Anim Prac
1986;16:851–73.
4. DeBowes LJ. Odontoclastic resorptive lesions in cats. Waltham Focus
1994;4:2–8.
5. Eisner ER. Chronic subgingival tooth erosion in cats. Vet Med 1989;84:
378–87.
6. Lyon KF. Subgingival odontoclastic resorptive lesions. Classification,
treatment, and results in 58 cats. Vet Clin N Am Sm Anim Prac
1992;22:1417–32.
7. Van Wessum R, Harvey CE, Hennet P. Feline dental resorptive lesions.
Vet Clin N Am Sm Anim Prac 1992;22:1405–15.
8. Lund EM, Bohacek LK, Dahlke JL, et al. Prevalence and risk factors for
odontoclastic resorptive lesions in cats. J Am Vet Med Assoc 1998;212:
392–5.
9. Coles S. The prevalence of buccal cervical root resorptions in Australian
cats. J Vet Dent 1990;7:14–6.
10. Wiggs RB, Lobprise HB. Veterinary dentistry: principles and practice.
Philadelphia: Lippincott-Raven, 1997:487–90.
11. Builder PL. Opening paper. Vet Rec 1955;67:386–90.
12. Okuda A, Harvey CE. Etiopathogenesis of feline dental resorptive lesions.
Vet Clin N Am Sm Anim Prac 1992;22:1385–403.
13. Mulligan TW. Feline cervical line lesions. Vet Med Report 1990;2:343–9.
14. Thompson RR, Wilcox GE, Clark WT, et al. Association of calicivirus
infection with chronic gingivitis and pharyngitis in cats. J Sm Anim Pract
1984;25:207–10.
15. Zetner K, Steurer I. The influence of dry food on the development of feline
neck lesions. J Vet Dent 1992;9:4–6.
16. Seawright AA, English PB, Gartner RJW. Hypervitaminosis A of the cat.
Adv in Vet Sci and Comp Med, 1970;14:1–27.
17. Donoghue S, Scarlett JM, Williams CA, et al. Diet as a risk factor for feline
external odontoclastic resorption (abstract). J Nutr 1994;124:2693S–4S.
18. Dawson-Saunders B, Trapp RG. Basic and clinical biostatistics. Norwalk,
Connecticut: Appleton and Lange, 1994:114–6, 149–55.
19. Kleinbaum DG, Kupper LL, Morgenstern H. Epidemiologic research.
London: Lifetime Learning Pub, 1982:461–82.
20. Schneck GW, Osborn JW. Neck lesions in the teeth of cats. Vet Rec
1976;99:100.
21. Harvey CE, Emily PP. Small animal dentistry. St. Louis: Mosby, 1993:
221–5.
22. Harvey CE, Shafer F. Epidemiology of canine and feline oral diseases.
Proc Vet Dent Forum 1992;6:45–6.
23. Berger M, Schawalder P, Stich H, et al. Feline dental resorptive lesions in
captive and wild leopards and lions. J Vet Dent 1996;13:13–21.
 Ultrasonographic Characteristics of Both
Adrenal Glands in 15 Dogs With
Functional Adrenocortical Tumors
Ultrasonographic examination of both adrenal glands was performed in 15 dogs with functional
adrenocortical tumors (FAT). Bilateral adrenal tumors were diagnosed in three of 15 dogs, and
unilateral tumors were diagnosed in 12 of 15 dogs. Adrenal tumors were characterized by
adrenal gland enlargement with loss of the normal shape and parenchymal structure. The
contralateral adrenal gland could be imaged in all dogs with unilateral tumors. Based on size,
shape, and parenchymal structure, the contralateral adrenal gland was similar to adrenal
glands of normal dogs. The results of this study show that: 1) both adrenal glands should be
imaged routinely in dogs with hyperadrenocorticism; 2) bilateral adrenocortical tumors seem to
be more frequent than previously assumed; 3) one normal adrenal gland does not exclude the
existence of a contralateral FAT; and 4) the functional atrophy of the contralateral adrenal gland
in dogs with FAT may not be apparent ultrasonographically. J Am Anim Hosp Assoc 1999;35:193–9.

Angelika Hoerauf, DVM Introduction

Claudia Reusch, DVM, PhD
O tumor (FAT; adenoma or carcinoma).1 The differentiation between these
From the Department of
Veterinary Internal Medicine,
University of Zurich,
Winterthurerstr. 260, CH–8057
Zurich, Switzerland.
Hyperadrenocorticism (i.e., Cushing’s syndrome) is a commonly diag-
nosed endocrine disorder in the dog. Most frequently it is associated with
pituitary-dependent bilateral adrenocortical hyperplasia (PDH). In 10%
to 20% of cases, the syndrome is triggered by a functional adrenocortical
two forms is important from a therapeutic as well as a prognostic point of
view. Dogs with FAT are much less responsive to o,- p´-DDDa therapy
than are dogs with PDH.2 On the other hand, unilateral adrenalectomy in
the case of a confirmed FAT can lead to full recovery.
During the last years, a number of hormonally based tests as well as
multiple imaging modalities (e.g., radiography, computed tomography,
ultrasonography) have been used for the differential diagnosis of
hyperadrenocorticism. With abdominal radiography, adrenocortical tu-
mors were identified in only 50% of the cases,3,4 while computed tomog-
raphy enabled visualization of both adrenal glands in all dogs.5–7 However,
this technique is time-consuming and not readily available.
The ultrasonographic detection of an adrenocortical tumor was first
described by Kantrowitz, et al. 8 and Poffenberger, et al. 9 At that time,
visualization of nonenlarged adrenal glands by means of ultrasonography
seemed impossible due to their anatomical position and the small differ-
ence in echogenicity compared to the surrounding tissue. However, with
the currently available higher-quality ultrasound equipment and increas-
ing ultrasonographer experience, it has now become possible to detect
and characterize the adrenal glands of healthy dogs.10–14 Information on
the size of the adrenal glands in healthy dogs was provided for the first
time by Barthez, et al.15 Further studies could show that ultrasonography
is a helpful procedure in the evaluation of adrenal glands in dogs sus-
pected of having hyperadrenocorticism. In dogs with PDH, ultrasono-
graphic characteristics of the adrenal glands include bilaterally
symmetrical adrenomegaly, increased adrenal thickness, maintenance of

JOURNAL of the American Animal Hospital Association 193

194 JOURNAL of the American Animal Hospital Association May/June 1999, Vol. 35

Table 1
Results of Adrenocorticotropin Hormone (ACTH) Stimulation and Low-Dose
Dexamethasone Suppression (LDDS) Tests in 15 Dogs With Hyperadrenocorticism
Due to a Functional Adrenocortical Tumor

ACTH Stimulation LDDS Test

(cortisol µg/dl) (cortisol µg/dl)
Case No. Pre 1 Hr Post Pre 4 Hrs Post 8 Hrs Post
1* 5.6 26.4 6.4 6.8 6.3
2* 4.2 18.3 3.5 4.1 3.6
3* 5.2 25.7 5.3 4.4 4.5
4 2.7 9.6 4.3 3.3 3.5
5 2.6 41.9 4.8 5.3 4.0
6 9.1 34.9 9.3 9.3 10.6
7 0.2 19.4 4.3 3.1 3.1
8 4.3 43.6 7.0 4.1 4.0
9 9.2 17.8 8.6 12.6 4.8
10 4.2 12.2 5.3 6.1 4.8
11 ND† ND 5.4 3.4 3.2
12 4.7 26.0 5.3 4.8 4.8
13 5.1 21.7 4.6 3.7 4.4
14 ND ND 4.1 3.2 2.8
15 7.9 31.2 4.5 7.0 4.4

* Dogs with bilateral adrenal gland tumors

†
ND=not done

the normal adrenal shape and contour, and homogeneous
echogenicity in the majority of cases.16–18 In contrast,
FATs are mostly unilateral, with an irregular, rounded
shape and mixed echogenicity.19–21 Ultrasonographic ap-
pearance and size of the contralateral adrenal glands in
dogs with unilateral adrenal tumors have been reported
in only three cases.22,23 The goal of this study was to
characterize both adrenal glands in dogs with FAT.
Materials and Methods
Cases
This prospective study was completed with dogs evalu-
ated at the Department of Veterinary Internal Medicine,
University of Munich, Germany between January 1, 1995
and April 1, 1996. To be included, each dog must have
been suspected of having hyperadrenocorticism on the
basis of history and the results of physical examination,
complete blood count, serum biochemical analysis, and
urinalysis. The diagnosis of hyperadrenocorticism must
have been confirmed with abnormal results on at least
one of the following two screening tests: an adrenocorti-
cotropin hormone (ACTH) stimulation test (plasma cor-
tisol concentration greater than 20 µg/dl one hour after
intramuscular [IM] administration of one vial of syn-
thetic ACTH [0.25 mg Tetracosactid-hexaacetateb]) and
a low-dose dexamethasone suppression test (LDDS)
(plasma cortisol concentration greater than 1.4 µg/dl
eight hours after intravenous [IV] administration of 0.01
mg/kg dexamethasonec) [Table 1].
The LDDS test was used to both confirm and classify
the diagnosis of hyperadrenocorticism as either FAT or
PDH. Each of the following two criteria was considered
proof of adrenal suppression in a LDDS test (indica-
tive of PDH): a four-hour postdexamethasone plasma
cortisol concentration less than 50% of the basal
plasma concentration or a four-hour postdexameth-
asone plasma concentration less than 1.4 µg/dl. 24 To
be included in this study (i.e., suspect FAT), each dog
must have had test results indicative of no suppres-
sion [Table 1].
The group was made up of four male and 11 female
dogs (five mixed-breeds, three dachshunds, three schnau-
zers, one shih tzu, one fox terrier, one Rhodesian
ridgeback, and one poodle) aged five to 16 years (me-
dian, 12 yrs) and weighing 3 to 30 kg (median, 14 kg)
[Table 2].
May/June 1999, Vol. 35 Adrenocortical Tumor Ultrasonography 195

Table 2
Breed, Age, Sex, and Body Weight of 15
Dogs With Hyperadrenocorticism Due to a
Functional Adrenocortical Tumor

Case Age Weight

No. Breed (yrs) Sex† (kg)
1* Standard schnauzer 14 F 20
2* Shih tzu 13 F 3
3* Standard schnauzer 15 F 21
4 Dachshund 11 M 11
5 Dachshund 15 M 12
6 Fox terrier 9 F 14
Figure 1—Longitudinal sonogram of the left adrenal gland of a
7 Mixed-breed dog 12 F 12 healthy control dog imaged in a sagittal plane. The head is to the
8 Rhodesian ridgeback 11 M 30 left. The adrenal gland is peanut shaped and hypoechoic to the
surrounding tissue. Adrenal gland thickness is represented by “x”
9 Miniature schnauzer 8 M 12 calipers; the long axis of the adrenal gland is represented by “+”
10 Mixed-breed dog 5 F 15 calipers; phrenicoabdominalis vein is represented by “➔” calipers.
11 Mixed-breed dog 16 F 23
12 Mixed-breed dog 12 F 25 median values of adrenal gland measurements are given.
13 Dachshund 13 F 8 The Mann-Whitney U-test was used for determining the
14 Poodle 7 F 10 differences between the two groups. Differences were
15 Mixed-breed dog 15 F 21 considered significant for a p value of 0.05 or less.

* Dogs with bilateral adrenal gland tumors Results

†
F=female; M=male
Ultrasonographic Technique
The dogs were placed in dorsal recumbency. None of the
dogs were sedated or anesthetized. A real-time mechani-
cal sector scannerd with a 7.5-megahertz (MHz) trans-
ducer was used. Ultrasonographic examination of the
adrenal glands was performed by one person (Hoerauf)
as described by Saunders.20
An adrenal gland tumor was defined as adrenomegaly
with loss of the normal shape and structure. Tumor size
was determined by measuring the greatest craniocaudal
and dorsoventral dimension in a longitudinal plane.
Contralateral adrenal glands were also measured in
the longitudinal plane. Length was defined as the great-
est craniocaudal dimension, and thickness was defined
as the greatest dorsoventral dimension perpendicular to
the long axis [Figure 1].
Statistical Analysis
The ultrasonographically-determined adrenal gland sizes
in dogs with FATs were compared with the adrenal
gland sizes determined in 20 healthy dogs (body weights,
2.5 to 32 kg; median, 14 kg) in an earlier study. 13 The
statistical evaluation was carried out with the help of the
statistics program, Statistical Package for the Social
Sciences,e using nonparametric procedures. Ranges and
Both adrenal glands could be visualized in 15 (100%) of
15 dogs examined. In three (case nos. 1, 2, 3) of 15 dogs,
bilateral adrenal gland tumors were diagnosed ultrasono-
graphically. The craniocaudal dimensions of the six bi-
lateral adrenal gland tumors in these three cases were
between 18.7 and 44.5 mm (median, 30.2 mm), and the
dorsoventral dimensions were between 10.6 and 24.8
mm (median, 18.7 mm) [Table 3]. In case nos. 1 and 2,
the left and right adrenal glands had a rounded shape and
were similar in size and echogenicity. The parenchyma
was of mixed echogenicity and was slightly hyperechoic
compared to the parenchyma of adrenal glands in normal
dogs. In both cases (case no. 1, right adrenal gland; case
no. 2, left adrenal gland), one of the adrenal glands had
mineralized foci. In case no. 3, the left adrenal gland had
a similar appearance without the mineralization. The
right adrenal gland of this dog was smaller than the left
one; the cranial pole was mineralized; and the caudal
pole had a mixed echogenicity with multiple hypoechoic
foci.
Unilateral adrenal gland tumors were seen in 12 of 15
dogs. The tumor was located on the left side in three
(case nos. 4, 5, 6) dogs and on the right side in nine (case
nos. 7–15) dogs. The craniocaudal dimensions of the
tumors were between 17.4 and 52 mm (median, 29.9
mm), and the dorsoventral dimensions were between 9.2
and 45 mm (median, 18.3 mm) [Table 3]. Ten of 12
neoplastic adrenal glands were rounded; in two cases,
only the cranial pole was thickened in a circular fashion.
196 JOURNAL of the American Animal Hospital Association May/June 1999, Vol. 35

Table 3
Ultrasonographic Measurements of Adrenal
Gland Length and Thickness in 15 Dogs
With Hyperadrenocorticism Due to
Functional Adrenocortical Tumor

Left Adrenal Gland Right Adrenal Gland

Case Length Thickness Length Thickness

No. (mm) (mm) (mm) (mm)
1* 37.3 24.8 33.0 22.0
Figure 2A
2* 19.7 16.8 18.7 14.3
3* 44.5 20.5 27.3 10.6 Figures 2A, 2B—(A) Longitudinal sonogram of the right-sided
adrenocortical tumor of case no. 9. The head is to the left. The
4† 20.8 9.2 9.6 4.2 adrenal gland appears as a rounded mass of mixed echogenicity
5† 32.0 28.0 15.7 4.2 (compare to Figure 1). (B) Longitudinal sonogram of the right-
6† sided adrenocortical tumor of case no. 9 invading the vena cava
33.1 18.7 20.0 3.2
caudalis (white arrows).
7‡ 19.6 6.1 17.4 17.8
8‡ 25.6 7.7 52.0 34.0
9‡ 17.5 4.2 24.6 12.0
10‡ 15.4 3.0 20.7 11.7
11‡ 22.9 3.5 27.7 15.5
12‡ 17.9 4.6 45.0 45.0
13‡ 21.3 7.5 20.3 14.0
14‡ 17.8 3.6 41.8 35.2
15‡ 25.1 5.8 43.0 30.2
Normal dogs13
Range 13.2–26.3 3.0–5.2 12.4–22.6 3.1–6.0
(n=20)
Median 17.4 4.1 16.7 4.3

* Dogs with bilateral adrenal gland tumors Figure 2B

†
Dogs with left-sided adrenal gland tumors
‡
Dogs with right-sided adrenal gland tumors
Note: Values for the contralateral adrenal glands are
given in bold numbers
Echogenicity of adrenal gland tumors was homogeneous
in two cases (similar to the renal cortex in appearance).
Ten adrenal gland tumors were of mixed echogenicity
with hypo- and hyperechoic foci. Mineralization with
shadow formation was demonstrated in four tumors. In-
vasion of the caudal vena cava was evident in two unilat-
eral right-sided tumors [Figure 2].
The contralateral adrenal glands in the 12 dogs with
unilateral adrenal gland tumors were 9.6 to 25.6 mm
long (median, 18.8 mm) and 3.0 to 7.7 mm thick (me-
dian, 4.2 mm) [Table 3]. They had a shape similar to
adrenal glands in normal dogs. In dogs with right-sided
tumors, the left adrenal gland was peanut shaped; whereas
in cases with left-sided tumors, the right adrenal gland
was comma shaped. All contralateral adrenal glands were
hypoechoic compared to surrounding tissues and had a
double-layered structure like adrenal glands in normal
dogs [Figure 3].
A statistical comparison between normal control dogs
and dogs with unilateral adrenal gland tumors was made
with respect to the length and thickness of the contralat-
eral adrenal gland. Measurements made by Hoerauf and
Reusch13 were utilized as normal values. No significant
difference was seen in length or thickness of the con-
tralateral adrenal gland compared to adrenal glands in
normal dogs [Figure 4].
Discussion
In the past few years, ultrasonography of adrenal glands
has gained importance in the differential diagnosis of
canine hyperadrenocorticism. In dogs with PDH, bilat-
eral symmetrical adrenal gland enlargement with normal
shape and echogenicity is seen ultrasonographically. In
comparison, a FAT most frequently is characterized by
an irregular, rounded shape with mixed echogenicity.
Previous studies regarding size and appearance of the
May/June 1999, Vol. 35
Figure 3A
Figures 3A, 3B—(A) Longitudinal sonogram of the right-sided
adrenocortical tumor of case no. 12. The head is to the left. The
adrenal gland appeared as a large, rounded mass (white arrows)
of mixed echogenicity (compare to Figure 1). Note the foci of
mineralization associated with shadowing (black arrows). (B)
Longitudinal sonogram of the contralateral (left) adrenal gland of
case no. 12. The head is to the left, white arrows on cranial and
caudal margin. The adrenal gland is peanut shaped, hypoechoic
to the surrounding tissue, and has double-layered structure. The
adrenal gland did not change in size, shape, and echogenicity
compared to the left adrenal gland of normal dogs (compare to
Figure 1).
Figure 3B
contralateral adrenal gland in dogs with a unilateral FAT
have not been published.
Cortisol production by the adrenal gland tumor low-
ers ACTH secretion via a negative feedback mechanism,
resulting in atrophy of the zona reticularis and fasciculata
of the unaffected adrenal gland. Histologically, the con-
tralateral adrenal gland is characterized by a thinned
cortex, consisting primarily of zona glomerulosa, and a
thickened adrenal capsule.25 In view of these patho-
physiological findings, it was presumed that the con-
tralateral adrenal gland in the presence of an adrenal
gland tumor would be considerably smaller than a nor-
mal adrenal gland and, therefore, difficult to detect ultra-
sonographically.11 Recently, Liste, et al.22 reported a dog
with a unilateral adrenal gland tumor and an atrophied
contralateral adrenal gland, whereas Besso, et al.23 re-
ported normal-sized contralateral adrenal glands in two
dogs with FAT.
Adrenocortical Tumor Ultrasonography 197
Figure 4—Comparison of the ultrasonographically determined
length and thickness of the left adrenal gland in 20 healthy dogs
and the contralateral adrenal gland in 12 dogs with hyperadreno-
corticism due to a functional unilateral adrenocortical tumor
(n.s.=not significant).
Therefore, the main focus of the authors’ study was to
determine if it is consistently possible to detect the con-
tralateral adrenal gland in dogs with a unilateral adrenal
tumor ultrasonographically, and to evaluate its size,
shape, and echogenicity, as well as to describe the ultra-
sonographic characteristics of adrenal gland tumors in a
relatively large number of dogs.
In this study, the adrenal glands were examined
ultrasonographically in 15 dogs in which FAT was sus-
pected based on the results of history and laboratory
findings. Four male and 11 female middle-aged dogs
were included in the study. In all cases, both adrenal
glands could be visualized.
Bilateral neoplastic changes were diagnosed in three
(20%) of 15 dogs. The frequency of bilateral adrenal
gland tumors in this study is considerably higher than
previously reported.4,26 However, it should be noted that
the finding of bilateral tumors does not necessarily imply
the presence of bilateral cortisol-producing adrenal gland
tumors. It is important to realize that the ultrasonographic
image does not give any information on the type of
tumor, nor its function. In the literature, six cases of
concurrent hypercortisolism (four cases with PDH and
two cases with FAT) and pheochromocytoma have been
described.27 Unfortunately, there is no information about
the ultrasonographic appearance of the adrenal glands in
these dogs. In two dogs (case nos. 1, 2) of the authors’
study, it is likely that both tumors were of adrenocortical
origin because of their similar appearance. In case no. 3,
the significance of the difference in size, shape, and
echogenicity between the left and right adrenal glands
remains unclear. Furthermore, the possibility exists that
in cases of a unilateral adrenal gland tumor, enlargement
of the contralateral adrenal gland may occur due to a
pituitary adenoma. 1 In these cases, ultrasonography gives
conflicting results. A definitive diagnosis can be made
only through further diagnostic tests. In the authors’
study, this situation is very unlikely because the ultra-
sonographic feature of an adrenal gland tumor was de-
198 JOURNAL of the American Animal Hospital Association
fined as an overall enlargement with loss of the normal
shape and structure. These characteristics are in contrast
to adrenal gland enlargement due to pituitary adenoma.
Unilateral adrenal tumors were diagnosed ultrasono-
graphically in 12 dogs; the tumors occurred on the right
side in nine dogs and on the left side in only three dogs.
This preference for the right side is in contrast to previ-
ous findings where an equal distribution was observed.4
Ultrasonographic examination of neoplastic adrenal
glands resulted in definite variation from normal with
respect to size, shape, and echogenicity. Adrenal tumors
were characterized by a considerable enlargement (me-
dian, craniocaudal dimension of 29.9 mm; median, dor-
soventral dimension of 18.3 mm), which resulted in a
significant change in the overall shape of the entire or-
gan, or one pole. The two-layered appearance observed
in the adrenal glands of normal dogs was absent; in most
cases a mixed echogenicity was seen, which was attrib-
uted to necrotic foci in the parenchyma of the tumor.9
Differentiation between adrenal adenoma and adenocar-
cinoma based on size, parenchymal structure, and miner-
alization observed in four unilateral tumors was not
possible, since 50% of adenomas and carcinomas can
show calcium deposition in the parenchyma.4 However,
the presence of an adenocarcinoma was presumed based
on invasion of the caudal vena cava in two cases.
The contralateral adrenal gland could be imaged in all
dogs with unilateral FAT. These findings contradict the
postulate made by Saunders, et al.11 who presumed that
the nonneoplastic adrenal gland would atrophy and,
therefore, felt that ultrasonographic characterization was
impossible. Based on size, shape, and parenchymal struc-
ture, the contralateral adrenal gland was similar to adre-
nal glands of normal dogs. Similarly, the length and
thickness of the contralateral adrenal gland (median
length, 18.8 mm; median thickness, 4.2 mm) were not
significantly smaller than those of normal dogs (median
length of left adrenal gland, 17.4 mm; median thickness
of left adrenal gland, 4.1 mm). It must be emphasized
that the contralateral adrenal gland thickness in all dogs
with unilateral FAT was not less than in any of the
normal dogs. The atrophy of the adrenal cortex does not
appear to result in a shrinkage of the adrenal gland that
can be demonstrated by current ultrasonographic means.
These results parallel the findings of a previous study.23
Besso, et al.23 reported on ultrasonographically normal-
sized contralateral adrenal glands in two dogs with FAT
that were proven to have an atrophied cortex at necropsy.
On the other hand, Liste, et al.22 reported on a smaller
contralateral adrenal gland in one dog with FAT mea-
sured by ultrasonography, and Emms, et al.6 reported on
smaller contralateral adrenal glands in five dogs with
FAT detected by computed tomography. Because of these
conflicting results, the authors theorize that the size of
the contralateral adrenal gland in dogs with FAT de-
pends on the duration of the disease.
May/June 1999, Vol. 35
Conclusion
The results of this study show that in cases of hyper-
adrenocorticism, both adrenal glands should be imaged
routinely. Bilateral adrenocortical tumors seem to be
more frequent than previously assumed. The diagnosis
of bilateral FAT is important with respect to treatment
and prognosis. The finding of one normal adrenal gland
does not exclude the existence of a FAT on the other
side. The authors were able to demonstrate that in dogs
with FAT, the size, shape, and parenchymal structure of
the contralateral adrenal gland was similar to those of
normal dogs. Therefore, the functional atrophy of this
gland may not be apparent ultrasonographically.
a
Lysodren; Bristol-Myers Squibb, Princeton, NJ
b
Synacthen; CIBA Pharma, Wehr, Germany
c
Hexadreson; Vemie, Kempen, Germany
d
Sim 7000 CFM Challenge; Esaote Biomedica, Italy
e
SPSS for Windows, Version 6.0; SPSS Inc., Chicago, IL
References
1. Feldman EC, Nelson RW. Hyperadrenocorticism. In: Feldman EC, Nelson
RW, eds. Canine and feline endocrinology and reproduction. Philadelphia:
WB Saunders, 1996:187–265.
2. Kintzer PP, Peterson ME. Challenges in the management of canine
hyperadrenocorticism (Cushing’s disease). Proceed, Am Coll Vet Int Med
1994:165–9.
3. Penninck DG, Feldman EC, Nyland TG. Radiographic features of canine
hyperadrenocorticism caused by autonomously functioning adrenocortical
tumors: 23 cases (1978–1986). J Am Vet Med Assoc 1988;192:1604–8.
4. Reusch C, Feldman EC. Canine hyperadrenocorticism due to adrenocortical
neoplasia: pretreatment evaluation of 41 dogs. J Vet Int Med 1991;5:3–10.
5. Bailey MQ. Use of x-ray-computed tomography as an aid in localization of
adrenal masses in the dog. J Am Vet Med Assoc 1986;188:1046–9.
6. Emms SG, Wortman JA, Johnston DE, Goldschmidt MH. Evaluation of
canine hyperadrenocorticism, using computed tomography. J Am Vet Med
Assoc 1986;189:432–9.
7. Voorhout G, Stolp R, Lubberink ME, Van Waes PF. Computed
tomography in the diagnosis of canine hyperadrenocorticism not
suppressible by dexamethasone. J Am Vet Med Assoc 1988;192:641–6.
8. Kantrowitz BM, Nyland TG, Feldman EC. Adrenal ultrasonography in the
dog. Detection of tumors and hyperplasia in hyperadrenocorticism. Vet
Radiol 1986;27:91–5.
9. Poffenbarger EM, Feeney DA, Hayden DW. Gray-scale ultrasonography in
the diagnosis of adrenal neoplasia in dogs: six cases (1981–1986). J Am
Vet Med Assoc 1988;192:228–32.
10. Voorhout G. X-ray-computed tomography, nephrotomography and
ultrasonography of the adrenal glands of healthy dogs. Am J Vet Res
1990;51:625–31.
11. Saunders HM, Pugh CR, Rhodes WH. Expanding applications of
abdominal ultrasonography. J Am Anim Hosp Assoc 1992;28:369–74.
12. Grooters AM, Biller DS, Merryman J. Ultrasonographic parameters of
normal canine adrenal glands: comparison to necropsy findings. Vet Radiol
1995;36:126–30.
13. Hoerauf A, Reusch C. Darstellung der Nebennieren mittels Ultraschall:
Untersuchungen bei gesunden Hunden, Hunden mit nicht-endokrinen
Erkrankungen sowie mit Cushing-Syndrom. Kleintierpraxis 1995;40:
351–60.
14. Kaser-Hotz B, Saunders HM. Sonographie der Nebennieren beim Hund.
Schweiz Arch Tierheilk 1995;137:258–64.
15. Barthez PY, Nyland TG, Feldman EC. Ultrasonographic evaluation of the
adrenal gland in normal dogs and dogs with pituitary-dependent
hyperadrenocorticism. Proceedings, Am Coll Vet Int Med, 1994:160.
16. Barthez PY, Nyland TG, Feldman EC. Ultrasonographic evaluation of the
adrenal glands in dogs. J Am Vet Med Assoc 1995;207:1180–3.
May/June 1999, Vol. 35
17. Hoerauf A, Reusch C. Ultrasonographic evaluation of adrenal gland in
healthy dogs, dogs with no evidence of endocrine disease and dogs with
Cushing’s disease. Vet Radiol 1995;36:434.
18. Grooters AM, Biller DS, Theisen SK, Miyabayashi T. Ultrasonographic
characteristics of the adrenal glands in dogs with pituitary-dependent
hyperadrenocorticism: comparison with normal dogs. J Vet Int Med
1996;10:110–5.
19. Voorhout G, Stolp R, Rijnberk A, Van Waes PF. Assessment of survey
radiography and comparison with x-ray computed tomography for
detection of hyperfunctioning adrenocortical tumors in dogs. J Am Vet
Med Assoc 1990;196:1799–803.
20. Saunders HM. Adrenal disease. Proceedings, 2nd International Symposium
Vet Echography, 1993:31–4.
21. Hoerauf A, Reusch C. Ultrasonographic evaluation of adrenal glands in
dogs with Cushing’s disease due to functional adrenal tumors and dogs
with Addison’s disease. Vet Radiol 1996;37:448.
22. Liste F, Cuevas M, Gascon M, Garcia de Jalon J, Cuevas J. Ultrasono-
graphic diagnosis of an adrenocortical carcinoma in a dog. Vet Rec
1997;140:339–41.
Adrenocortical Tumor Ultrasonography 199
23. Besso JB, Penninck DG, Gliatto JM. Retrospective ultrasonographic
evaluation of adrenal lesion in 26 dogs. Vet Radiol 1998;38:448–55.
24. Feldman EC, Nelson RW, Feldman MS. Use of low- and high-dose
dexamethasone tests for distinguishing pituitary-dependent from adrenal
tumor hyperadrenocorticism in dogs. J Am Vet Med Assoc 1996;209:
772–5.
25. Capen CC. Endocrine system. In: Capen CC, ed. Special veterinary
pathology. Toronto: Decker Inc., 1988:369–437.
26. Ford SL, Feldman EC, Nelson RW. Hyperadrenocorticism caused by
bilateral adrenocortical neoplasia in dogs: four cases (1983–1988). J Am
Vet Med Assoc 1993;202:789–92.
27. Von Dehn BJ, Nelson RW, Feldman EC, Griffey SM. Pheochromocytoma
and hyperadrenocorticism in dogs: six cases (1982–1992). J Am Vet Med
Assoc 1995;207:322–4.
 Iatrogenic Hyperadrenocorticism
in 28 Dogs
Twenty-eight dogs with iatrogenic hyperadrenocorticism were studied. The most common
clinical signs were cutaneous lesions (27/28), polydipsia (21/28), polyuria (19/28), and lethargy
(16/28). The most predominant findings on biochemical profile were elevated alkaline
phosphatase (ALP, 15/28) and alanine transferase (ALT, 14/28); hypercholesterolemia (14/28);
elevated aspartate transferase (AST, 12/28); and elevated triglycerides (12/18). Baseline
cortisol levels of all 28 dogs were at the lower end of the reference range and exhibited
suppressed or no response to adrenocorticotropic hormone (ACTH) stimulation. The mean time
for each dog to show initial improvement of clinical signs after corticosteroid withdrawal was six
weeks, with another mean time of 12 weeks to demonstrate complete remission.
J Am Anim Hosp Assoc 1999;35:200–7.

Hui-Pi Huang, DVM, PhD Introduction

Heng-Leng Yang, DVM, MVM
Soa-Ling Liang, DVM, MVM
Yu-Hsin Lien, DVM, MVM
Kuang-Yang Chen, DVM, PhD
O of the pituitary-adrenocortical axis.1–7
Iatrogenic hyperadrenocorticism (IHAC) results from excessive adminis-
tration of corticosteroids.1–4 This condition can be induced by topical,
parenteral, or oral medications containing corticosteroids.5–7 As in spon-
taneous hyperadrenocorticism, the most common clinical signs for IHAC
are polyuria and polydipsia, together with cutaneous and conformational
abnormalities.1–4 The development of signs of glucocorticoid excess de-
pends on the dosage and duration of the exposure.3 The diagnosis of
IHAC is based on a history of corticosteroid administration, clinical and
physical findings consistent with hyperadrenocorticism, and evaluation
Adrenocorticotropic hormone (ACTH) release is easily suppressed by
exogenous corticosteroids due to feedback mechanisms of the hypotha-
lamic-pituitary-adrenocortical (HPA) axis.5 This results in failure of the
adrenal gland to respond to ACTH administration.5,8–12 In the presence of
clinical signs of hyperadrenocorticism, results of diagnostic tests to assess
the HPA axis show differentiation between IHAC and secondary iatro-
genic or spontaneous hypoadrenocorticism.1–4
The purpose of this study is to report the clinical manifestations of
IHAC in 28 dogs.

Materials and Methods

Twenty-eight dogs with cutaneous abnormalities and progressive illness,
which were referred to the National Taiwan University Animal Hospital
(NTUAH) for further investigation, were enrolled in this study. Breeds
represented included six Maltese terriers; five Pomeranians; four mon-
grels; three each of Chihuahuas, miniature pinschers, and shih tzu; and
one each of rough collie, miniature poodle, Shiba Inu, and West Highland
white terrier. There were 18 intact males, nine intact females, and one
spayed female. The mean age was 5.0±2.8 years and ranged from 10
months to 14 years. The mean body weight was 5.9 kg and ranged from
From the Department of
1.6 to 14.5 kg.
Veterinary Medicine,
National Taiwan University, The diagnosis of IHAC in the present study was based on a history of
142 Chou-San Road, treatment with oral, parenteral, or topical corticosteroids (for various skin
Taipei 106, Taiwan. disorders) for more than one month prior to referral to NTUAH; clinical

200 JOURNAL of the American Animal Hospital Association

May/June 1999, Vol. 35 latrogenic Hyperadrenocorticism 201

Table 1
Breeds of Dogs Affected With Iatrogenic Hyperadrenocorticism as Well as Route, Type,
Dose, and Frequency of Glucocorticoid Used

Case / Breed Route* Glucocorticoids† Frequency‡

Chihuahua Oral P BID
West Highland white terrier Oral Unknown BID
Shihba Inu Oral Unknown BID
Otic T BID
Chihuahua Oral P BID
Otic H BID
Pomeranian Oral Unknown BID
Topical H BID
Maltese Parenteral Unknown SID
Mongrel Parenteral T SID
Mongrel Parenteral T SID
Pomeranian Oral P BID
Parenteral T Q3D
Shih tzu Oral Unknown BID
Parenteral D Q3D
Pomeranian Oral Unknown BID
Parenteral T Q3D
Shih tzu Oral Unknown BID
Parenteral T Q3D
Mongrel Oral Unknown BID
Parenteral D Q3D
Miniature poodle Oral P BID
Parenteral T QW
Mongrel Oral Unknown BID
Parenteral T Q3D
Maltese Oral Unknown BID
Parenteral T QW
Pomeranian Oral Unknown BID
Parenteral D Q3D
Maltese Oral Unknown BID
Parenteral D Q3D
Maltese Oral Unknown BID
Parenteral D QW
Otic H BID
Topical H BID
Miniature pinscher Oral Unknown BID
Parenteral T QW
Otic H BID
Topical B BID
Rough collie Oral Unknown BID
Parenteral T QW
Topical H BID
Miniature pinscher Topical B BID
Chihuahua Topical B TID
Miniature pinscher Topical B QID
Maltese Topical B TID
Pomeranian Otic T TID
Maltese Otic T TID
Shih tzu Opthalmic D BID

* Parenteral=subcutaneously in all dogs

†
P=prednisolone; T=triamcinolone; H=hydrocortisone; D=dexamethasone; B=betamethasone
‡
BID=twice daily; SID=once daily; Q3D=every three days; QW=every week; TID=three times daily; QID=four times
daily
202 JOURNAL of the American Animal Hospital Association
Figure 1—Excessive bilateral hair loss on truncal regions, as
well as posterior and lateral aspect of the thighs in a dog with
iatrogenic hyperadrenocorticism.
and cutaneous abnormalities which were consistent with
canine hyperadrenocorticism; and a suppressed cortisol
response to ACTH administration. The duration of treat-
ment for the various skin disorders, administration routes,
clinical presentation, and findings on physical examina-
tions were recorded. For each case, a complete blood cell
count (CBC),a a serum biochemical profile,b and an
ACTH stimulation test were also carried out at initial
consultation. The ACTH stimulation test was performed
using synthetic ACTH c at 0.25 mg intramuscularly.
Samples for serum cortisol determination were collected
prior to and one hour after ACTH administration. The
cortisol levels were measured using a validated
radioimmunoassay.d
The blood sodium to potassium ratios were monitored
every two to four weeks, after IHAC was diagnosed and
throughout the period of IHAC resolution.
Results
The mean period of treatment for skin disorders prior to
referral to NTUAH was 9.4 months and ranged from one
to 36 months. Among these 28 dogs, two were medicated
orally; three were medicated both orally and topically;
May/June 1999, Vol. 35
Figure 2—Lateral survey radiograph demonstrating the
formation of calcinosis cutis over the spinous processes of the
thoracic vertebrae in a dog with iatrogenic hyperadrenocorticism.
This is the same dog as seen in Figure 1.
Figure 3—As the iatrogenic hyperadrenocorticism progressed,
alopecia often extended from the head to the tail in affected
dogs.
three were medicated parenterally only; 13 were medi-
cated both orally and parenterally; and seven were medi-
cated only topically (skin, ear, or eye) during the
treatment period. Routes of medicine administration and
frequency are summarized in Table 1.
Among these 28 dogs, 27 showed skin lesions at the
time of referral. The most apparent clinical abnormali-
ties were cutaneous signs, including localized or gener-
alized thinning of the hair coat or excessive hair loss (23/
28) and localized or generalized alopecia (19/28). Exces-
sive hair loss or alopecia was bilateral and involved
mainly the frontal and temporal regions of the head,
pinna, truncal regions, and the medial and posterior as-
pects of the thighs [Figure 1]. Seventeen dogs (17/28)
suffered from localized or generalized pyoderma, and
nine dogs (9/28) had developed cutaneous hyperpigmen-
tation. Calcinosis cutis was found in four dogs (4/28)
[Figure 2]. Thirteen dogs medicated both orally and pa-
rentally were found to have severe cutaneous lesions,
including generalized papules, pustules, crusts, scale,
May/June 1999, Vol. 35 latrogenic Hyperadrenocorticism 203

Table 2
Clinical Findings on Physical Examinations
of 28 Dogs With Iatrogenic
Hyperadrenocorticism

Clinical Findings No. of Dogs

Cutaneous lesions
Localized or generalized thin coat 23
Localized or generalized alopecia 19
Localized or generalized pyoderma 17
Hyperpigmentation 9
Calcinosis cutis 4 Figure 4—Bilateral alopecia on frontal region, ear base, and
pinna. The condition of iatrogenic hyperadrenocorticism in this
Chief complaints on presentation
dog was caused by topical medication alone.
Polydipsia 21
Polyuria 19
Lethargy 16
Polyphagia/ravenous appetite 15
Physical examination findings
Hepatomegaly 15
Pot-belly 13
Obese 8
Muscular atrophy 8

alopecia, and hyperpigmentation. The lesions usually

developed from the head initially and then extended to
the tail [Figure 3]. However, seven dogs medicated topi-
cally were found to have mild cutaneous lesions, which
were characterized by bilateral and localized alopecia
involving the temporal regions, neck, forearms, and me-
dial and posterior aspects of the thighs [Figure 4].
Apart from cutaneous abnormalities, polydipsia (21/
28) and polyuria (19/28) were most commonly reported
by the owners. More than 50% of these dogs (16/28)
were lethargic at the time of referral. Fifteen dogs (15/
28) were polyphagic or had ravenous appetites.
Hepatomegaly (15/28) was the most common finding
on physical examination. Of these 15 dogs, 13 had a pot-
bellied appearance and eight were obese. Muscular atro-
phy of both thighs and temporal regions was found in
eight dogs (8/28).
The clinical findings and results of physical examina-
tions of these 28 dogs are summarized in Table 2.
Eosinopenia (18/28) was the most predominant find-
ing on hematological profiles [Tables 3, 4]. Remaining
hematological parameters were unremarkable [Table 3].
The most predominant findings on biochemical profiles
were elevations of ALP (15/28), ALT (14/28), choles-
terol (14/28), AST (12/28), and triglycerides (12/28)
[Tables 3, 4]. Hyperglycemia was also found in 10 dogs
(10/28) [Tables 3, 4]. The remaining biochemical param-
eters were all within the reference ranges [Table 3].
Figure 5—Hair color change from the original black color to
white on the frontal and temporal regions in a dog with resolving
iatrogenic hyperadrenocorticism. This is the same dog as in
Figure 4.
Subnormal baseline cortisol levels and a suppressed
or poor response in cortisol levels after ACTH adminis-
tration were consistently found in all 28 dogs [Table 5].
The mean concentration of baseline cortisol before
ACTH stimulation was 0.5 µg/dl and ranged from 0.1 to
1.6 µg/dl (reference range, 0.5 to 6 µg/dl). The mean
concentration of cortisol one hour after ACTH adminis-
tration was 1.1 µg/dl and ranged from 0.1 to 2.8 µg/dl
(reference range, 8 to 18 µg/dl). Sixteen dogs (16/28)
showed no response in cortisol levels after ACTH ad-
ministration, whereas 12 dogs (12/28) had a suppressed
response to ACTH.
After a diagnosis of IHAC was made, no medication
was prescribed to these dogs except in those dogs (17/
28) that suffered from secondary, generalized pyoderma.
These dogs were prescribed cephalexin orally (25 to 35
mg/kg, bid) for four to eight weeks, depending on the
severity of the skin conditions.
Twenty-eight dogs (28/28) recovered from their IHAC
without progressing into secondary hypoadrenocorticism.
204 JOURNAL of the American Animal Hospital Association May/June 1999, Vol. 35

Table 3
Results of Laboratory Tests From 28 Dogs With Iatrogenic Hyperadrenocorticism

Value* Reference Range Mean Standard Deviation Range

RBC 4.2–7.9x106/µl 6.2 1.3 4.1–8.1
Hb 9.6–19.1 g/dl 14.7 2.9 9.5–19.3
Hct 28.4–53.1% 42.8 7.9 28.5–54.0
WBC 7,500–18,400/µl 15,567 7,163 7,100–46,000
neut 4,275–16,560/µl 12,298 1,401 4,606–44,160
eos 75–1,472/µl 249 265 0–882
lymph 375–5,888/µl 2,148 1,339 242–3,920
mono 75–1,288/µl 934 420 89–2,516
Alb 2.6–3.7 g/dl 3.4 0.4 2.5–4.5
ALKP 27–219 U/l 907 896 17–3,788
ALT 20–123 U/l 390 376 14–1,792
AST 6–60 U/l 91 70 6–431
BUN 7–27 mg/dl 18 8.6 4–70
Chol 115–243 mg/dl 289 125 120–500
Crea 0.5–1.3 mg/dl 0.9 0.4 0.4–3.6
Glu 82–127 mg/dl 156 63 97–313
TP 5.4–7.8 g/dl 6.7 0.6 5.3–7.7
Trig 30–87 mg/dl 119 75 26–388
Na 136–151 mmol/l 140 6 127–155
K 3.2–4.9 mmol/l 4.0 0.4 3.7–4.8
Cl 99–116 mmol/l 103 9 89–118
Na/K 27–42 35.7 3.9 29.4–51.3

* Hb=hemoglobin concentration; Hct=microhematocrit; RBC=red blood cell count; WBC=white blood cell count;
neut=neutrophil; eosi=eosinophil; baso=basophil; lym=lymphocyte; mono=monocyte; Alb=albumin; ALKP=alkaline
phosphatase; ALT=alanine transferase; AST=aspartate transferase; BUN=blood urea nitrogen; Cl=chloride;
Chlo=cholesterol; Crea=creatinine; Glu=glucose; Na=sodium; K=potassium; TP=total protein; Trig=triglyceride

Mean time from initial withdrawal of medication to clini-
cal improvement was six weeks. Cessation of polydipsia,
polyuria, and polyphagia were the first signs of recovery,
followed by weight loss and a tight-bellied appearance.
At this stage, liver size had reduced, and hepatic enzyme
levels were within the reference ranges for most dogs
(24/28). Pyoderma and excessive hair loss or alopecia
remained at this stage. It took a further mean time of 12
weeks (range, eight to 24 weeks) before there was a
complete remission of these cutaneous lesions. How-
ever, color changes in patchy patterns of new hair growth
were found in dogs (19/19) which previously had partial
or generalized alopecia. The original white hair coat
became darker, whereas the original dark or black hair
coat became lighter or grayish [Figure 5]. Apart from
color change in hair coat, calcinosis cutis (4/4) was the
only cutaneous lesion that appeared to be permanent.
Dogs with IHAC induced by topical medication exhib-
ited complete remission of the hair coat changes at ap-
proximately five weeks after topical medications were
withdrawn. Similar hair color changes were also ob-
served in these dogs.
Three dogs (3/28) had concurrent diabetes mellitus
while suffering from IHAC. Of these three dogs, the
IHAC in two dogs was caused by systemic medication
(parenteral triamcinolone once daily, intermittently for
three years), with IHAC induced by topical 0.1%
betamethasone ointment (twice daily over the entire body
for three years to treat recurrent papules and pustules) in
the third dog. All three dogs were ovariohysterectomized
after the diagnosis of diabetes mellitus was made. Com-
plete remission of IHAC also occurred in these three
dogs. However, diabetes mellitus persisted and was con-
trolled using isophane insuline (0.25 to 1 U/kg subcuta-
neously, once daily).
Discussion
The cases of IHAC described in the present study showed
many of the clinical signs consistent with hyperadreno-
corticism described in other studies.1–4 While many
May/June 1999, Vol. 35 latrogenic Hyperadrenocorticism 205

Table 4 Table 5
Abnormalities in Hematological and Cortisol Levels Before and After
Biochemical Profiles* From 28 Dogs With Adrenocorticotropic Hormone (ACTH)
Iatrogenic Hyperadrenocorticism Stimulation From 28 Dogs With Iatrogenic
Hyperadrenocorticism
No. of Dogs
Abnormal Findings on Hematological Profile PreACTH Cortisol PostACTH Cortisol
Case (µg/dl) (µg/dl)
Eosinopenia 18
Lymphopenia 7 1 0.1 0.2
Monocytosis 7 2 0.5 0.4
Abnormal Findings on Biochemical Profile 3 0.3 0.3
Elevated ALP level 15 4 0.2 0.3
Elevated ALT level 14 5 0.5 0.2
Hypercholesterolemia 14 6 0.1 2.8
Elevated AST level 12 7 0.1 0.1
Hypertriglyceridemia 12 8 0.3 0.4
Hyperglycemia 10 9 0.1 2.6
Abnormal Adrenocortical Function 10 0.1 0.3
No response in cortisol levels 16 11 1.4 2.4
after ACTH stimulation 12 1.6 2.6
Suppressed cortisol level after 12 13 0.1 0.1
ACTH stimulation 14 1.6 2.3
15 0.1 0.1
* ALP=alkaline phosphatase; ALT=alanine
aminotransferase; AST=aspartate aminotransferase; 16 0.1 0.1
ACTH=adrenocorticotropic hormone 17 0.1 0.1
18 0.5 1.0
19 0.4 2.6
breeds were included in this study, toy breeds were over- 20 0.9 2.2
represented. This was presumed to reflect the popularity
21 0.8 2.0
of these breeds in this area13 rather than a predisposition.
22 1.6 2.6
Although twice as many males as females were af-
23 0.3 0.6
fected with IHAC, as yet no study has established that
IHAC has a sex predisposition. 24 0.1 0.1
Clinical manifestations in these dogs were consistent 25 1.1 2.2
with spontaneous hyperadrenocorticism.1–4 Cutaneous 26 1.1 1.0
abnormalities were the most common signs. Dogs with 27 0.1 0.3
generalized thin hair coat or alopecia were found to have 28 0.8 2.0
been medicated orally, parenterally, or both for pre-
existing skin disorders. In the cases of generalized thin
hair coat, alopecia, or pyoderma, the lesions were ini-
tially noted on the face and head. Toward the end of the cation. Observable signs were only evident in dogs which
course, the back, flanks, hind legs, and tail became in- were medicated orally, parenterally, or both for pre-
volved. Although the exact compounds of corticoster- existing skin disorders. None of these signs developed in
oids could not be confirmed in some cases using oral dogs that were given only topical treatment.
medication, the severity of excessive hair loss was re- Eosinopenia and lymphopenia, which have been fre-
lated to routes of administration. Regardless of duration, quently reported in dogs with spontaneous hyperadreno-
topical medications (skin, ear, or eye preparations) in- corticism,1–4 were the most common findings in the
duced mild and localized rather than generalized hair hematological profiles in these dogs. However, only dogs
loss or alopecia. which were medicated via oral or parenteral administra-
Apart from skin manifestations, polydipsia, polyuria, tion showed these findings. In the present study, elevated
lethargy, and polyphagia were also consistent with spon- levels of ALP, ALT, and AST were the most common
taneous hyperadrenocorticism.1–4 The severity of these abnormalities in the biochemical profiles, and these find-
signs was also found to be related to the routes of medi- ings agreed with most published studies.1–4 All dogs
206 JOURNAL of the American Animal Hospital Association
demonstrating a pot-belly and hepatomegaly also had
elevated levels of hepatic enzymes. In those dogs with
hypercholesterolemia and hypertriglyceridemia, raised
levels of hepatic enzymes were also found.
Diabetes mellitus is one of the medical complications
associated with hyperadrenocorticism. Glucocorticoids
increase gluconeogenesis, act as an insulin antagonist,
and can induce a diabetic state.14 In the present study, the
concurrent diabetes mellitus in three dogs was not
thought to be caused by glucocorticoid abuse alone be-
cause these dogs were obese, aged, and intact females.15
Progesterone, one of the insulin antagonists, rises dra-
matically during diestrus. Therefore, older bitches may
occasionally develop diabetes during diestrus.15 Together
with the action of another insulin antagonist, exogenous
glucocorticoids, diabetes mellitus then developed. In the
authors’ three cases, ovariohysterectomy was performed
and the glucocorticoids medication discontinued, and
the signs of IHAC showed complete remission. How-
ever, the condition of diabetes mellitus persisted but was
well controlled with insulin.
The HPA axis in dogs is easily suppressed by exog-
enous glucocorticoids. These glucocorticoids suppress
the HPA axis in 12 to 36 hours after administration and
suppress the levels of cortisol via the negative feedback
mechanism. 5,7,11,12 The ACTH stimulation test has been
proven to be a sensitive indicator of adrenocortical sup-
pression and an excellent means of determining a return
of the HPA axis to normal function.1,2,4,5 However, dogs
with hypoadrenocorticism also have a suppressed HPA
axis.1,3,4 The cause of hypoadrenocorticism can be either
spontaneous or iatrogenic. Secondary iatrogenic hypo-
adrenocorticism was considered as the major differential
diagnosis in the present study. Adrenal gland atrophy
due to a severely depressed HPA axis can be caused by
long-term corticosteroid abuse.3,11,12,16 Secondary iatro-
genic hypoadrenocorticism can be induced by long-term
corticosteroid use with sudden withdrawal. 16,17 The
pathophysiological changes characteristic of hypoadre-
nocorticism are serum electrolyte alternations. Lack of
aldosterone secretion, as a result of adrenocortical insuf-
ficiency, leads to impaired serum sodium and chloride
conservation and potassium excretion. Significant hy-
ponatremia and hyperkalemia develop, and the sodium
to potassium ratio decreases.16,17 The ACTH stimulation
test and the blood sodium to potassium ratio can be used
as aids in the differential diagnosis between IHAC and
secondary iatrogenic hypoadrenocorticism.1,16,17 In the
present study, the blood sodium to potassium ratio was
monitored every two to four weeks for each dog until
complete remission. The blood sodium to potassium ra-
tios of the 28 dogs in this study were within reference
range during the course of IHAC. All dogs recovered
from IHAC without developing secondary iatrogenic hy-
poadrenocorticism.
May/June 1999, Vol. 35
It is known that topical application of corticosteroids
rapidly suppresses the HPA axis, with plasma cortisol
concentrations becoming significantly depressed within
seven hours after the first treatment.8,9 Repeated topical
applications of corticosteroids will continue to suppress
plasma concentrations of ACTH and cortisol and con-
tinue to reduce the response to exogenous ACTH.7,11,12
Prolonged treatment for three weeks or more induced
marked suppression of the adrenal gland’s response to
exogenous ACTH in dogs.10–12 The duration and dosage
of systemic corticosteroid administration to induce clini-
cal signs of IHAC in dogs have not been determined but
appear to be dependent on the type of corticosteroids
used and their duration of action (i.e., higher doses of
longer-acting preparations appeared to induce IHAC
more rapidly than the other glucocorticoid preparations
used). Although the exact compounds of corticosteroids
used could not be confirmed in all 28 dogs of this study,
the compounds for both parenteral and topical prepara-
tion may be short-acting (as with hydrocortisone), inter-
mediate-acting (as with triamcinolone), or long-acting
(as with betamethasone and dexamethasone).2,4 These
compounds are widely available in parenteral, skin, oph-
thalmic, and aural preparations.
In the present study, the mean time for initial clinical
signs of IHAC to develop was around nine months,
although signs did develop in as little as one month. It
has been shown that after cessation of corticosteroid
therapy, HPA axis returns to normal values by four weeks
after the last daily treatment in dogs treated with triam-
cinolone acetonide or with betamethasone valerate.12 In
the present study, the HPA axis of these dogs was not
monitored after medication was withdrawn due to lack
of cooperation from the owners.
It took an average of six weeks for dogs to show
clinical improvement or a return to normal hepatic en-
zyme levels after corticosteroid withdrawal. On average,
it took another 12 weeks for skin and hair coat conditions
to exhibit complete recovery, except for calcinosis cutis
and the hair color changes which persisted. For those
cases that were induced by topical application of corti-
costeroids, complete recovery of their hair coat and skin
conditions occurred within five weeks after the topical
application was discontinued.
Conclusion
Iatrogenic hyperadrenocorticism is a common endocri-
nopathy seen in the authors’ hospital. Twenty-eight dogs
with a history of long-term oral, parenteral, or topical
glucocorticoid use were examined. The most common
clinical findings were consistent with spontaneous hy-
peradrenocorticism, including cutaneous lesions and el-
evated levels of ALP, ALT, and AST. The ACTH
stimulation test was the best test to differentiate sponta-
neous hyperadrenocorticism from IHAC. The blood so-
May/June 1999, Vol. 35
dium to potassium ratio was a helpful indicator to
differentiate IHAC from secondary iatrogenic hypoadre-
nocorticism. The pathophysiological change characteris-
tic of hypoadrenocorticism is a decreased serum sodium
to potassium ratio. Among these cases, clinical signs and
abnormalities in laboratory tests were less remarkable in
dogs with IHAC induced by topical corticosteroids in
comparison to those with the same condition caused by
parenteral or oral corticosteroids. Also, a shorter period
of time was required to exhibit complete remission of
this condition in dogs when it was induced by topical
corticosteroids rather than by parenteral or oral cortico-
steroids.
a suppression associated with topical otic administration of glucocorticoids in
Sysmex K-1000; Toa Medical Electronics Co., Ltd., Japan
b
Kodak Etachem DT 60 II; Eastman Kodak Company, Rochester, NY
c
Cortrosyn; Organon, Oss, Holland
d 1987;191:685–8.
Coat-A Count Cortisol; Diagnostic Products Cooperation, Webster, TX
e 13.
Insulatard HM; Novo Nordisk A/S, Denmark
References
1. Nelson RW, Couto CG. Essentials of small animal internal medicine. St.
Louis: Mosby Year Book, 1992:587–97.
2. Chastain CB, Gabjam VK. Clinical endocrinology of companion animals.
Philadelphia: Lea & Febiger, 1986:409–30.
3. Rijnberk A, ed. Clinical endocrinology of dogs and cats. Dordrecht: Kluwer
Academic Publishers, 1996:88–91.
4. Feldman EC, Nelson RW. Canine and feline endocrinology and
reproduction. Philadelphia: WB Saunders, 1996:333–5.
latrogenic Hyperadrenocorticism 207
5. Eichenbaum JD, Macy DW, Severin GA, Paulsen ME. Effect in large dogs
of ophthalmic prednisolone acetate on adrenal gland and hepatic function.
J Am Anim Hosp Assoc 1988;24:705–9.
6. Murphy CJ, Feldman E, Bellhorn R. Iatrogenic Cushing’s syndrome in a
dog caused by topical ophthalmic medications. J Am Anim Hosp Assoc
1990;26:640–2.
7. Glaze MB, Crawford MA, Nachreiner RF, Casey HW, Nafe LA, Kearney
MT. Ophthalmic corticosteroid therapy: systemic effects in the dog. J Am
Vet Med Assoc 1988;192:73–5.
8. Kemppainen RJ, Lorenz MD, Thompson FN. Adrenocortical suppression in
the dog after a single dose of methylprednisolone acetate. Am J Vet Res
1981;42:22–4.
9. Kemppainen RJ, Lorenz MD, Thompson FN. Adrenocortical suppression in
the dog given a single intramuscular dose of prednisolone or triamcinolone
acetonide. Am J Vet Res 1982;42:204–6.
10. Moore GE, Hoening M. Duration of pituitary and adrenocortical
suppression after long-term administration of anti-inflammatory doses of
prednisolone in dogs. Am J Vet Res 1992;53:716–20.
11. Moriello KA, Fehrer-Sawyer SL, Meyer DJ, Feder B. Adrenocortical
dogs. J Am Vet Med Assoc 1988;193:329–31.
12. Zenoble RD, Kemppainen RJ. Adrenocortical suppression by topically
applied corticosteroids in healthy dogs. J Am Vet Med Assoc
Huang H-P, Chaing G-H, Wang C-H. Canine hematology reference values
in Veterinary Hospital of National Taiwan University. Memoirs of the
College of Agriculture, National Taiwan University 1995;35:120–9.
14. Feldman EC, Nelson RW. Canine and feline endocrinology and
reproduction. Philadelphia: WB Saunders, 1996:206–15.
15. Feldman EC, Nelson RW. Canine and feline endocrinology and
reproduction. Philadelphia: WB Saunders, 1996:339–83.
16. Feldman EC, Nelson RW. Canine and feline endocrinology and
reproduction. Philadelphia: WB Saunders, 1996:266–81.
17. Kintzer PP, Peterson ME. Hypoadrenocorticism in dogs. In: Bonagura JD,
Kirk RW, eds. Kirk’s current veterinary therapy XII small animal practice.
Philadelphia: WB Saunders, 1995:425–9.
 Radiographic Findings in Dogs
With Naturally-Occurring Primary
Hypoadrenocorticism

Survey radiographs often are obtained in dogs with primary hypoadrenocorticism in adrenal
crisis as part of the routine evaluation of a critically ill dog. In this study, standardized methods
of cardiac, pulmonary vasculature, and vena cava mensuration were used in 22 dogs with
naturally-occurring primary hypoadrenocorticism, and the findings were compared with those in
22 breed-matched, clinically normal dogs. Most (81.8%) untreated dogs with primary
hypoadrenocorticism had one or more radiographic abnormalities, including small size of the
heart (45.5%), cranial lobar pulmonary artery (36.4%), caudal vena cava (54.5%), or liver
(36.4%). Megaesophagus was not found in any of the dogs with hypoadrenocorticism, and
therefore, compared to the other common radiographic findings, should be considered a rare
finding. J Am Anim Hosp Assoc 1999;35:208–12.

Carlos Melián, DVM Introduction

Joseph Stefanacci, VMD
Mark E. Peterson, DVM,
Diplomate ACVIM
Peter P. Kintzer, DVM,
Diplomate ACVIM
RS
From the Department of Medicine,
The Animal Medical Center (Melián,
Stefanacci, Peterson),
510 East 62nd Street,
New York, New York 10021 and the
Departments of Environmental Studies and
Medicine (Kintzer), Tufts Northeastern
Veterinary Medical Center,
North Grafton, Massachusetts 01536.
Address all correspondence and reprint
requests to Dr. Peterson.
Doctor Melián’s current address is
Perez Galdós, 16, 35002
Las Palmas, Spain.
Doctor Kintzer’s current address is
Boston Road Animal Hospital,
1235 Boston Road,
Springfield, Massachusetts 01119.
In dogs, naturally-occurring primary hypoadrenocorticism (i.e., Addison’s
disease) is an uncommon disease caused by deficient adrenal production
of glucocorticoids and mineralocorticoids.1–3 Hyperkalemia and hy-
ponatremia are the most common laboratory abnormalities. Common
clinical signs include anorexia, vomiting, lethargy, weakness, and col-
lapse. These signs may be chronic or intermittent, but over a third of dogs
present as an acute adrenal crisis at the time of initial diagnosis.3,4
Survey thoracic and abdominal radiographs are often obtained in dogs
with primary hypoadrenocorticism in adrenal crisis as part of the routine
evaluation of a critically ill dog. Radiographic findings, as described in
textbooks, include an abnormally small heart, pulmonary vessels, caudal
vena cava, and liver, and less commonly, a dilated, air-filled esopha-
gus.2,5,6 Decreases in heart size have been reported in dogs with experi-
mentally-induced hypoadrenocorticism (before and after bilateral
adrenalectomy),7 but only a few reports mention radiographic findings in
dogs with naturally-occurring hypoadrenocorticism.3,4,8–10
One problem with published reports of dogs with naturally-occurring
hypoadrenocorticism is that the radiographic size of the heart, cranial
lobar pulmonary arteries, and caudal vena cava was assessed subjectively.
Some guidelines have been given for more exact determination of the size
of these structures; however, most of these methods have been used to
identify cardiac or vasculature enlargement rather than reduction in size.11–13
The purpose of this study was to critically evaluate thoracic and
abdominal radiographs taken in dogs with naturally-occurring hypoadreno-
corticism. Standardized methods of cardiac, pulmonary vasculature, and
vena cava mensuration were used, and the findings were compared with
those in normal dogs of the same breed. Hepatic size also was evaluated
and compared.

208 JOURNAL of the American Animal Hospital Association

May/June 1999, Vol. 35
Materials and Methods
Dogs
Medical records and survey radiographs of 22 dogs with
naturally-occurring primary hypoadrenocorticism were
evaluated at The Animal Medical Center and Tufts Uni-
versity Veterinary Teaching Hospital. All of these dogs
were considered to have moderate to severe adrenal cri-
sis at the time of examination. Twenty-one of these dogs
have each been included in one of two other studies on
diagnosis and treatment of primary hypoadreno-
corticism.3,4 Twenty-two breed-matched, clinically nor-
mal dogs were used to establish a normal reference range.
Dogs With Hypoadrenocorticism: The 22 dogs with
primary hypoadrenocorticism ranged in age from one to
10 years (mean±standard deviation [SD], 5.1±2.4 yrs).
Breeds included West Highland white terrier (n=6),
mixed-breed dogs (n=4), Labrador retriever (n=2), stan-
dard poodle (n=2), Airedale terrier (n=1), basset hound
(n=1), bichon frise (n=1), Chihuahua (n=1), German
shepherd dog (n=1), Great Dane (n=1), miniature poodle
(n=1), and pointer (n=1). Body weights ranged from 2.7
kg to 43.9 kg (mean±SD, 17±13.9 kg). Fourteen (63.6%)
of the 22 dogs were female, and eight were male; 12
females and four males were neutered.
A tentative diagnosis of primary hypoadrenocorticism
was made on the basis of clinical signs and routine
laboratory findings consistent with moderate to severe
disease (e.g., depression, weakness, collapse, vomiting,
anorexia, hyperkalemia, and hyponatremia). The diag-
nosis was confirmed in all dogs by determination of low
to low-normal baseline serum cortisol concentrations
that failed to increase one hour after intravenous (IV)
administration of 0.25 mg synthetic adrenocorticotropic
hormonea (ACTH).2–4
Normal Dogs: Twenty-two clinically normal dogs un-
dergoing elective procedures (e.g., elective surgery, an-
nual vaccination, and heartworm test) were used as
controls. The dogs ranged in age from one to 13 years
(mean±SD, 7.1±3.6 yrs). These dogs were matched by
breed and weight to the dogs with hypoadrenocorticism.
Twelve (54.5%) of the 22 dogs were female, and 10 were
male; 10 females and four males were neutered. Body
weights ranged from 3.1 kg to 45 kg (mean±SD,
17.3±14.1 kg). None of these dogs had serum cortisol
concentrations determined; however, all 22 dogs were
examined and were considered healthy on the basis of
history and physical examination findings as well as
results of routine laboratory testing (i.e., complete blood
count and serum biochemical analysis, which included
the measurements of both sodium and potassium con-
centrations).
Radiographs
Right lateral survey radiographs of the thorax were used
for measurements of the heart, pulmonary artery, and
Radiography in Dogs With Hypoadrenocorticism 209
Figure 1—Diagram of the lateral view of the thorax of a dog,
illustrating the vertebral heart size measurement method. The
long axis (L) and short axis (S) heart dimensions are transposed
onto the vertebral column and recorded as the number of
vertebrae beginning with the cranial edge of the fourth thoracic
(T4) vertebra. These values are then added to obtain the
vertebral heart size; in this example, 5.9 vertebra (long axis) plus
4.5 vertebra (short axis) equals a vertebral heart size of 10.4
vertebra.
caudal vena cava. Presence or absence of megaesopha-
gus was determined on right lateral and dorsoventral
survey radiographs of the thorax by use of previously
described radiographic parameters.14,15 Liver size was
assessed on right lateral and ventrodorsal abdominal ra-
diographs. In dogs with hypoadrenocorticism, all radio-
graphs evaluated were taken before treatment with IV
fluids and corticosteroids. In the clinically normal con-
trol dogs, all radiographs evaluated were taken before
any elective procedures were performed.
Heart Size: Heart size was evaluated by a vertebral
scale system.11 The long axis of the heart was measured
as the distance from the ventral border of the left main
stem bronchus to the most distant ventral contour of the
cardiac apex [Figure 1]. This dimension reflects the com-
bined size of the left atrium and the left ventricle. The
measurement was made by use of a caliper, which was
then repositioned over the thoracic vertebrae beginning
with the cranial edge of the fourth thoracic (T4) vertebra
and extending caudally and parallel to the vertebra. The
maximal short axis of the heart in the central third re-
gion, perpendicular to the long axis, was recorded in the
same manner starting at T4 [Figure 1]. The short and long
axis dimensions then were added to yield a verte-
brae:heart sum as an expression of heart size in relation
to a vertebral indicator of body length. The overall size
of the heart was thus expressed as total units of vertebral
210 JOURNAL of the American Animal Hospital Association
length to the nearest 0.1 vertebra and termed the verte-
bral heart size.11
Cranial Lobar Pulmonary Artery: The pulmonary
artery:fourth rib ratio was calculated to assess the size of
the pulmonary artery.13 The width of the most isolated
cranial lobar pulmonary artery in the right lateral view
was used for measurement. The width of this artery was
measured in millimeters at the point where it crosses the
fourth rib. The width of the fourth rib at its proximal one-
third length was measured in millimeters. These mea-
surements were used to calculate the ratio.
Caudal Vena Cava: The width of the caudal vena
cava was measured in millimeters at equal distances
from the heart and diaphragm. The length of the fifth
thoracic (T5) vertebra in millimeters was then measured,
and the caudal vena cava:length of the T5 vertebra ratio
was calculated.11
Liver Size: The size of the liver was assessed by
determining the axis of the stomach; the position of the
pyloric region, right kidney, and transverse colon; and
the relationship of the caudal ventral margin of the he-
patic silhouette to the costal arch. A diagnosis of
microhepatica was based on a cranioventral shift in the
axis of the gastric silhouette; a cranial displacement of
the pylorus, right kidney, and transverse colon; and find-
ing the caudal ventral margin of the hepatic silhouette
well within the costal arch.12
In one dog with hypoadrenocorticism and microhe-
patica, radiographs were repeated and liver size was
assessed again after 48 hours of treatment with IV fluids
and corticosteroids [Figure 3].
Statistical Analysis
Results are given as the range and mean±SD. Statistical
analysis was performed by the unpaired Student’s t-test
and chi-square test, as appropriate. 16 Significance was
defined as a value of p of 0.05 or less.
Results
The mean±SD vertebral heart size (9.2±0.5 vertebra) in
the 22 dogs with hypoadrenocorticism was significantly
(p less than 0.0001) smaller than that of the 22 clinically
normal dogs (10.5±0.7 vertebra). Ten (45.5%) of the 22
dogs with hypoadrenocorticism had a vertebral heart
size below the low end of the reference range (9.2 to 11.6
vertebrae) established by assessment of the 22 normal
dogs [Figure 2].
The mean pulmonary artery:fourth rib ratio
(0.37±0.16) in the 22 dogs with hypoadrenocorticism
was significantly (p of 0.0003) lower than that of the 22
clinically normal dogs (0.56±0.15). Eight (36.4%) of the
dogs with hypoadrenocorticism had a pulmonary
artery:fourth rib ratio below the reference range (0.3 to
0.8) established by assessment of the 22 normal dogs
[Figure 2].
May/June 1999, Vol. 35
Figure 2—Scatterplots of the measurements for vertebral heart
size, pulmonary artery:fourth rib ratio, and the caudal venal
cava:fifth thoracic vertebra ratio in 22 dogs with hypoadreno-
corticism ( ) and 22 breed-matched clinically normal dogs ( ).
The horizontal bars indicate mean values.
The mean vena cava:fifth thoracic vertebra ratio in
the 22 dogs with hypoadrenocorticism (0.61±0.13) was
significantly (p less than 0.0001) lower than that of the
22 clinically normal dogs (0.79±0.12). Twelve (54.5%)
of the dogs with hypoadrenocorticism had a vena
cava:fifth thoracic vertebra ratio below the reference
range (0.6 to 1.0) established by assessment of the 22
normal dogs [Figure 2].
The size of the liver was significantly (p of 0.0018;
chi-square test) smaller in the 22 dogs with hypoadreno-
corticism than in the 22 clinically normal dogs. In eight
(36.4%) of the dogs with hypoadrenocorticism, the liver
was considered small [Figure 2], whereas the liver size
was considered within reference range in the 22 normal
dogs. In one dog with microhepatica in which abdominal
radiographs were repeated 48 hours after treatment, the
liver size normalized after the correction of hypovolemia
and initiation of corticosteroid treatment [Figure 3].
Of the 22 dogs with hypoadrenocorticism, 18 (81.8%)
had one or more radiographic abnormalities consistent
with their disease (i.e., microcardia, small size of the
cranial lobar pulmonary artery or caudal vena cava, or
microhepatica). Three of the 18 dogs had all four abnor-
malities; four dogs had three abnormalities; three dogs
had two abnormalities; and eight dogs had one abnor-
mality. Only four (18.2%) of the 22 dogs with hypoadreno-
corticism had no radiographic abnormalities.
Discussion
The results of this study indicate that most untreated
dogs with primary hypoadrenocorticism have one or more
radiographic abnormalities, including small size of the
heart, cranial lobar pulmonary artery, caudal vena cava,
and liver. Each of these abnormalities were found in
33% to 50% of the 22 dogs with hypoadrenocorticism
studied in this report. The cause for the development of
May/June 1999, Vol. 35
Figure 3A—Right lateral abdominal radiograph of a three-year-
old, female, spayed West Highland white terrier with confirmed
hypoadrenocorticism. Notice the cranioventral shift in the axis of
the gastric silhouette and cranial displacement of the pylorus.
The caudoventral border of the liver is round and located more
cranially, lying well within the costal arch.
Figure 3B—Right lateral abdominal radiograph of the dog in
Figure 2A after 48 hours of treatment (i.e., intravenous fluids and
corticosteroids) for hypoadrenocorticism. The liver size is
increased when compared to pretreatment radiographs, with the
caudoventral margin extending beyond the costal arch. The
stomach axis is parallel with the ribs and has moved more
caudally, compared with the pretreatment radiograph.
these abnormalities is not known, but all can be attrib-
uted to hypovolemia.5,17
Radiographic evaluation of the size of the heart is
often done empirically, rather than by use of established
methods of measuring cardiac and other anatomic struc-
tures.18 Differences in conformation of the thorax among
dog breeds and the variation resulting from the influence
of cardiac and respiratory cycles have prevented the
development of an accurate reference range applicable to
all dog breeds. A guideline of 2.5 to 3.5 intercostal
spaces on lateral radiographic views indicating normal
heart size in dogs is used by many radiologists and
cardiologists.18 Limitations of this method include varia-
tion between dogs in the axis position of the heart and
conformation of the thorax and the effect of phase of
respiration, superimposition of the ribs, and imprecise
measurement points.11 Comparative methods of measur-
Radiography in Dogs With Hypoadrenocorticism 211
ing cardiac size can be used by establishing ratios
between cardiac dimensions and other anatomical struc-
tures, preferably in the same animal prior to the develop-
ment of the disease or in a normal dog of the same
breed.18
Recently, a method of assessing canine heart size by a
vertebral scale system was developed to overcome most
of these limitations,11 and this scale was used in this
study. The vertebral scale system has not been used to
evaluate smaller-than-normal cardiac size, but this
method appeared to work well in the dogs of this study.
The prevalence of microcardia in the dogs of this study
was similar to that reported subjectively in previous
studies of primary hypoadrenocorticism.3,4
As with evaluation of cardiac size, radiographic evalu-
ation of the size of the cranial lobar pulmonary artery or
caudal vena cava is often empirical. Some guidelines
have been proposed that determine large size of pulmo-
nary artery on the basis of the pulmonary artery:fourth
rib ratio and large size of the caudal vena cava on the
basis of the caudal vena cava:length of the T5 vertebra
ratio.11,13 The reference ranges for these ratios deter-
mined by this study were similar to those previously
established by other studies.11,13 These ratios have not
been commonly used in determining smaller-than-nor-
mal vessel size, but both measurements appeared to work
well in the dogs of this study. The prevalence of reduced
size of the pulmonary artery or caudal vena cava in the
dogs of this study was similar to that reported subjec-
tively in previous studies of primary hypoadreno-
corticism.3
Radiographic evaluation of liver size is not mentioned
commonly among radiographic findings in dogs with
hypoadrenocorticism.4,10 However, in this study, micro-
hepatica was as common as other typical findings. Small
liver size most likely was related to hypovolemia, as
supported by normalization of liver size in one dog after
fluid and corticosteroid replacement.
A dilated esophagus (i.e., megaesophagus), although
reported in five dogs with hypoadrenocorticism,3,4,8–10
was not found in any of the dogs in this study. Therefore,
compared to the reduced cardiac, vessel, and liver size,
megaesophagus should be considered rare, a finding con-
firmed in two other studies of large series of dogs with
hypoadrenocorticism.3,4
Conclusion
Many dogs affected with hypoadrenocorticism have one
or more radiographic abnormalities (i.e., small size of
the heart, cranial lobar pulmonary artery, caudal vena
cava, or liver). The pathogenesis of these abnormalities
is unclear, but generally can be attributed to systemic
volume depletion.5,17 In accordance with this, it is well
recognized that dogs with volume depletion from other
causes (e.g., dehydration, shock, anemia) may exhibit
similar radiographic signs. However, if such radiographic
212 JOURNAL of the American Animal Hospital Association
abnormalities are found in a dog suspected of having
hypoadrenocorticism, it is important for the veterinarian
to rule out this disease.
a
Cortrosyn; Organon Inc., West Orange, NJ
References
1. Feldman EC, Tyrrell JB. Hypoadrenocorticism. Vet Clin N Am
1977;7:555–81.
2. Hardy RM. Hypoadrenal gland disease. In: Ettinger SJ, Feldman EC, eds.
Textbook of veterinary internal medicine. 4th ed. Philadelphia: WB
Saunders, 1995:1579–93.
3. Peterson ME, Kintzer PP. Pretreatment clinical and laboratory findings in
dogs with hypoadrenocorticism: 225 cases (1979–1993). J Am Vet Med
Assoc 1996;208:85–91.
4. Melián C, Peterson ME. Diagnosis and treatment of naturally occurring
hypoadrenocorticism in 42 dogs. J Sm Anim Pract 1996;37:268–75.
5. Feldman EC, Nelson RW. Hypoadrenocorticism (Addison’s disease). In:
Feldman EC, Nelson RW, eds. Canine and feline endocrinology and
reproduction. 2nd ed. Philadelphia: WB Saunders, 1996:266–306.
6. Ticer JW. Roentgen signs of endocrine disease. Vet Clin N Am
1977;7:465–86.
7. Rendano VT, Alexander JE. Heart size changes in experimentally induced
adrenal insufficiency in the dog: a radiographic study. J Am Vet Rad Soc
1976;17:57–66.
May/June 1999, Vol. 35
8. Bartges JW, Nielson DL. Reversible megaesophagus associated with
atypical primary hypoadrenocorticism in a dog. J Am Vet Med Assoc
1992;201:889–91.
9. Burrows CF. Reversible mega-oesophagus in a dog with hypoadre-
nocorticism. J Sm Anim Pract 1987;28:1073–8.
10. Whitley NT. Megaesophagus and glucocorticoid-deficient hypoadre-
nocorticism in a dog. J Sm Anim Pract 1995;36:132–5.
11. Buchanan JW, Bücheler J. Vertebral scale system to measure canine heart
size in radiographs. J Am Vet Med Assoc 1995;206:194–9.
12. Pechman RD. The liver and spleen. In: Thrall DE, ed. Veterinary diagnostic
radiology. 2nd ed. Philadelphia: WB Saunders, 1994:426–35.
13. Thrall DE, Losonsky JM. A method for evaluating canine pulmonary
circulatory dynamics from survey radiographs. J Am Anim Hosp Assoc
1976;12:457–62.
14. Suter PF, Lord PF. Swallowing problems and esophageal abnormalities. In:
Suter PF, ed. Thoracic radiography: a text atlas of thoracic diseases of the
dog and cat. Wettswil, Switzerland: Suter PF, 1984:324–31.
15. Watrous BM. The esophagus. In: Thrall DE, ed. Textbook of veterinary
diagnostic radiology. 2nd ed. Philadelphia: WB Saunders, 1994:236–7.
16. Zar JH. Biostatistical analysis. 2nd ed. Englewood Cliffs, NJ: Prentice-
Hall, 1984:61–78, 150–61.
17. Remington JW. Circulatory factors in adrenal crisis in the dog. Am J
Physiol 1951;165:306–18.
18. Suter PF, Lord PF. Cardiac diseases. In: Suter PF, ed. Thoracic radiogra-
phy: a text atlas of thoracic diseases of the dog and cat. Wettswil,
Switzerland: PF Suter, 1984:351–516.
 Ultrasonographic Evaluation of the
Adrenal Glands in Six Dogs With
Hypoadrenocorticism
The purpose of this study was to determine the value of ultrasonographic characterization of
the adrenal glands in dogs with hypoadrenocorticism. Measurements of adrenal glands were
obtained in six dogs with hypoadrenocorticism. The adrenal glands on both sides were shorter
(range: left adrenal gland length, 10.0 to 19.7 mm; right adrenal gland length, 9.5 to 18.8 mm)
and thinner (range: left adrenal gland thickness, 2.2 to 3.0 mm; right adrenal gland thickness,
2.2 to 3.4 mm) than in normal dogs (range: left adrenal gland length, 13.2 to 26.3 mm; right
adrenal gland length, 12.4 to 22.6 mm; left adrenal gland thickness, 3.0 to 5.2 mm; right
adrenal gland thickness, 3.1 to 6.0 mm). Statistical analysis revealed a significant reduction in
size of the left adrenal gland (p less than 0.05) in dogs with hypoadrenocorticism compared to
the left adrenal gland in normal dogs. The results of this study show that atrophy of the adrenal
glands in dogs with hypoadrenocorticism seems to lead to an ultrasonographic-measurable
reduction in size of the adrenal glands. J Am Anim Hosp Assoc 1999;35:214–8.

Angelika Hoerauf, DVM Introduction

Claudia Reusch, DVM, PhD
O atrophy of all layers of the adrenal cortex, believed to be the end result of
From the Department of
Veterinary Internal Medicine,
University of Zurich,
Winterthurerstr. 260, CH-8057
Zurich, Switzerland.
Adrenal insufficiency is an uncommon endocrine disorder in dogs, result-
ing in inadequate adrenal secretion of mineralocorticoids, glucocorti-
coids, or both. In dogs, primary hypoadrenocorticism (Addison’s disease)
is the most common form of the disease. It is caused by destruction or
an immune-mediated process.1 Secondary hypoadrenocorticism is a rare
condition which is characterized by deficient secretion of adrenocortico-
tropic hormone (ACTH) by the pituitary gland.
The disease typically occurs in young to middle-aged female dogs. The
clinical signs are often vague, vary in severity, and include anorexia,
lethargy, dehydration, weight loss, vomiting, or diarrhea. A tentative
diagnosis of primary hypoadrenocorticism is often made on the basis of
routine laboratory findings (e.g., mild anemia, eosinophilia, lymphocyto-
sis, azotemia, hyperkalemia, and hyponatremia). The definitive diagnosis
of hypoadrenocorticism requires the demonstration of a low baseline
serum cortisol concentration with a subnormal or absent response to
exogenous ACTH administration.2 Since it often takes one to several days
before results of specific tests are available, the therapy of a dog pre-
sented in hypoadrenal crisis initially is limited to symptomatic treatment.
It would be extremely useful to support the suspicion of hypoadre-
nocorticism by another diagnostic procedure which offers an immediate
diagnosis.
During the last few years, multiple imaging modalities (e.g., computed
tomography, scintigraphy, and ultrasonography) have been used for char-
acterizing adrenal glands. With computed tomography and scintigraphy,
normal as well as enlarged adrenal glands can be visualized.3,4 However,
both methods are time-consuming and not readily available.

214 JOURNAL of the American Animal Hospital Association

May/June 1999, Vol. 35 Ultrasonography in Dogs With Hypoadrenocorticism 215

Table 1
Breed, Sex, Age, and Weight of Six Dogs
With Hypoadrenocorticism

Case Age Weight

No. Breed Sex* (yrs) (kg)
1 Collie F 5 35
2 West Highland M 8 8
white terrier
3 Pitbull terrier M 8 26
4 Cocker spaniel F 6 11
cross
5 Airedale terrier M 7 23
Figure 1—Longitudinal sonogram of the left adrenal gland of a
6 Afghan hound M 5 22 healthy control dog imaged in a sagittal plane. Adrenal thickness
is represented by the “x” calipers; the long axis of the adrenal
* F=female; M=male gland is represented by the “+” calipers; phrenicoabdominalis
vein is represented by the “ ” caliper.

Department of Veterinary Internal Medicine, University

The use of ultrasonography, by contrast, has found
very wide acceptance during the last few years. Visual-
ization of nonenlarged adrenal glands by means of ultra-
sonography seemed impossible for a long time due to
their anatomical position and the small difference in
echogenicity compared to the surrounding tissue. Only
recently has it become possible, with higher quality
equipment and increasing ultrasonographer experience,
to detect and characterize the adrenal glands of healthy
dogs.5–9 Information on the size of the adrenal glands of
healthy dogs was provided for the first time by Barthez,
et al.10 In further studies it was demonstrated that ultra-
sonography is a helpful procedure in the evaluation of
adrenal glands in dogs suspected of having Cushing’s
disease. In dogs with pituitary-dependent hyperadreno-
corticism, ultrasonographic characteristics of the adrenal
glands include bilaterally-symmetrical adrenomegaly,
increased adrenal thickness, maintenance of normal ad-
renal shape and contours, and homogeneous adrenal
echogenicity in the majority of cases.11–13 In contrast,
functional adrenal tumors are mostly unilateral, of ir-
regular and rounded shape, and have mixed echogeni-
city.14–16 Results of ultrasonographic evaluation of the
adrenal glands in dogs with adrenal insufficiency have
not been reported so far; the purpose of the present study
was to determine the value of ultrasonographic charac-
terization of the adrenal glands in dogs with
hypoadrenocorticism.
Materials and Methods
Cases
This prospective study was completed with four dogs
evaluated at the Department of Veterinary Internal Medi-
cine, University of Munich, Germany between March 1,
1994 and March 1, 1995 and two dogs evaluated at the
of Zurich, Switzerland. To be included, each dog must
have been suspected to have hypoadrenocorticism on the
basis of history and physical examination and the results
of a complete blood count, serum biochemical analysis,
and urinalysis. Dogs having any notable disorder involv-
ing other organ systems and dogs having previous ste-
roid therapy were not included. The diagnosis of
hypoadrenocorticism must have been confirmed by a
low baseline serum cortisol concentration that failed to
increase one hour after intramuscular (IM) administra-
tion of one vial of synthetic ACTH.a
The group was made up of four male and two female
dogs, between five and eight years of age (median, 6.5
yrs), with body weights ranging from 8 to 35 kg (median,
22.5 kg) [Table 1].
Ultrasonographic Technique
The ultrasonographic examination was performed prior
to having the results of the ACTH stimulation testing.
The dogs were placed in dorsal recumbency. None of the
dogs were sedated or anesthetized. A real-time mechani-
cal sector scannerb with 7.5-Mhz transducer was used.
Imaging of the adrenal glands was performed by one
person (Hoerauf) using standard techniques that have
been described previously.6 The maximum length of each
adrenal gland was measured in the longitudinal plane
using electronic calipers. Adrenal thickness was defined
as the greatest dorsoventral dimension and was assessed
as a single measurement made perpendicular to the long
axis [Figure 1]. The shape and acoustic texture of each
adrenal gland were assessed subjectively and described.
Statistical Analysis
The ultrasonographically-determined adrenal gland sizes
in dogs with hypoadrenocorticism were compared with
the adrenal gland sizes determined in 20 healthy dogs
216 JOURNAL of the American Animal Hospital Association
Figure 2—Longitudinal sonogram of the left adrenal gland of a
dog with hypoadrenocorticism imaged in a sagittal plane. Adrenal
thickness is represented by the “x” calipers; the long axis of the
adrenal gland is represented by the “+” calipers; phrenico-
abdominalis vein is represented by the “ ” caliper.
Figure 3—Schematic drawing of the longitudinal sonogram of
the left adrenal gland of a dog with hypoadrenocorticism from
Figure 2.
(body weight, 2.5 to 32 kg; median, 14 kg) in an earlier
study.8 The statistical evaluation was carried out with the
help of the statistics program, Statistical Package for the
Social Sciencesc (SPSS), using nonparametric proce-
dures. Ranges and median values of adrenal gland mea-
surements are given. The Mann-Whitney U-test was used
for determining the differences between two groups. Dif-
ferences were considered significant for p of 0.05 or less.
Results
The left adrenal gland was identified and measured in
each of the six dogs, whereas the right adrenal gland was
identified only in four (case nos. 2, 4, 5, 6) of the six
dogs. The two (case nos. 1, 3) dogs in which it was not
possible to visualize the right adrenal gland were nar-
row-chested breeds (collie, pitbull terrier) and weighed
more than 25 kg (35 kg and 26 kg, respectively).
The length of the left adrenal gland varied between
10.0 and 19.7 mm (median, 13.1 mm), and the thickness
was between 2.2 and 3.0 mm (median, 2.4 mm). The
May/June 1999, Vol. 35
Figure 4—Comparison of the ultrasonographically-determined
length and thickness of the left adrenal gland in 20 healthy dogs
and six dogs with hypoadrenocorticism (minimum, maximum,
median); * = significant at p of 0.05 or less.
right adrenal gland length ranged from 9.5 to 18.8 mm,
and the thickness ranged from 2.2 to 3.4 mm. Median
values for the right adrenal gland measurements were
not given because of the low number. Table 2 contains
the ultrasonographically-determined lengths and thick-
nesses of the left and right adrenal glands in the six dogs
with confirmed hypoadrenocorticism as well as the val-
ues recently obtained from a control group of 20 healthy
dogs.8
In all the dogs with hypoadrenocorticism, the left
adrenal gland had the same shape (i.e., peanut shape) as
that of healthy dogs. However, both poles were consider-
ably thinner compared to the adrenal glands of healthy
dogs. In the dogs with hypoadrenocorticism, the comma
shape typical of the right adrenal gland in healthy dogs
was reduced to a straight line as a result of the reduction
in size of the organ.
Both adrenal glands, like those of healthy dogs, were
hypoechoic compared to the surrounding tissue. The two-
layer internal structure, as is typical of the adrenal glands
in healthy dogs, was not identifiable in the dogs with
hypoadrenocorticism [Figures 2, 3].
In five of the six dogs with hypoadrenocorticism, the
thickness of the left and right adrenal glands was mark-
edly less than the thickness of the left and right adrenal
glands respectively in the healthy control dogs; in one
dog with hypoadrenocorticism (case no. 5), the thickness
of both adrenal glands was in the lower part of the
reference range [Table 2].
A statistical comparison between healthy dogs and
dogs with hypoadrenocorticism, with respect to the length
and thickness of the adrenal glands, was carried out only
for the left adrenal gland. The values previously ob-
tained by Hoerauf and Reusch were used as standard
values for healthy dogs.8 The length and thickness of the
left adrenal gland in dogs with hypoadrenocorticism were
significantly less than those of healthy dogs (p less than
0.05) [Figure 4].
May/June 1999, Vol. 35
Table 2
Ultrasonographic Measurements of Adrenal Gland Length and Thickness in Six Dogs With
Hypoadrenocorticism and in 20 Normal Dogs
Left Adrenal
Gland Length Gland Thickness Gland Length Gland Thickness
Case No. (mm)
1 12.1
2 11.8
3 10.0
4 14.1
5 19.7
6 17.8
Range, dogs with hypoadrenocorticism 10.0–19.7
Median 13.1
Range, normal dogs8 (n=20) 13.2–26.3
Median 17.4
Discussion
Acute adrenal insufficiency is a life-threatening disease
known to be fatal if not treated immediately with inten-
sive therapeutic measures. A tentative diagnosis can be
made on the basis of a number of clinical signs and
laboratory findings, but a definitive diagnosis is only
possible with the help of adrenal function tests, such as
the ACTH stimulation test, that have an inherent time
delay until results are known. It thus would be useful if
the clinical diagnosis could be confirmed by means of an
imaging technique (e.g., ultrasonography), which can be
evaluated immediately.
It currently is believed that most cases of hypoadreno-
corticism result from an autoimmune destruction of the
adrenal cortices with bilateral atrophy of all three adre-
nal zones (Addison’s disease).17 This is visible in the
form of a bilateral, symmetrical reduction in size. In
humans, it has been possible to demonstrate this bilateral
atrophy with the help of computed tomography and to
utilize this finding in the differential diagnosis of
hypoadrenocorticism.18,19 In veterinary medicine, how-
ever, the diagnosis of hypoadrenocorticism in dogs using
an imaging technique has not been described.
Ultrasonography of the canine adrenal glands has be-
come increasingly important during the last few years.
Several studies have shown that ultrasonography is help-
ful in the diagnosis and differentiation of hyperadreno-
corticism in dogs.6,12,13 The purpose of the authors’ study
was to determine if it is possible to ultrasonographically
visualize and characterize the adrenal glands in dogs
suffering from hypoadrenocorticism, and to what extent
an ultrasonographic examination would be useful in di-
agnosing hypoadrenocorticism.
Ultrasonography in Dogs With Hypoadrenocorticism 217
Left Adrenal Right Adrenal Right Adrenal
(mm) (mm) (mm)
2.2 Not found Not found
2.2 12.3 2.8
2.4 Not found Not found
2.4 13.1 2.5
3.0 18.8 3.4
2.4 9.5 2.2
2.2–3.0 9.5–18.8 2.2–3.4
2.4
3.0–5.2 12.4–22.6 3.1–6.0
4.1 16.7 4.3
In the present study, the adrenal glands of six dogs,
with confirmed hypoadrenocorticism by laboratory meth-
ods, were examined ultrasonographically. The dogs were
of medium age; four of them were male, and two were
female. In four of the dogs, it was possible to precisely
visualize and measure both adrenal glands. These were
predominantly dogs of smaller breeds. By contrast, in
the cases of two large dogs (greater than 25 kg), it was
only possible to visualize the left adrenal gland. The
right adrenal gland is considerably more difficult to lo-
cate, particularly in narrow-chested dogs and dogs weigh-
ing more than 10 kg. This is due to intestinal-gas
interference and the adrenal gland’s anatomical position
within the costal arch. This, together with the fact that
atrophy of the adrenal glands is to be expected in dogs
with hypoadrenocorticism, explains why it was not pos-
sible to visualize the right adrenal gland in these two
dogs.
Ultrasonographic examination of the adrenal glands
of dogs with hypoadrenocorticism showed them to clearly
differ in size from the adrenal glands in dogs of a healthy
control group. The adrenal glands on both sides were
shorter (range: left adrenal gland length, 10.0 to 19.7
mm; median, 13.1 mm; range: right adrenal gland length,
9.5 to 18.8 mm) and thinner (range: left adrenal gland
thickness, 2.2 to 3.0 mm; median, 2.4 mm; range: right
adrenal gland thickness, 2.2 to 3.4 mm) than in healthy
dogs. It must be emphasized that the adrenal gland thick-
ness in five of six dogs with hypoadrenocorticism was
smaller than in any of the normal dogs. This result is
important, as the results of a previous study demon-
strated that there is no linear relationship between adre-
nal thickness and body weight in contrast to adrenal
218 JOURNAL of the American Animal Hospital Association
length and body weight,11 making adrenal thickness a
more consistent indicator of adrenal gland size. Statisti-
cal evaluation showed a significant reduction in length
and thickness of the left adrenal gland in dogs with
hypoadrenocorticism compared to left adrenal gland
length and thickness in normal dogs (p less than 0.05).
The change in shape is to be seen as the result of the
general reduction in size.
Comparing the results of the authors’ study with pre-
vious measurements of adrenal glands in normal dogs
made by other investigators,7,10 smaller measurements in
dogs with hypoadrenocorticism were found. Only
Douglass, et al.20 found a wider range of adrenal gland
size for dogs with no adrenal disease. In contrast to the
study by Grooters, et al.7 and Barthez, et al.,10 this study
did not take into account the possibility of previous
corticosteroid therapy and chronic nonadrenal illness
causing changes in adrenal gland size. Therefore, mea-
surements from this study may not reflect normal dog
reference values.
Conclusion
The results of this study show that atrophy of the adrenal
glands in dogs with hypoadrenocorticism seems to lead
to an ultrasonographically-measurable reduction in size
of the adrenal glands. Ultrasonographic examination of
the adrenal glands thus represents an extremely useful
additional technique for the diagnosis of hypoadreno-
corticism, and it is highly suitable as a screening method,
especially in emergency cases of acute Addison’s dis-
ease. To what extent continuous therapy with cortico-
steroids likewise brings about a measurable reduction in
size of the adrenal glands, which would necessitate dif-
ferential diagnostic consideration, has not yet been in-
vestigated and will constitute the subject of a further
study. In addition, it must be emphasized that due to the
reduction in size of the organ, it is more difficult to
visualize the adrenal glands in dogs with hypoadreno-
corticism than it is in healthy dogs. This emphasizes the
need for high-quality equipment and a high level of
experience on the part of the operator.
a
Synacthen; CIBA Pharma, Wehr, Germany
b
Sim 7000 CFM Challenge, Esaote Biomedica, Italy
c
SPSS for Windows, Version 6.0; SPSS, Inc., Chicago, IL
References
1. Schaer M, Riley WJ, Buergelt CD, et al. Autoimmunity and Addison’s
disease in the dog. J Am Anim Hosp Assoc 1986;22:789–93.
2. Feldman EC, Nelson RW. Hypoadrenocorticism. In: Feldman EC, Nelson
RW, eds. Canine and feline endocrinology and reproduction. Philadelphia:
WB Saunders, 1996:266–306.
3. Mulnix JA, Van Den Brom WE, Lubberink AAME, De Bruijne JJ,
Rijnberk A. Gamma camera imaging of bilateral adrenocortical tumors in
the dog. Am J Vet Res 1976;37:1467–71.
May/June 1999, Vol. 35
4. Voorhout G, Stolp R, Rijnberk A, Van Waes PF. Assessment of survey
radiography and comparison with x-ray computed tomography for
detection of hyperfunctioning adrenocortical tumors in dogs. J Am Vet
Med Assoc 1990;196:1799–803.
5. Voorhout G. X-ray-computed tomography, nephrotomography and
ultrasonography of the adrenal glands of healthy dogs. Am J Vet Res
1990;51:625–31.
6. Saunders HM, Pugh CR, Rhodes WH. Expanding applications of
abdominal ultrasonography. J Am Anim Hosp Assoc 1992;28:369–74.
7. Grooters AM, Biller DS, Merryman J. Ultrasonographic parameters of
normal canine adrenal glands: comparison to necropsy findings. Vet Radiol
1995;36:126–30.
8. Hoerauf A, Reusch C. Darstellung der Nebennieren mittels Ultraschall:
Untersuchungen bei gesunden Hunden, Hunden mit nicht-endokrinen
Erkrankungen sowie mit Cushing-Syndrom. Kleintierpraxis 1995;40:
351–60.
9. Kaser-Hotz B, Saunders HM. Sonographie der Nebennieren beim Hund.
Schweiz Arch Tierheilk 1995;137:258–64.
10. Barthez PY, Nyland TG, Feldman EC. Ultrasonographic evaluation of the
adrenal gland in normal dogs and dogs with pituitary-dependent
hyperadrenocorticism. Proceed, Am Coll Vet Int Med, 1994:160 (abstr.).
11. Barthez PY, Nyland TG, Feldman EC. Ultrasonographic evaluation of the
adrenal glands in dogs. J Am Vet Med Assoc 1995;207:1180–3.
12. Hoerauf A, Reusch C. Ultrasonographic evaluation of adrenal gland in
healthy dogs, dogs with no evidence of endocrine disease and dogs with
Cushing’s disease. Proceed, European Assoc Vet Diag Imaging, 1995:47
(abstr.).
13. Grooters AM, Biller DS, Theisen SK, Miyabayashi T. Ultrasonographic
characteristics of the adrenal glands in dogs with pituitary-dependent
hyperadrenocorticism: comparison with normal dogs. J Vet Int Med
1996;10:110–5.
14. Kantrowitz BM, Nyland TG, Feldman EC. Adrenal ultrasonography in the
dog. Detection of tumors and hyperplasia in hyperadrenocorticism. Vet
Radiol 1986;27:91–5.
15. Saunders HM. Adrenal disease. Proceed, 2nd Intern Symposium Vet
Echography, 1993:31–4.
16. Hoerauf A, Reusch C. Ultrasonographic evaluation of adrenal glands in
dogs with Cushing’s disease due to functional adrenal tumors and dogs
with Addison’s disease. Proceed, European Assoc Vet Diagn Imaging,
1996;abstr.:7–6.
17. Boujon CE, Bornand-Jaunin V, Schaerer V, Rossi GL, Bestetti GE.
Pituitary gland changes in canine hypoadrenocorticism: a functional and
immunocytochemical study. J Comp Path 1994;111:287–95.
18. Sun ZH, Nomura K, Toraya S, et al. Clinical significance of adrenal
computed tomography in Addison’s disease. Endocrinol Jpn 1992;39:
563–9.
19. Garcia Pascual L, Simo R, Mesa J, Gonzalez Atienza J, Vincente de Vera
P, Solduga C. Clinical usefulness of adrenal gland computed tomog-
raphy in the etiologic diagnosis of Addison’s disease. Med Clin (Barc)
1993;101:132–5.
20. Douglass JP, Berry CR, James S. Ultrasonographic adrenal gland
measurements in dogs without evidence of adrenal disease. Vet Radiol
1997;38:124–30.
Novartis
4C Ad
New

page 219
 Nonresponsive Generalized Bacterial
Infection Associated With Systemic
Lupus Erythematosus in a Beauceron
A case of concurrent canine systemic lupus erythematosus (SLE) and generalized bacterial
infection in a six-year-old female Beauceron is reported. The dog presented with purulent nasal
and ocular discharges, skin lesions (including seborrhea, hyperkeratotic areas, and papules as
well as ecchymoses around the eyes, on both sides of the pinnae, and on the vulva),
generalized lymph node enlargement, a mitral murmur, and lameness. Later, facial swelling, a
retrobulbar abscess, and a cough also developed. Occurrence of a generalized bacterial
infection was established by culture of group-C, β-hemolytic Streptococcus from the throat, the
mouth, a biopsy site (popliteal lymph node area), the retrobulbar abscess, and the lung.
The diagnosis of SLE was based on the clinical signs and particularly on the occurrence
of antinuclear antibody (ANA) and antidoublestranded-desoxyribonucleic acid (ds-DNA)
antibody. Interestingly, the latter type of antibodies were also detected in two young female
puppies whelped by this dog. Salient histological findings included an extreme cell depletion of
the lymph nodes and spleen and severe pneumonitis and peribronchiolitis.
The results of this case indicate that a definite diagnosis of canine SLE can, at times, be
made on the basis of the presence of serum ANA and ds-DNA antibodies.
J Am Anim Hosp Assoc 1999;35:220–8.

C. Clercx, DVM, Introduction

Diplomate ECVIM-CA Systemic lupus erythematosus (SLE) is a multiorgan disease character-
K. McEntee, DVM, ized by the simultaneous occurrence of a plethora of immunological
Diplomate ECVIM-CA abnormalities. Systemic lupus erythematosus has been reported in hu-
mans and many animals, including dogs and cats.1–5 The underlying
S. Gilbert, DVM pathogenesis is poorly understood and appears to involve genetic, envi-
ronmental, infectious, and hormonal factors as well as the immune sys-
L. Michiels, DVM
tem.2,6–8 The pathology in SLE is produced by autoimmune reactions
F. Snaps, DVM, DScV, Cert VR, against many tissue antigens and circulating immune complexes (CIC)
Diplomate ECDI involving antinuclear antibody (ANA).4,9,10 This leads to a wide range of
disease manifestations, all of which may exist alone or in combination,
E. Jacquinet, DVM making this disease exceedingly hard to diagnose. In this paper the
authors describe a case of SLE in a dog that demonstrated only the
D. Desmecht, DVM, AgES
minimum of criteria deemed necessary for a diagnosis of SLE.1–3 This
M. Henroteaux, DVM dog also had a concurrent end-stage generalized bacterial infection.

W.E. Bernadina, DVM

From the Departments of Small Animal Clinical Sciences (Clercx, McEntee, Gilbert,
C Michiels, Henroteaux), Diagnostic Imaging (Snaps), and Pathology (Jacquinet,
Desmecht), Faculty of Veterinary Medicine, University of Liège, 4000 Liège, Bel-
gium and the Department of Infectious Diseases and Immunology (Bernadina),
Faculty of Veterinary Medicine, University of Utrecht, Utrecht, The Netherlands.

Address requests for reprints to Dr. Clercx.

220 JOURNAL of the American Animal Hospital Association

Novartis
4C Ad
New

page 221
222 JOURNAL of the American Animal Hospital Association
Figure 1—View of the nares and planum nasale, showing severe
hyperkeratosis and nasal discharge in a six-year-old Beauceron
with systemic lupus erythematosus.
Case Report
A six-year-old, female Beauceron was referred to the
Small Animal Clinic of the Veterinary Faculty of the
University of Liège (SACVFUlg) for emaciation, dyspha-
gia, poor hair coat, and purulent discharges from both
eyes and nostrils. Previously, the dog had been referred
to the SACVFUlg at the age of two with a right, unilat-
eral, bloody nasal discharge, which had been diagnosed
as nasal aspergillosis (Aspergillus fumigatus). Intranasal
infusions with enilconazole had been applied. Regular
episodes of mucopurulent rhinitis (treatable with
antibiotic therapy) were noted ever since. At the age of
three, the dog was mated with a Beauceron, and two
females from the resulting litter (kept in the same house-
hold as their mother) were also available for this study.
Three months prior to referral to SACVFUlg, after
spending time in Greece, the dog manifested facial swell-
ing and a diffuse hemorrhagic diathesis with bleeding
from multiple skin and mucosal sites, which was suc-
cessfully treated with antibiotics.
At presentation, the dog was very thin (body weight,
34 kg), had a poor hair coat, purulent nasal and ocular
discharges, generalized seborrhea, hyperkeratotic lesions
[Figure 1], and papules and ecchymoses around the eyes,
on both sides of the pinnae, and on the vulva. The nails
were brittle. Physical examination revealed a general-
ized lymph node enlargement, a mitral murmur, and
signs of a chronic left anterior cruciate ligament rupture.
A suppurative skin lesion was observed at the tip of the
tail, which had developed three years earlier and had
never completely healed. Dysphagia was also noted when
the dog drank or was fed dry food.
Apart from demonstrating a slightly elevated erythro-
cyte sedimentation rate (2 mm/hr; reference range, 0 to 1
mm/hr) and reticulocytosis (4%), the results were all
within reference ranges. The direct Coombs’ test was
negative. A full serum biochemical profile was normal
except for a decreased α2 globulin (3.4 g/L; reference
May/June 1999, Vol. 35
range, 6.0 to 15.2 g/L). Serological tests for leishmania-
sis (immunofluorescent technique), babesiosis, toxoplas-
mosis, and Lyme disease were negative. Blood smears
from peripheral blood failed to detect any blood-borne
parasites. Additional samples of serum were taken and
stored frozen for use in immunoassays. Thoracic radio-
graphs showed nonsignificant bronchial calcification.
Electrocardiography and echocardiography were within
normal limits. Fine-needle aspirations were performed
on the left and right prescapular and popliteal lymph
nodes, and cytological examination revealed hypocellu-
larity. No fecal examination was performed. Urinalysis
revealed no abnormalities. Permission for the dog to be
hospitalized was denied, and the dog was sent home
without treatment.
On examination one month later, the dog’s general
condition had improved, and her body weight had in-
creased to 35.5 kg. Skin lesions and lymph node enlarge-
ments were less pronounced; however, the cardiac
murmur and dysphagia persisted. No abnormalities were
detected on neurological examination. Swallowing of a
barium contrast agent during fluoroscopy was normal.
No osteomyelitis was identified on radiographs of the
tail.
No definitive diagnosis could be made; therefore,
complete oral, pharyngeal, laryngeal, and intranasal ex-
aminations as well as tail amputation were performed
under general anesthesia. Severe stomatitis, glossitis,
and gingivitis were noted. Rhinoscopy revealed severe
atrophy of the right nasal cavity turbinates and the pres-
ence of exophytic scar tissue at the distal end of this
cavity. Radiography of the frontal sinuses was normal.
Serology for aspergillosis using double-diffusion on
agarose was positive, but no mycotic agent was cultured
from either the nose or the mouth. Group-C, β-hemolytic
Streptococcus and Staphylococcus aureus grew in
samples taken from the throat and mouth of the dog.
Mycobacterial cultures from the mouth and the left and
right prescapular and popliteal lymph node specimens
were negative; bacterial cultures from the lymph
nodes were also negative. Hematology was within
reference ranges, whereas blood chemistry revealed a
polyclonal increase in gammaglobulins despite a
slight hypoproteinemia.
Amoxicillin-clavulanatea (12.5 mg/kg body weight,
bid for 10 days) was given with a transient therapeutic
effect. Despite a good general condition and normal
appetite, intermittent bilateral hemorrhagic and purulent
nasal discharge, right forelimb lameness, and a cough
affected the dog within two weeks of antibiotic withdrawal.
The dog was again presented to the clinic due to the
sudden onset of swelling of the right eye. Facial swelling
was noted, with exopthalmia and lateral deviation of the
right ocular globe. Orthopedic findings were compatible
with a chronic left anterior cruciate ligament rupture.
Joint fluid aspiration and evaluation was recommended
Novartis
4C Ad
New

page 223
224 JOURNAL of the American Animal Hospital Association May/June 1999, Vol. 35

Figure 3A
Figure 2—Lateral view of the thoracic radiograph of the dog
from Figure 1, showing a diffuse, mixed, interstitial-alveolar
Figures 3A, 3B—Histology of a cervical lymph node of the dog
pattern with patchy consolidations in the diaphragmatic lobes
from Figure 1. (3A) Note the atrophy involving both B- and T-cell
and probable hilar lymph node enlargement.
regions. (3B) Note the multifocal histiocytic and eosinophilic
subacute lymphadenitis (Hematoxylin and eosin stain, 40X).

but declined. Auscultation findings were unremarkable.

Thoracic radiographs showed a moderate, generalized
pulmonary interstitial pattern; calcification of the right
shoulder was also identified. Opthalmological examina-
tion pointed to a probable retrobulbar mass as the under-
lying process.
Under anesthesia, retrobulbar aspiration and bronchos-
copy with bronchoalveolar lavage were carried out, and
biopsies were taken, including the whole left popliteal
lymph node and bone marrow. Bronchoscopy revealed
no macroscopic abnormalities. Histological examination
of the bone marrow was unremarkable. Histology of the
lymph node, however, revealed cellular depletion. The
retrobulbar mass was an abscess; Enterobacter and a
Figure 3B
group-C, β-hemolytic Streptococcus were grown in mi-
crobiological culture. Group-C, β-hemolytic Streptococ-
cus was also cultured from the bronchoalveolar lavage.
Culture on Sabouraud-chloramphenicol medium and my- cephalosporin and drainage of the popliteal abscess, no
cobacterial cultures using samples from the lymph node, improvement was noted in the dog’s condition. The
the lung, and the retrobulbar abscess were all negative, spleen became markedly enlarged. The owners requested
as were acid-fast stained smears of the same materials. euthanasia.
The amoxicillin-clavulanate treatmenta (12.5 mg/kg
body weight, bid for 10 days) was repeated. The face Immunological Test Results
rapidly returned to a normal appearance. However, the Several immunological laboratory investigations were
cough became harsh and frequent, and dyspnea ensued. performed on serum from this dog, using routine meth-
The biopsy area of the popliteal lymph node failed to ods. 11–14 Evaluation of the immune system is best
heal, and an abscess developed which yielded group-C, achieved through the evaluation of its component parts:
β-hemolytic Streptococcus in culture. Thoracic radio- humoral (B-lymphocytes, immunoglobulins), cellular
graphs were repeated and demonstrated a diffuse, mixed (T-lymphoctes, natural killer cells), complement, and
interstitial-alveolar pattern with patchy consolidations in the nonspecific phagocytic system.15 Humoral immunity
the diaphragmatic lobes [Figure 2]. Tracheobronchial was determined by the quantification of specific immu-
nodes were enlarged. Transthoracic fine-needle aspira- noglobulin classes in this dog’s serum; the patient had
tion biopsy induced hemoptysis; cytology of the aspirate normal immunoglobulin A (IgA), M (IgM), Gd (IgGd;
suggested an infectious process, and a group-C, β- an IgG subclass), and increased immunoglobulin Gab
hemolytic Streptococcus was cultured again. (IgGab; an IgG subclass) concentrations, indicating ad-
In spite of intensive supportive care, including treat- equate production. However, quantification of the hu-
ment with high doses of amoxicillin-clavulanate and moral response to a specific antigen (i.e., group-C
May/June 1999, Vol. 35
β-hemolytic Streptococcus) was not done. The authors
were not able to perform any T-cell function-specific
tests; however, as T-helper cells are required for the
production of immunoglobulins and concentrations were
normal in this dog, their presence and function were
considered to be adequate. The patient had normal serum
complement (C3b). Evaluation of neutrophil function
(nonspecific phagocytic system) includes adequacy of
numbers, chemotaxis, phagocytosis, and killing. 15 Al-
though neutrophil numbers were normal in this dog,
phagocytic function was severely depressed (neutrophil
phagocytic score, 10%; reference range, 75% to 100% in
healthy humans) [see Table]. Then the immune system
was evaluted for autoimmunity. The rheumatoid factor
(RF), ANA, antidoublestranded-desoxyribonucleic acid
(ds-DNA) antibody (Crithidia test), and circulating im-
mune complexes (CIC) tests were positive, the latter also
having ds-DNA antibody activity. Antinuclear antibody
and ds-DNA activities also were observed in both female
offspring; one also had slightly increased CIC. Based on
these results in this patient, a diagnosis of SLE was
made.
Necropsy and Histological Findings
The body was in very poor condition. A proliferative
ulcer (1 cm wide and 0.5 cm deep) at the tip of the tail
and another deep-penetrating ulcer (with a diameter of 2
cm and a depth of 0.7 cm) at the caudal side of the right
stifle were noted. The spleen, bronchial and cervical
lymph nodes, the lung, and the left atrium were enlarged.
Pathological diagnoses included splenic siderosclerosis,
mild pulmonary emphysema, chronic mitral endocardi-
tis, left-atrial subendocarditis and fibrosis (i.e., jet le-
sions), necrotizing chronic rhinitis, and mucopurulent,
acute, frontal sinusitis.
Histopathological examination of the cervical and
bronchial lymph nodes was characterized by atrophy of
both B- and T-cell dependent regions with subacute to
chronic, multifocal, histiocytic and eosinophilic lym-
phadenitis [Figure 3].
The stifle lesion was overlaid by an eosinophilic layer
of necrotic tissue covering a proliferative, homogeneous
tissue. The proliferative tissue was characterized as a
pleomorphic infiltrate of elongated inflammatory cells.
In the spleen, the white pulp was limited to small
follicles. The red pulp was a dense, eosinophilic tissue,
with few red blood cells (RBCs). Megakaryoblasts were
occasionally present, and there was evidence of hemo-
siderosis. Under the serosal capsule, there were large
foci of fibroblasts and golden-brown pigment containing
cells, both being evidence of siderosclerosis.
In the lung’s parenchyma, alveolar septae were thick-
ened by a cellular infiltrate and by numerous foci of
collagen and elastic fibers. Some alveoli were lined com-
pletely by type II pneumocytes. Inflammatory cells in
septae had round nuclei; they were mostly macrophages.
Bacterial Infection With Systemic Lupus Erythematosus 225
No exudate was present in the alveoli and bronchioli.
The tissue around the bronchi and bronchioli was com-
posed of irregular, fibrillar, and edematous collagen.
Diagnoses were interstitial, diffuse, subacute to chronic
pneumonitis; chronic peribronchiolitis; and chronic peri-
bronchitis.
Discussion
Major clinical differential diagnoses made initially in
this patient included disseminated aspergillosis and sys-
temic bacterial infection. Disseminated aspergillosis was
considered unlikely on the basis of the absolute failure to
detect the fungus, whether antemortem or postmortem,
in both fungal cultures and histological preparations.
Similarly, systemic bacterial disease arising solely from
an untreated or poorly treated infection of group-C β-
hemolytic Streptococcus, was not consistent with the
case history as fever was never detected and the adminis-
tration of systemic antibiotics did not prevent recurrence
of the infections or eliminate potential septic foci.
In the present study, T-cells were not examined in
vitro for capacity to proliferate in response to fungal and
bacterial antigens, or to modulate such responses. Both
specific and nonspecific IgG type (indicating isotype
switching) of humoral responses were detected in this
patient. This finding provides strong, albeit indirect, evi-
dence that there were no major aberrations in either the
inductive phase of antibody production (i.e., macro-
phage processing of antigens) or the productive phase,
including T- and B-cell activity. As antigen-specific an-
tibody is the best mediator of clearance of exogenous
bacteria, 16 it seems reasonable to assume that the pro-
gression or recurrence of infections was not due to im-
paired T- and B-cell function per se.
Immunological data revealed that the affected dog
had serological features of SLE. These included ANA
and ds-DNA as well as RF and CIC which have been
implicated in a wide spectrum of pathological lesions
and clinical manifestations.1,5,9,10 Three interrelated pro-
cesses are at work in SLE. First, there is a lack of
suppressor T-cell activity17 and T-lymphocyte dysfunc-
tion.18 Second, there is a consequential loss of control on
B-cell responses, resulting in polyclonal B-cell activa-
tion5 and production of autoantibodies to various self
antigens. The majority of self antigens are intracellular,
being mostly intranuclear10 but also intracytoplasmic in
location. 19 Lesser numbers of autoantibodies are directed
against surface antigens such as those on RBCs, neutro-
phils, platelets, or clotting factors. Third, loss of self
tolerance results in autoimmune disease and, in turn, the
autoantibodies initiating a wide spectrum of pathological
lesions and clinical manifestations. 1,5
Although pathogenesis of the clinical signs and patho-
logical findings observed in this case is uncertain, sev-
eral possible mechanisms deserve mention. Possibly, the
dog’s immune system may have been overwhelmed by
226 JOURNAL of the American Animal Hospital Association May/June 1999, Vol. 35

Table
Immunological Parameters Tested in a Six-Year-Old Female Beauceron,
Clinically Healthy Daughters, and Controls

Parameter Tested* Beauceron Daughter 1 Daughter 2 Controls†

Ig Levels mg/ml
IgA 3.3 2.1 2.6 3.5
IgM 3.6 2.6 2.8 3.5
IgG(ab) 19.2 10.9 8.7 15.0
IgG(d) 3.6 1.6 2.6 5.1
RF Positive Negative Negative Negative
ANA 1:200 1:50 1:50 <1:10
ds-DNA antibody 1:200 1:50 1:50 <1:10
CIC‡ 2.9 1.8 1.4 1.6
ds-DNA antibody/CIC Positive Negative Negative Negative
phag.act§ % 10% 81% 74% Not done
C3b level Normal Normal Normal Normal

* Ig=immunoglobulin; RF=rheumatoid factor; ANA=antinuclear antibody; ds-DNA=doublestranded deoxyribonucleic

acid; CIC=circulating immune complex; ds-DNA antibody/CIC=ds-DNA antibody activity detected in CIC;
phag.act=neutrophil phagocytic activity; C3b=serum complement
†
Detected in pooled serum from 10 randomly chosen dogs presented at the Utrecht Veterinary Clinic
‡
Protein content of precipitated CIC in mg/ml serum
§
Expressed as percentage of neutrophils that had phagocytosed three or more fluorescent E. coli bacteria (Phagotest
Becton Dickinson; reference range, 75% to 100% in healthy humans)

antigens at too high of a level. The lack of suppressor T-
cell activity may have produced the pathology observed.
The resulting sustained production of antibody to both
invading antigens and autoantigens creates a condition
of continuous usage of B-cells and constant demand for
phagocytic clearance potential. Neutrophils from the pa-
tient were shown to be markedly deficient, compared to
neutrophils from controls, in the capacity to phagocyte
opsonized Escherichia coli-fluorescein isothiocyanate
(FITC) bacteria. However, phagocytic tests carried out
during infection episodes, as were done in this case, may
be difficult to interpret as active infection and the use of
antibiotics may alter test results. Since neutrophil phago-
cytic activity was not measured during infection-free
periods, it is possible that persistent loading of neutro-
phils with CIC produced during infection resulted in the
low level of phagocytic activity observed in this case.
Also, the persistent infection may cause T- and B-cells
to become sequestered when they confront these large
concentrations of antigens. A similar mechanism, called
negative selection, has been shown to be operative in in
vivo experiments.20 Ensuing B-cell depletion (negative
selection) and depressed neutrophil phagocytic activity
(due to overloading) cause even more antigens to remain
in the system, with further aggravation of cell depletion
and phagocytic activity. Additional freeing of autoanti-
gens, such as DNA (cell death), could have aggravated
B-lymphocyte depletion of the spleen and lymph nodes
to produce the postmortem finding (i.e., remarkable
depletion of the reticuloendothelial system). Exhaustion
of T- and B-cell areas in the spleen and lymph nodes may
especially have been hastened by events during the dog’s
last phases of life. Indeed, whether the infection ob-
served in this patient followed SLE-induced, B-cell
depletion or whether it was pre-existing but hastened by
B-lymphocyte depletion remains uncertain.
The initiating cause of the autoimmune response
which is the fundamental triggering mechanism in SLE
is still poorly understood. Genetic factors are essential,
as shown by numerous studies conducted in humans as
well as in dogs.6,21 The SLE disease process could be
triggered by diverse stimuli, including infection7 and
endocrine and environmental factors (e.g., ultraviolet
radiation),2,5 all of which have been implicated clini-
cally. In this study, the familial aspect of the disease is
illustrated by the finding of ds-DNA in two young fe-
male offspring. Also, hormones may have influenced
development of disease, as signs occurred following
pregnancy. Interestingly, the animal’s first suspect epi-
sode of cutaneous vasculitis occurred after returning
home from a visit to sunny Greece, thus emphasizing a
possible aggravation of the lesions with exposure to
May/June 1999, Vol. 35
sunlight. On the other hand, a primary infectious process
could also have been the triggering mechanism causing
exacerbation of signs. Two infectious processes may
have been involved in the authors’ case study. First, the
skin lesion at the tip of the tail could have been a chronic
site of free entrance for pathogens for many months. The
other possible initiating antigen of infectious origin may
have been Aspergillus fumigatus, even though fungal
infection occurred four years previously. Indeed, al-
though SLE affects middle-aged dogs (between two and
12 years of age; mean age, five years), the disease most
probably precedes the clinical diagnosis by many months
or even years.3 Available data about breed predilection is
inconsistent; although poodles and German shepherd
dogs appear over-represented,1–3,18,22,23 the Beauceron is
usually absent from the breeds quoted. Overall analysis
of data shows no pronounced sex predisposition.1–3,5,22–24
A question still needing to be addressed is whether
the pathology observed in the described case can be
attributed to a SLE-related phenomena. The pathology in
dogs usually is manifested in the form of so-called major
and minor signs.2,9 This case had no clear major features
characteristic of SLE. Arthritis is, overall, the major
presenting sign.9 Immune-mediated arthritis was pos-
sible since lameness was shifting, but this had not been
clinically confirmed. Although the appearance of lupus
cutaneous lesions is highly variable,2,23 the presence of
erosions, ulcers, crusts, and scales involving the face
was compatible with SLE and could not be ascribed to
any other specific disease. The observed reticulocytosis
suggested a previous anemia; petechia, ecchymoses, and
spontaneous bleedings were recorded in the history. The
previous occurrence of a hemolytic process was also
suggested by the presence of marked siderosclerosis,
although it can also be associated with systemic infec-
tion by a hemolytic bacteria.
In the present study, a diagnosis of SLE was made on
the basis of serological findings. In canine SLE, ANA
titers obtained by indirect immunofluorescence technique
are high (higher than 100).24 According to some authors,
the highest titers are associated with more severe clinical
signs, and, conversely, ANA titers decrease with appro-
priate treatment and tend to increase again when recur-
rence occurs.3 Why no exceptionally high ANA titer was
detected in the patient of this case report is uncertain, but
one is tempted to think of the ongoing B-cell depletion as
a possible explanation. Nevertheless, in the patient stud-
ied here, the definite diagnosis of SLE was made on the
strong presence of ds-DNA antibody detectable at rela-
tively high titers, a phenomenon not ascribed to any
other disease. The Crithidia test for ds-DNA antibody
represents a sensitive and highly specific substrate for
the detection of SLE-associated ANA in humans.12 In
humans, the presence of ds-DNA antibody is a hallmark
of SLE, since they are found in 80% of the cases. The
Crithidia test’s use in dogs is controversial, since ds-
Bacterial Infection With Systemic Lupus Erythematosus 227
DNA antibodies are rarely found in canine SLE. 3,18 How-
ever, when positive, this test could represent an accurate
diagnostic element for canine SLE.
Conclusion
On the basis of findings produced in this study, a diagno-
sis of end-stage generalized bacterial infection occurring
in association with SLE was made in a six-year-old
female Beauceron. It appears that, because of underlying
SLE, the patient was strongly biased toward the induc-
tion of an inadequate immune response. This diagnosis
could not be made on clinical findings alone, but re-
quired serological data. Among these, a Crithidia ds-
DNA antibody test, when positive, may represent an
accurate diagnostic element for canine SLE; however, its
utility, in this regard, requires further investigation.
a
Synulox; SmithKline Beecham, Wavre, Belgium
Acknowledgments
The authors thank Dr. Nicole Lafontaine, from the Labo-
ratory of Immunology of the “Centre Hospitalier
Universitaire” of the University of Liége, Belgium, for
the phagocytic test.
References
1. Grindem CB, Johnson KH. Systemic lupus erythematosus. Literature
review and report of 42 new canine cases. J Am Anim Hosp Assoc
1982;19:489–503.
2. Halliwell REW. Autoimmune blood diseases. In: Halliwell REW, Gorman
NT, eds. Veterinary clinical immunology. Philadelphia: WB Saunders,
1989:324–36.
3. Fournel C, Chabanne L, Caux C, et al. Canine systemic lupus erythemato-
sus. I. A study of 75 cases. Lupus 1992;1:133–9.
4. Jones DRE. Canine systemic lupus erythematosus: new insights and their
implications. J Comp Path 1993;108:215–28.
5. Tizard IR. The systemic imunological diseases (chapter 33). In: Tizard IR,
ed. Veterinary immunology. An introduction. 5th ed. Philadelphia: WB
Saunders, 1996:426–31.
6. Lewis RM, Schwartz RS. Canine lupus erythematosus. Genetic analysis of
an established breeding colony. J Exp Med 1971;136:417.
7. Lewis RM, Andre-Schwartz J, Harris GS, et al. Canine systemic lupus
erythematosus. Transmission of serologic abnormalities by cell-free
filtrates. J Clin Invest 1973;52:1893–907.
8. Hubert B, Teichner M, Fournel C, et al. Spontaneous familial systemic lupus
erythematosus in a canine breeding colony. J Comp Path 1988;98:81–9.
9. Bennett D. Immune-based non-erosive inflammatory joint disease of the
dog. 1. Canine sytemic lupus erythematosus. J Sm Anim Pract 1987;
28:871–89.
10. Monier JC, Ritter J, Caux C, et al. Canine systemic lupus erythematosus. II.
Antinuclear antibodies. Lupus 1992;1:287–93.
11. Wood DD, Hurvitz AI, Schultz RD. A latex test for canine rheumatoid
factor. Vet Immunol Immunopathol 1980;1:103–11.
12. Aarden LA, de Groot ER, Feltkamp TEW. Immunology of DNA. III.
Crithidia Lucilliae, a simple substrate for the determination of anti-dsDNA
with the immunofluorescence technique. Ann NY Acad Sci 1975;254:
505–14.
13. Breedveld FC, Heurkens AHM, Lafeber GJM, et al. Immune complexes in
sera of patients with rheumatoid vasculitis induced polymorphonuclear cell
mediated injury to endothelial cells. Clin Immunol Immunopathol
1988;48:202–8.
(Continued on next page)
228 JOURNAL of the American Animal Hospital Association May/June 1999, Vol. 35

References (cont’d) 19. Monestier M, Novick KE, Karam ET, et al. Autoantibodies to histone,
DNA, and nucleosome antigens in canine systemic lupus erythematosus.
14. Wisselink MA, Bernadina WE, Willemse A, et al. Immunologic aspects of Clin Exp Immunol 1995;99:37–41.
German shepherd dog pyoderma (GSP). Vet Immunol Immunopathol
1988;19:67–77. 20. Bellgrau D, Talmage DW. T cells can be cytotoxic without making IL-2; a
model of separate pathways of induction. Proc Natl Acad Sci USA
15. Greshwin LJ. Immunological assessment of the small animal patient. In: 1986;83:3412–6.
Kirk RW, Bonagura JD, eds. Current veterinary therapy, small animal
practice XI. Philadelphia: WB Saunders, 1992:441–8. 21. Monier JC, Fournel C, Lapras M, et al. Systemic lupus erythematosus in a
colony of dogs. Am J Vet Res 1988;49:46–51.
16. Felsburg PJ. Primary immunodeficiencies. In: Kirk RW, Bonagura JD, eds.
Current veterinary therapy, small animal practice XI. Philadelphia: WB 22. Drazner FH. Systemic lupus erythematosus in the dog. Comp Cont Ed
Saunders, 1992:448–53. 1980;2:243–54.

17. Abdou NI, Sagawa A, Pascual E, et al. Suppressor T cell abnormality in
idiopathic systemic lupus erythematosus. Clin Immunol Immunopathol
1976;6:192–9.
18. Monier JC, Dardenne M, Rigan D, et al. Clinical and laboratory features of
canine lupus syndromes. Arthritides Rheum 1980;23:294.
23. Scott DW, Walton DK, Manning TO, et al. Canine lupus erythematosus. I.
Systemic lupus erythematosus. J Am Anim Hosp Assoc 1983;19:461–79.
24. Kass PH, Strombeck DR, Farver TB, et al. Application of the log-linear
model in the prediction of the antinuclear antibody test in the dog. Am J
Vet Res 1985;46:2336–9.
 Reflex Sympathetic Dystrophy in a Dog
Reflex sympathetic dystrophy is a well-recognized syndrome in human patients following injury
to an extremity. The syndrome may include hyperesthesia and autonomic changes. The
autonomic changes are initial vasodilatation followed by vasoconstriction (e.g., edema followed
by cyanosis, and cool skin); hyper- or hypohydrosis; atrophic changes in the skin, subcutis, and
muscles; and osteoporosis. Early treatment with a short course of steroids and infiltration of the
painful site with lidocaine may alleviate symptoms. If that fails, sympathetic ganglionic block
with lidocaine (and possibly steroids) or surgical sympathectomy may provide resolution. A
case of reflex sympathetic dystrophy in a dog is presented, involving bilateral distal hind-limb
edema and hyperesthesia. J Am Anim Hosp Assoc 1999;35:229–31.

An LaBarre, DVM, MS Case Report

Bonnie E. Coyne, DVM, MS,
Diplomate ACVS
C noted to be flaccid and painful. At drain removal four days later, the tail
From the Companion Veterinary Clinic,
105 Auburn Folsom Road,
Auburn, California 95603.
A young, adult, male neutered Labrador retriever mixed-breed was pre-
sented for evaluation of bilateral hind-limb edema. The previous year the
dog had been found with puncture wounds of the left medial thigh, which
was very swollen. He was treated with surgical debridement, drainage,
and Keflex (500 mg tid, duration unknown). The next day, the tail was
was still flaccid and painful, and a lump was noted at the tail head. At that
time, bilateral swelling of the hind limbs was noted, and the dog was
treated with prednisone (complete dosing information unavailable). The
owner stated that the dog’s edema had never resolved. The owner as-
sumed that the dog had had pelvic fractures, but there had been no
radiographs taken to define the bony injuries.
On presentation, the dog had a fever of 104.2˚ F and severe bilateral
edema of the hind limbs distal to the hocks. There were fluctuant,
3- to 4-cm pockets anterior to the hocks and pitting edema distally. Pain
was elicited upon manipulation of the tail, and the dog did not voluntarily
lift it above horizontal plane. Examination of the head, eyes, ears, nose,
and throat; auscultation of the chest; and palpation of the abdomen were
normal.
Results of a serum biochemistry panel were normal. Aside from 3+
ammonium magnesium phosphate crystals, a urinalysis was normal.
Results of the hemogram included a leukocytosis of 24.7 x103/µl (refer-
ence range, 6.0 to 17.0 x103/µl) with an absolute neutrophil count of 22.9
x103/µl (reference range, 3.0 to 11.5 x103/µl). Fluid aspiration of a visibly
dilated, fluctuant lymph cistern anterior to the hock on the right hind limb
revealed approximately 30 cc of a clear, yellow fluid with a specific
gravity of 1.017, a protein of 2.3 g/dl, a red blood cell count of 81/µl, and
nucleated cells of 75/µl consisting of a low number of heterogenous
lymphocytes that were mostly medium sized, occasionally granular, and
rarely reactive. This was consistent with lymphedema. There was no
evidence of inflammation or infection. A bacterial culture was negative.
Lumbar spinal radiographs showed no evidence of previous pelvic
fractures, but they did demonstrate mature callus formation and deformity
between the first and second coccygeal vertebrae (suspect as an old
fracture site) and slight indistinctness of the articular facets at the lum-
bosacral junction. No urinary calculi were seen.
The dog was given an injection of amoxicillina (5 mg/lb body weight)
and continued on the same dose orallyb bid for seven days. By the next

JOURNAL of the American Animal Hospital Association 229

230 JOURNAL of the American Animal Hospital Association
morning, body temperature was normal but the edema
persisted with only slight improvement. He had eaten
and urinated but had not passed any stool. A large vol-
ume of stool had been seen on the previous abdominal
radiographs. Rectal examination did not reveal any ob-
struction or lesions internally other than the coccygeal
lesion which was palpable dorsally. The dog exhibited
severe pain if the tail was lifted above the horizontal
plane. It was speculated that the large volume of stool
within the colon might be increasing pressure on internal
lymphatics, resulting in the hind-limb edema. A dioctyl
sodium succinate (DSS) enemac was administered, and
the dog was started on furosemided (1 mg/lb tid).
Further diagnostics including abdominal ultrasound,
intravenous pyelogram, barium enema, and lymphogram,
to define pelvic and lymphatic structure and function so
as to rule out obstructive masses or other causes of
edema, were discussed but declined by the owner. Am-
putation of the tail to reduce leverage on the previously
identified, suspect fracture site and possibly reduce pe-
ripheral nerve compression was agreed upon, pending
the dog’s response to continued medical treatment at
home.
The dog did not improve at home on furosemide and
amoxicillin therapy. Surgery to amputate the tail at the
body-tail interface was performed. Bilateral chemical
sympathetic block of the distal two sympathetic ganglia1
was done by placing a 3.5-inch spinal needle approxi-
mately 1.5 inches lateral to the dorsal spinus process,
just anterior to the transverse processes of the second
and fourth lumbar vertebrae, deep to the psoas muscle
and injecting lidocainee (6 cc, 1%) into each area. Toe
web temperature was monitored. An immediate increase
in temperature of the extremities is possible in human
patients with reflex sympathetic dystrophy following
successful block due to immediate vasodilatation. This
was not observed in this dog. However, the edema
gradually resolved approximately 95% within the next
24 hours. After several days, the edema recurred.
Vasodilator therapy with enalapril maleate f (10 mg bid),
used in an attempt to reverse the suspect autonomic
vasoconstriction, had little effect. Combined therapy with
enalapril maleate (10 mg bid) and furosemide (50 mg
bid) had a greater effect. The combination of drug therapy
and physical therapy with retrograde massage controlled
the edema to the owner’s satisfaction. The size of the
cisterna was reduced 25% to 75%, depending on the time
of day and the dog’s activity level (the owner stated the
edema was worse in the evening after a day’s activity).
The owner declined injections of absolute alcohol or
phenol for ganglionic block and surgical sympathec-
tomy. The most effective control was later achieved with
phenoxybenzamine HClg (10 mg bid). Edema was re-
duced 75%.
May/June 1999, Vol. 35
Discussion
Reflex sympathetic dystrophy is a well-recognized
syndrome in human patients following injury to an
extremity.2 It is a sequel in 5% of trauma cases. Fifty
percent of cases are postfracture, but the condition may
also follow blunt trauma, inflammation, laceration,
surgery, injection, angina, vascular disease, myocardial
infarction, stroke, cervical arthritis, degenerative joint
disease, frostbite, or burns. It is remarkable, even
frightening, that symptoms may occur after almost any
injury, anywhere in the distribution of the nerve, even
after injuries that are as seemingly minor as a cut on the
finger. Acute symptoms in human patients are aching or
burning sensations which persist after healing in the area
of the injured nerve’s distribution. Initially the skin is
red and dry, later becoming cold, cyanotic, and sweaty.
Edema, hypertrichosis, and overgrowth of nails may also
develop. Chronic symptoms may include irreversible
atrophy of the skin (which becomes smooth and
glossy), subcutis, and muscles (tapering of digits
and contractures), and degeneration of the joints
(i.e., ankylosis) and bones (i.e., osteoporosis).3 Patho-
physiology of this syndrome in people is not completely
understood, but inappropriate reflex sympathetic tone
probably causes hyperesthesia, venous constriction, and
edema. The atrophy of the skin, subcutis muscle, and
bone may also be caused by altered nerve trophism.
Lymph vessels are not known to have any neurological
modification in function. Treatments in human patients
include physical therapy, injection of the appropriate
area with local anesthetics (this is most effective early in
the course of the disability), intravenous regional sympa-
thetic block (injection of guanethidine or reserpine into
the temporarily occluded vessel of the area), chemical
sympathectomy (injecting phenol or absolute alcohol),
surgical sympathectomy, or oral therapy (using phenoxy-
benzamine or propanolol) 4 to block sympathetic signals.
Edema has been associated with hypoalbuminemia or
renal, cardiac, or hepatic dysfunction. These causes were
ruled out by the physical examination and results of
blood chemistries in this case. Inflammation and infec-
tion were ruled out by the fluid analysis and culture. No
physical cause for vascular or lymphatic obstruction was
documented, although a complete investigation was de-
clined by the owner. Lymphogram and/or venogram
might define or distinguish these possible causes of
edema. The spinal trauma itself would not cause lym-
phatic obstruction. The history of spinal trauma and
persistent bilateral edema in a dog whose initial soft-
tissue injuries were unilateral led the authors to suspect
autonomic dysfunction as a cause of functional obstruc-
tion in this dog. Resolution of the edema after sympa-
thetic block with lidocaine was supportive of the
diagnosis. It is probable that more permanent ganglionic
May/June 1999, Vol. 35 Sympathetic Dystrophy 231

block with phenol or absolute alcohol or surgical sympa- Acknowledgments

thectomy would decrease or resolve the edema. Clini- The authors would like to acknowledge the assistance of
cians should be aware of reflex sympathetic dystrophy as Dinah C. Gibbs, RPT; Christine Bonacci, MD; Jeffery
a differential for edema following limb or nerve injury, Reinking, MD; James Willis, MD; and Candy Stafford,
especially when all other etiologies have been ruled out. RVT.

a
Amoxicillin injectable; GC Hanford Mfg Co, Syracuse, NY
b
Amoxicillin; SmithKlein Beecham, Philadelphia, PA
References
c 1. Evans HE. The automatic nervous system. In: Millers’ anatomy of the dog.
Diocytl Sodium Succinate 250 mg in glycerine; Vedco Inc., St Joseph, MO
d 3rd ed. Philadelphia: WB Saunders, 1993:776–9.
Furosemide; Hoechst Group, Warren, NJ
e 2. Posner JB. Disorders of sensation. In: Wyngaarden JB, Smith LH, eds.
Lidocaine 1%; Schein Pharmaceutical Inc., Florham Park, NJ Cecil textbook of medicine. Philadelphia: WB Saunders, 1988:2128–37.
f
Enalapril maleate; Merck & Co Inc., Rahway, NJ 3. Tasker RR, Dostrovsky JO. Deafferentation and central pain. In: Wall PD,
g Melzack R, eds. Textbook of pain. 2nd ed. New York: Churchill
Phenoxybenzamine HCl; SmithKlein Beecham, Philadelphia, PA
Livingstone, 1989:163.
4. Bonica JJ. Causalgia and other reflex sympathetic dystrophies. In: Bonica
JJ, Loeser JD, Chapman CR, Fordyce WE, eds. The management of pain.
2nd ed. Philadelphia: Lea & Febiger, 1990:220–43.
 Sensory Polyganglioradiculoneuritis
in a Dog
Generalized reduction of nociception and conscious and unconscious proprioception were
found in an approximately eight-year-old, male, Maltese mixed-breed dog presented for
difficulty prehending food and experiencing ataxia of three months duration. Results of needle
electromyogram, motor nerve conduction velocity, and cerebrospinal fluid analysis were
normal. A diagnosis of sensory polyneuropathy was suspected. No underlying cause could be
determined. Neurological signs progressed to quadriparesis over the following four months
despite treatment attempts with prednisone and procarbazine. Necropsy confirmed a sensory
polyganglioradiculoneuritis, but no inciting cause could be established.
J Am Anim Hosp Assoc 1999;35:232–5.

Cheryl L. Chrisman, DVM, MS, EdS, Case Report

Diplomate ACVIM
Simon R. Platt, BVM&S, MRCVS
A.M.S. Chandra, BVSc, PhD
Alexander deLahunta, DVM, PhD,
Diplomate ACVIM
G. Diane Shelton, DVM, PhD,
Diplomate ACVIM
C Five months prior to presentation, the dog acutely developed an unpro-
From the Department of Small Animal
Clinical Sciences (Chrisman, Platt),
P.O. Box 100126,
and the Department of Pathobiology
(Chandra),
P.O. Box 100144,
College of Veterinary Medicine,
The University of Florida,
Gainesville, Florida 32610;
the Department of Anatomy (deLahunta),
College of Veterinary Medicine,
Cornell University,
Ithaca, New York 14853;
and the Comparative Neuromuscular
Laboratory (Shelton),
University of California-San Diego,
LaJolla, California 92093.
An approximately eight-year-old, neutered male, Maltese mixed-breed
dog was presented for difficulty prehending food and a progressive,
generalized ataxia of three months duration. The ataxia had improved
slightly on a previous one-month course of prednisolone (1 mg/kg body
weight, per day, per os [PO]) therapy, but the dog was on no medication at
this time. The dog had been found as a stray six years previously and had
been in good health except for chronic flea allergy dermatitis and peri-
odontal disease. Annual vaccinations (given six months prior to presenta-
tion) and frequent dentistry had been performed. Flea control was
maintained with pyrethrins. No heartworm medication had been given,
but he was negative for heartworm antibody.
ductive cough and ocular discharge two days after teeth cleaning and two
weeks after being suspected of inhaling particles from a malfunctioning
air conditioner. The particles were later analyzed and found to be silica
gel. Results of a complete blood count (CBC), serum biochemistry pro-
file, and tracheal wash cytology and culture were normal at that time.
Thoracic radiographs showed a mild, right caudal lung-lobe pneumonia.
Keratoconjunctivitis sicca was diagnosed, as the Schirmer’s tear test was
less than 4 mm in each eye. Cyclosporin 0.2% ophthalmic ointmenta was
administered in each eye daily. The cough continued even after the
pneumonia resolved radiographically.
Three months prior to presentation, the dog developed dysmetria of all
four limbs, worse in the pelvic limbs. Bilateral conscious proprioceptive
deficits of the pelvic limbs also were present. Spinal reflexes of all four
limbs were normal. A diffuse encephalomyelitis was suspected, and se-
rum titers for canine distemper virus, canine herpes virus, Toxoplasma
gondii, Erhlichia canis, Rickettsia rickettsii, Borrelia burgdorferi, and
Babesia canis were evaluated and were negative except for a distemper
virus immunoglobulin G (IgG) of 1:25, considered to be due to vaccina-
tion. An antinuclear antibody test was negative.
On presentation, the dog weighed 3.55 kg and had a mild generalized
erythematous dermatitis, right otitis externa, and bilateral conjunctivitis.
A cough was easy to elicit on tracheal manipulation, but there were no

232 JOURNAL of the American Animal Hospital Association

May/June 1999, Vol. 35
increased lung sounds. Body temperature was 102˚ F.
The dog was mentally alert and responsive, but had
reduced facial nociception, evidenced by little response
to needle pricking of nasal mucosa, ears, lips, and tongue.
Jaw tone and swallowing appeared normal. Difficulty
prehending food was observed and seemed to be due to a
lack of touch sensation and proprioception of the lips
and tongue. Smell and taste seemed appropriate, as the
dog could find the food and would eat well once food
was placed in his mouth. The tongue often protruded
beyond the lips. The palpebral reflex was reduced bilat-
erally, although the dog could close his eyelids. Vision,
hearing, and normal vestibular nystagmus were present.
Other cranial nerves were normal.
The dog had extreme generalized dysmetria, worse in
the pelvic limbs than the thoracic limbs. Conscious
prorioception was absent in all four limbs. Limb strength
was normal. The patellar reflex was absent on the left
and severely depressed on the right. The sciatic notch
response was depressed bilaterally, but cranial tibial,
gastrocnemius, and flexor reflexes of the pelvic and
thoracic limbs were normal. A crossed extensor reflex
was present in all four limbs. Anal tone, defecation, and
urination were normal. Although the cutaneous trunci
response was present, little behavioral response was ob-
served to pin pricking of the skin along the dorsal spinous
processes of the vertebral column, suggesting reduced
superficial nociception. Deep nociception, tested by the
application of a hemostat forcep to the digits of all four
feet, also was reduced. A diffuse disorder affecting sen-
sory nerves carrying nociception and unconscious and
conscious proprioception of the head, body, and limbs
was suspected. Sensory nerves carrying the special senses
of olfaction, vision, audition, and equilibrium appeared
to be spared. A sensory polyneuropathy due to a meta-
bolic disorder, endocrinopathy, toxicity, or autoimmune
disease was suspected.
Results of a CBC, urinalysis, and serum biochemistry
profile and serum cholinesterase were normal, with the
exception of a serum alkaline phosphatase of 292 U/L
(reference range, 8 to 56 U/L) and an alanine aminotrans-
ferase of 323.2 U/L (reference range, 15 to 58 U/L),
thought to be due to the previous prednisone therapy. A
low-dose dexamethasone suppression test was normal.
Precortisol was 1.03 µg/dl (reference range, 0.5 to 6.0
µg/dl), and post low-dose dexamethasone (0.01 mg/kg
body weight, intravenously [IV]) was 1.09 µg/dl at four
hours and 0.57 µg/dl at eight hours (reference range, less
than 1.2 µg/dl). A basal thyroxine (T4) level was slightly
low at 0.77 µg/dl (reference range, 0.9 to 2.6 µg/dl), but the
thyroid stimulating hormone level was normal at 0.15
ng/ml (reference range, 0.03 to 0.39 ng/ml), so hypothy-
roidism was considered unlikely. A small liver and
bronchointerstitial lung disease were seen on abdominal
and thoracic radiographs, respectively. No evidence of
megaesophagus or neoplasia was seen. Skull and cervi-
Polyganglioradiculoneuritis 233
cal radiographs, performed under general anesthesia,
were normal, except for a slight thickening and opacity
in the left osseous bulla and lucency around the roots of
several teeth.
No fibrillation potentials, positive sharp waves, or
other abnormalities were found on electromyographic
(EMG) examination of muscles of the head, body, and
limbs. A sciatic tibial motor nerve conduction velocity
was 59 meters per second (reference range, 55 to
75 m/sec). The evoked M responses were 5.83 and 6.75
millivolts, biphasic and considered to be normal. No
decremental response to repetitive stimulation was ob-
served. These EMG findings were indicative of normal
motor nerve function. Sensory nerve stimulation was
attempted but was inconclusive due to technical difficul-
ties. An H-reflex was not performed. Analysis of cere-
brospinal fluid (CSF) taken from the cerebellomedullary
cistern indicated no red blood cells (RBCs), three white
blood cells (WBC/µl; reference range, 0 to 6 WBC/µl),
and protein less than 20 mg/dl (reference range, less than
20 mg/dl). Cytological evaluation of 50 CSF WBCs
showed that 13 were small lymphocytes and 37 were
reactive mononuclear phagocytes, indicating nerve root
or central nervous system (CNS) disease. Serum
antineuronal nuclear antibody type I and calcium chan-
nel antibodies were negative.
A diagnosis of sensory polyneuropathy of presumed
autoimmune or toxic etiology was made. Since no toxic-
ity could be established, the dog was treated for sus-
pected autoimmune disease with oral prednisone
(3 mg/kg body weight per day). Minimal improvement
was seen after 23 days of treatment. Another immuno-
suppressive drug, procarbazineb (50 mg/mm2 per day),
was added to the prednisone. The dog was maintained on
the multiple vitamin and mineral supplements initiated
prior to presentation. A CBC was performed every two
weeks to monitor for leukopenia and anemia, which, if
present, would signify bone-marrow suppression from
the procarbazine. Procarbazine was discontinued after
three weeks when no improvement was seen. The cough
had increased, and pneumonia was suspected, but tho-
racic radiographs were normal. Oral amoxicillinc (15
mg/kg body weight, q 12 hrs) and enrofloxacin d
(5 mg/kg body weight, q 12 hrs) were given, for 14 days
each, in succession, but the cough continued.
Neurological deficits progressed over the next four
months, despite continued oral prednisone (3 mg/kg body
weight, per day). The dog developed severe depression
and dyspnea and was reevaluated. Proprioceptive and
nociceptive deficits remained, but now the dog had little
voluntary movement of the limbs. Previously normal
spinal reflexes were now depressed or absent. The left
pupil was miotic. Aspiration pneumonia was suspected
based on thoracic radiographs, and the peripheral WBC
count of 53.70 x10 3/µl (reference range, 6 to 17
WBC/µl) with 44.6 x103/µl neutrophils (reference range,
234 JOURNAL of the American Animal Hospital Association
2.5 to 12.5 x103/µl) and 3.8 x103/µl bands (reference
range, 0 to 0.3 x103/µl). Hepatomegaly was seen on
abdominal radiographs. The serum alkaline phosphatase
was 2,500 U/L (reference range, less than 80 U/L), and
alanine aminotransferase was 167.1 IU (reference range,
less than 60 IU). Due to the progression and severity of
signs, the dog was euthanized.
A complete necropsy was done within a few hours
after euthanasia. At necropsy, the cranioventral lung
lobes were consolidated and mottled tan and red. The
liver was enlarged with an accentuated reticular pattern.
The CNS was removed and fixed by immersion in 10%
neutral buffered formalin. Selected peripheral nerves,
skeletal muscles, and ganglia, as well as other organs
were collected in 10% formalin. All tissue samples were
embedded in paraffin, cut into 5-µm thick sections, and
stained with hematoxylin and eosin (H&E) stain. Sec-
tions of the CNS also were stained with luxol fast blue
(LFB) and Bodian’s method for axons. Additionally,
gluteraldehyde-fixed samples of the sciatic nerve were
postfixed in osmium tetroxide and dehydrated in serial
alcohol solutions and propylene oxide prior to embed-
ding in araldite resin. Sections (1 µm) were stained with
toluidine blue for light microscopy. A sample of the
biceps femoris muscle was flash frozen in isopentane,
precooled in liquid nitrogen, and 8-µm thick sections
were evaluated with standardized histochemical stains
and enzyme reactions.
Histological examination of the lung tissue revealed
suppurative bronchopneumonia compatible with aspira-
tion. Under polarized light, sections of lung revealed a
few birefrigent crystals in areas of inflammation that
were interpreted as silica particles. Nodular collagenous
scars or granulomas, suggestive of chronic silicosis, were
not present in the lung. Vacuolar degeneration of hepato-
cytes was consistent with steroid hepatopathy.
Coronal sections of formalin-fixed spinal cord re-
vealed a “V-shaped” area of white discoloration corre-
sponding to the dorsal funiculi that was more prominent
in the lumbar and cervical spinal cord. The most remark-
able histological lesions were present in the white matter
of the spinal cord and consisted of severe, diffuse loss of
axons and demyelination in the dorsal funiculi (i.e.,
fasiculus cuneatus and gracilis) and the dorsal nerve
roots. Longitudinal sections of the spinal cord revealed
marked wallerian degeneration with linear chains of di-
gestion chambers in the dorsal columns. The dorsal roots
and spinal ganglia had perivascular and diffuse intersti-
tial infiltrates of predominantly lymphocytes with fewer
macrophages and plasma cells; the lumbar region was
affected more severely [Figure 1]. This ganglionitis was
accompanied by mild to marked loss of neuronal cell
bodies and proliferation of satellite cells. Histomor-
phology of the ventral roots and trigeminal nerve was
normal. The right and left sciatic nerves had some fas-
cicles with axonal loss, dilated myelin sheaths,
May/June 1999, Vol. 35
Figure 1 —Dorsal nerve roots from a dog diagnosed with a
sensory polyganglioradiculoneuritis, showing infiltrates of
lymphocytes and marked loss of axons and myelin (Hematoxylin
and eosin stain, 180X).
intramyelin infiltration by large lipid-laden macro-
phages, and focal areas of mononuclear cell (primary
lymphocytes) infiltration. Muscle innervated by the sci-
atic nerve was histologically normal, supporting the di-
agnosis of a primary sensory neuritis. No lesions were
observed in any part of the brain, multiple skeletal
muscles, or enteric ganglia (Meissner’s or Auerbach’s).
Discussion
From the clinical signs and necropsy findings, a sensory
polyganglioradiculoneuritis affecting all spinal nerves
was confirmed. Involvement of the trigeminal nerve was
suspected because of absent facial nociception and lip
and tongue proprioception, but this was not confirmed
histologically. Sensory polyganglioradiculoneuritis has
been rarely reported in dogs, and the lymphocytic neuri-
tis extending along the length of the sciatic nerve, as in
this case, has not been reported previously. In the previ-
ously reported eight cases, three were males and five
were females of varying large and small breeds with an
age of onset ranging from 1.5 to nine years.1–4 All ani-
mals were ataxic. Conscious proprioception was exam-
ined in six cases and was reduced. The patellar reflex
was examined in eight affected dogs and was reduced or
absent in seven. Nociception of the face was evaluated in
eight dogs and was reduced in six dogs, increased in one
dog, and normal in one dog. One other dog had anisoco-
ria. Other abnormalities seen that did not occur in the
case reported here were megaesophagus, head tilt, voice
loss, urinary and fecal incontinence, and hearing loss.
Crossed extensor reflex was not reported in the other
cases. The cause of the crossed extensor reflex, usually
associated with upper motor neuron disease, was not
determined in this dog. All dogs progressively wors-
ened, and although necropsy confirmed sensory ganglio-
radiculitis, the underlying cause was not determined in
any case.
May/June 1999, Vol. 35 Polyganglioradiculoneuritis 235

Underlying causes of sensory polyneuropathies in hu- logical examination of a spinal nerve dorsal root gan-
mans include infectious diseases such as leprosy, human glion biopsy antemortem. In this and previously reported
immunodeficiency virus, and Lyme disease.5 An attempt cases, the diagnosis was confirmed at necropsy.
to rule out infectious diseases by evaluating serum titers
to specific organisms was performed in this dog. Meta- a
Optimmune; Schering-Plough, Kenilworth, NJ
bolic and endocrine disturbances associated with diabe- b
Matulene; Roche, Pfizer Animal Health, Exton, PA
tes mellitus, uremia, and hypothyroidism can produce c
Amoxi-tabs; SmithKline Beecham, Pfizer Animal Health, Exton, PA
primary sensory neuropathies in humans.5 Serum tests to d
Baytril; Miles, Bayer Co., Shawnee Mission, KS
evaluate these disorders also were normal in this dog.
Inherited sensory polyneuropathy seemed unlikely in Acknowledgments
this case, because he was a mixed-breed dog with an The authors wish to thank Dr. Vanda Lennon, Dr. Stephen
onset of signs at nine years of age and the lesion was Huber, and Patricia A. Pignataro for their assistance with
inflammatory in nature. Paraneoplastic sensory poly- this case.
neuropathies occur in humans, but no neoplastic condi-
tion was found in this dog. Antineuronal antibody (ANA)
tests are often positive in humans with paraneoplastic References
sensory polyneuropathies, but they were negative in this 1. Cummings JF, deLahunta A, Mitchell WJ. Ganglioradiculitis in the dog: a
clinical, light- and electron microscopic study. Acta Neuropathol
dog.5 No response was seen after 21 days of the antican- 1983;60:29–39.
cer drug, procarbazine. 2. Wouda W, Vandevelde M, Oettli P, et al. Sensory neuronopathy in dogs.
This dog had no exposure to toxic substances, such as A study of four cases. J Comp Pathol 1983;93:437–50.
3. Steiss JE, Pook HA, Clark EG, et al. Sensory neuropathy in a dog. J Am
vincristine, thallium, and hexocarbons, known to pro- Vet Med Assoc 1987;190:205–8.
duce sensory neuropathies in humans.5 Although a spe- 4. Duncan ID, Cuddon PA. Sensory neuropathy. In: Kirk RW,
cific history of exposure to organophosphates was Bonogura JD, eds. Current veterinary therapy X small animal practice.
Philadelphia: WB Saunders, 1989:822–7.
lacking, a serum acetylcholinesterase level was obtained
5. Smith BE, Windebank AJ, Dyck PJ. Nonmalignant inflammatory sensory
in this case due to the occasional environmental use of polyganglionopathy. In: Dyck PJ, Thomas PK, eds. Peripheral neuropathy.
these substances. Philadelphia: WB Saunders, 1993:1525–31.
By process of elimination, an autoimmune process 6. Cojocaru M, Spataru E, Dinu E, et al. The role of certain immunologic
parameters in silicosis. Rom J Intern Med 1995;33:61–72.
was suspected even though the ANA was negative. A 7. Orozan K, Ucan H, Tutkak H, et al. Systemic lupus erythematosus arising
positive ANA supports an autoimmune disease diagno- in a patient with chronic silicosis [letter] Rheumatol Int 1997;16:217–8.
sis in dogs, but a negative ANA does not rule it out.
Xerostomia, xerophthalmia, and severe sensory neur-
opathy with prominent sensory ataxia have been reported
in humans with the immune-mediated disorder, Sjogren’s
syndrome.5 One of the pathological processes in these
patients can be dorsal root ganglionitis with infiltrates of
T-lymphocytes and macrophages. Humans with
Sjogren’s syndrome test positive for antibodies to ex-
tractable nuclear antigens such as ro (SS-A) and la (SS-
B).5 These antibodies were not evaluated in this dog,
because the appropriate reagents could not be located.
The role of silica gel inhalation causing immune-
mediated disease in this dog is questionable. Silicosis in
humans is associated with a hyperactivity of humoral-
mediated immunity, and systemic lupus erythematosus
has been associated with silicosis.6,7 Silica crystals were
present in the lung tissue of this dog, so it is possible that
silica exposure may have resulted in immune stimulation
and clinical sensory polyganglioradiculoneuritis in this
dog.

Conclusion
Although rare, sensory polyganglioradiculoneuritis
should be considered in any dog that has multiple sen-
sory deficits of the head and limbs with normal motor
functions. The diagnosis might be confirmed by histo-
 Mitoxantrone and Cyclophosphamide
Combination Chemotherapy for the
Treatment of Various Canine Malignancies
Thirteen dogs with histopathologically confirmed malignancies were treated with mitoxantrone
and cyclophosphamide combination therapy. One to four doses were administered at 21-day
intervals. Recombinant human granulocyte colony-stimulating factor was administered to
ameliorate myelosuppression in dogs with neutrophil nadirs less than 1,000/µl. While the
protocol appears to be safe for use in tumor-bearing dogs, an advantage over mitoxantrone
single-agent protocols in terms of tumor response was not demonstrated in this initial pilot
study. J Am Anim Hosp Assoc 1999;35:236–9.

Carolyn J. Henry, DVM, MS Introduction

Michael S. Buss, DVM, MS
Kathleen A. Potter, DVM, PhD
K. Jane Wardrop, DVM, MS
O fibrosarcoma, chondrosarcoma, myxosarcoma, oral and cutaneous malig-
From the Department of Veterinary
Medicine and Surgery (Henry),
University of Missouri-Columbia,
379 East Campus Drive,
Columbia, Missouri 65211
and the Departments of Veterinary Clinical
Sciences (Buss, Wardrop) and Veterinary
Microbiology and Pathology (Potter),
Washington State University,
Pullman, Washington 99164.
Doctor Buss’ current address is
Veterinary Internal Medicine Specialists of
Kansas City,
1033 Metcalf Avenue,
Overland Park, Kansas 66212.
Reprint requests should be addressed to
Dr. Henry.
Funding for this project was provided by the
Brink Endowment Fund, the Autzen
Endowment Fund, and the Adler Fund,
Pullman, Washington and the Department of
Veterinary Medicine and Surgery Faculty
Practice Plan,
University of Missouri-Columbia.
Mitoxantronea is a synthetic derivative of anthracene used to treat neo-
plastic disease in patients unable to tolerate the side effects and toxicity of
doxorubicin.1 Mitoxantrone produces milder and less frequent side ef-
fects than doxorubicin.2 In a study which enrolled 126 tumor-bearing
dogs, mitoxantrone produced a complete or partial remission in 23% of
treated dogs.3 Tumors for which remission was achieved included squa-
mous cell carcinoma, renal and mammary adenocarcinomas, lymphoma,
nant melanoma, mesothelioma, hemangiopericytoma, and thyroid, transi-
tional cell, hepatocellular, and rectal carcinomas. Mitoxantrone and
cyclophosphamide combination protocols have been evaluated as initial
therapy for metastatic breast cancer and for treatment of various refrac-
tory solid tumors in humans.4–7 Response rates are comparable to those
obtained using doxorubicin and cyclophosphamide combination proto-
cols, with lower toxicity. 4,6,7 Combination mitoxantrone and cyclophos-
phamide protocols have not been investigated widely in veterinary
medicine, in part due to the assumptions that myelosuppression would
preclude clinical utility. However, availability of recombinant colony-
stimulating factors that may be used to ameliorate myelosuppression has
made more aggressive protocols clinically feasible. The purpose of the
study reported here was to conduct a pilot study to evaluate the hemato-
logical and clinical responses of tumor-bearing dogs to mitoxantrone and
cyclophosphamide combination therapy.
Materials and Methods
The study population consisted of client-owned dogs with various malig-
nant tumors [Table 1] treated at the Washington State University Veteri-
nary Teaching Hospital and the University of Missouri Veterinary Medical
Teaching Hospital between June 1994 and September 1997. All diag-
noses were made by histological examination of biopsy tissues. Diag-
noses made at other veterinary institutions were confirmed by review of
histological specimens by one pathologist (Potter).
Baseline evaluations prior to study enrollment included a complete
blood count, serum biochemical analysis, urinalysis, and bone marrow
aspiration/cytology. Dogs with segmented neutrophil counts less than

236 JOURNAL of the American Animal Hospital Association

May/June 1999, Vol. 35 Mitoxantrone and Cyclophosphamide Combination Chemotherapy 237

Table 1
Mitoxantrone and Cyclophosphamide Chemotherapy for Treatment of
Various Canine Malignancies

Case Tumor Number of Survival

No. Tumor Type Location Doses Response* (days)
1 Myxosarcoma Pelvic inlet 2 PD 260
2 Fibrosarcoma Multiple subcutaneous 2 PD 720†
3 Squamous cell carcinoma Tongue 4 SD 501†
4 Lymphoma Stage IIIb Multicentric 3 CR 88
5 Undifferentiated sarcoma Left antebrachium 1 PD 32
6 Squamous cell carcinoma Tonsil 1 PD 234
7 Malignant fibrous histiocytoma Spleen 4 SD‡ 402
8 Adenocarcinoma Mammary gland 3 SD 160
9 Hemangiosarcoma Spleen and liver 2 PD 56
10 Undifferentiated sarcoma Coxofemoral joint 4 SD 109†
and sublumbar
lymph nodes
11 Hemangiosarcoma Subcutaneous 3 PD 229
12 Synovial cell sarcoma Stifle 4 SD‡ 1164†
13 Lymphoma Stage Vb Multicentric 4 PR 65

* PD=Progressive disease; greater than 50% increase in tumor volume or development of new or metastatic lesions;
SD=Stable disease; less than 50% increase or decrease in tumor volume, without development of new metastatic
lesions; CR=Complete remission; disappearance of all clinically detectable disease; PR=Partial remission; 50% or
greater reduction in tumor volume and no evidence of new lesions
†
Alive at completion of study
‡
No measurable disease at initiation of chemotherapy

3.0 x103 cells/µl, platelet counts less than 200 x103
cells/µl, or both were reevaluated one week later. If
neutropenia or thrombocytopenia persisted, the dog was
ineligible for study enrollment.
Tumors were staged using the World Health Organi-
zation classification scheme, with methods of diagnosis
of metastasis dependent on primary tumor type and loca-
tion.8 Minimum tumor staging techniques included pal-
pation of regional lymph nodes and three-view thoracic
radiographs. Primary tumor dimensions (i.e., longest di-
ameter, diameter perpendicular to longest diameter, and
height) were made by use of calipers, and tumor volume
(cm3) was calculated for ellipsoid masses using the for-
mula (length x width x height)π/6. Tumor staging and
measurement were reevaluated every 21 days throughout
the treatment period.
Mitoxantrone was administered at a dosage of 5 mg/m2
intravenously (IV) every 21 days for four treatments or
until development of disease progression or severe tox-
icity. Mitoxantrone was diluted in 20 ml of 0.9% sodium
chloride (NaCl) and administered over five minutes.
Cyclophosphamideb was administered at a dosage of
150 mg/m 2 IV on the same days as mitoxantrone.
Furosemidec (2 mg/kg body weight, q 12 hrs) was ad-
ministered orally on the day of chemotherapy adminis-
tration to promote diuresis and decrease the risk of cy-
clophosphamide-induced, sterile, hemorrhagic cystitis.
White blood cell (WBC) and platelet counts were
performed daily after the first chemotherapy dose and
were continued for a minimum of two days after the
observed neutrophil nadir. Based on the timing of each
dog’s initial nadir, scheduling of WBC and platelet count
evaluation after subsequent doses included a minimum
of three counts collected as follows: 48 hours prior to the
anticipated neutrophil nadir, during the anticipated na-
dir, and 48 hours later. If the segmented neutrophil count
dropped below 3.0 x103 cells/µl, prophylactic antibiotic
therapy was initiated using trimethoprim/sulfadiazined
(15 mg/kg body weight, per os [PO] bid). If the seg-
mented neutrophil count decreased to less than 1.0 x103
cells/µl, parenteral antibiotic therapy was initiated using
enrofloxacine (5 mg/kg body weight, IV) and cefazolin f
(22 mg/kg body weight, IV), and oral antibiotics were
discontinued. In addition, human recombinant granulo-
cyte colony-stimulating factorg (rhG-CSF) was adminis-
tered subcutaneously (5 µg/kg body weight, per day) for
five days or for a minimum of two days after resolution
of neutropenia. Dogs showing clinical signs of sepsis,
238 JOURNAL of the American Animal Hospital Association May/June 1999, Vol. 35

Table 2
Hematological Results for 13 Normal Dogs Following Administration of Mitoxantrone*
and Cyclophosphamide†

Number of Median Range of Median Days Range of Median Days

Dogs Neutrophil Neutrophil Postchemotherapy Median Platelet Platelet Postchemotherapy
Receiving Nadir Nadirs to Neutrophil Nadir Nadirs to Platelet
Dose Dose (cells/µl) (cells/µl) Nadir (platelets x103/µl) (platelets x103/µl) Nadir
1 13 320 9 – 4,032 8 80 39 – 189 9
2 10 553 380 – 2,666 8 110 64 – 487 9.5
3 7 600 132 – 1,242 8 96 36 – 273 8
4 5 300 218 – 954 9 91 75 – 338 9

* 5 mg/m2 , intravenously every 21 days for up to four doses

†
150 mg/m2 , intravenously every 21 days for up to four doses

including fever and lethargy, were evaluated with urine
and three blood cultures prior to administration of antibi-
otics. Absence of a left shift was not considered to be
evidence of a lack of sepsis.
Response to therapy was determined by tumor restag-
ing prior to each treatment. Primary tumor measurement
and evaluation for metastatic disease were performed by
appropriate methods (i.e., ultrasonography, radiography,
palpation). Complete remission (CR) was defined as dis-
appearance of all clinically detectable disease. Partial
remission (PR) was defined as 50% or greater reduction
in tumor volume and no evidence of new lesions. Stable
disease (SD) was defined as less than 50% increase or
decrease in tumor volume, without development of new
or metastatic lesions. Progressive disease (PD) was de-
fined as greater than 50% increase in tumor volume or
development of new or metastatic lesions.
Statistical analysis of survival times and remission
duration was anticipated to have little value since the
study included several tumor types. Therefore, response
to therapy was determined as the overall percentage of
CR, PR, SD, and PD for the entire study population and
CR, PR, SD, and PD for each tumor type. The median
values for neutrophil and platelet count nadirs were also
determined.
Results
Thirteen dogs with various tumor types were enrolled
into the study [Table 1]. Remissions (CR or PR) were
achieved in only two dogs, both of which had multicen-
tric lymphoma. Both dogs with lymphoma had received
prior treatment. One had received prednisone, and the
other had received one dose of L-asparaginase, although
both dogs were out of remission at the time of mitoxantrone
and cyclophosphamide administration. All other dogs
had SD or PD while on the study protocol. Three (case
nos. 3, 6, 8) of these dogs had been enrolled previously
in an immunotherapy trial using a monoclonal antibody
directed against a tumor antigen. These three dogs were
removed from the immunotherapy trial when their tu-
mors became unresponsive to therapy. None of the other
dogs had received prior medical treatment for their tu-
mors. Of the 13 dogs enrolled, two did not have measur-
able disease at the time of study enrollment, but they
were included to assess the hematological effects and
toxicity of the mitoxantrone and cyclophosphamide com-
bination protocol. One (case no. 12) of these dogs is still
alive with no signs of disease recurrence, and the other
(case no. 7) died 402 days after initial diagnosis with
hepatic metastasis. Of the remaining 11 dogs, three were
still alive at the time of data analysis. Two had stable
disease. One (case no. 2) underwent subsequent cyto-
reductive surgery and radiation therapy and had disease
stabilization at the time of data analysis.
Side effects of the combination chemotherapy proto-
col included fever and lethargy (n=1; case no. 8); loose
stools (n=4; case nos. 8, 10, 11, and 13) within three to
seven days after chemotherapy; decreased appetite (n=5;
case nos. 1, 3, 5, 8, and 11) one to seven days after
chemotherapy; and moderate to marked neutropenia. The
median neutrophil nadirs and median days to nadir are
listed in Table 2. Segmented neutrophil counts returned
to greater than 1.0 x103/µl within one to four days after
the nadirs (median, two days). Signs of sepsis were not
noted in any dog in association with neutropenia. The
dog that experienced fever and lethargy did so within 72
hours of chemotherapy administration, at which time her
neutrophil count was within normal limits. Blood and
urine cultures revealed no infectious organism, and signs
resolved within 36 hours. Platelet nadirs occurred an
average of eight to nine days after chemotherapy. Me-
dian platelet nadirs for doses one through four are
reported in Table 2. Although bleeding times and coagu-
lation profiles were not performed, no clinical evidence
May/June 1999, Vol. 35 Mitoxantrone and Cyclophosphamide Combination Chemotherapy
of bleeding, petechiation, or ecchymoses was noted in
any dogs during periods of thrombocytopenia. Necropsy
evaluations have been performed on 10 of the study
dogs. No evidence of chemotherapy-induced pathology
was noted in any organ, including heart and kidney.
Discussion
The combination protocol used in this study was well
tolerated by all dogs. Although myelosuppression was
moderate to marked in 12 of 13 dogs, no dog became
septic. Administration of rhG-CSF at the time of neutro-
phil nadir precludes the authors’ ability to assess the true
degree of myelosuppression that would occur secondary
to administration of mitoxantrone and cyclophospha-
mide. It is possible that the timing of nadirs and the
severity of myelosuppression were altered by rhG-CSF
administration. However, as these were client-owned
animals, the protocol included a margin of safety. The
necessity of rhG-CSF in these cases is unknown. In a
previous study, administration of mitoxantrone (5 mg/m2)
and cyclophosphamide (200 mg/m2) to normal dogs pro-
duced rapid and severe neutropenia (polymorphonuclear
cells less than 500/µl) in all dogs, with 6.7 (±2.5 standard
deviations) mean days of neutropenia.9 By comparison,
when dogs were treated with 5 µg/kg rhG-CSF begin-
ning two days after chemotherapy, duration and severity
of neutropenic episodes were decreased, with mean days
of neutropenia being 4.7 (±3.7 standard deviations). The
present study did not include control dogs. However, as
designed, the protocol appears to be safe and well toler-
ated in tumor-bearing dogs. Bone marrow recovery was
noted after each course of chemotherapy, indicating that
the repeated use of rhG-CSF did not result in develop-
ment of neutralizing antibodies to the recombinant hu-
man protein. The authors’ previous studies of repeated
rhG-CSF administration in dogs receiving chemotherapy
have demonstrated a similar suppression of antibody
formation in dogs receiving chemotherapy.9,10
The tumor responses to the mitoxantrone and cyclo-
phosphamide combination protocol were disappointing
in this small group of cases. Of the tumor types exam-
ined, only lymphomas responded to the protocol. Of the
five cases with disease stabilization, only three had in-
complete tumor excision or lymph node metastasis prior
to initiation of chemotherapy. Of these three, one died of
metastatic disease while enrolled in the study. Due to the
small number of cases and the inclusion of two dogs with
no measurable disease, it was not possible to determine
the efficacy of the combination protocol. However, pre-
liminary results suggest no advantage over mitoxantrone
single-agent protocols for the treatment of sarcomas and
carcinomas. The protocol induced remission in dogs with
lymphoma. The overall 18% response rate in dogs with
measurable disease suggests that the protocol may not
offer an advantage over mitoxantrone alone. A larger,
239
randomized prospective trial would be necessary to de-
termine the comparative efficacy of the two protocols.
a
Novantrone; Immunex Corporation, Seattle, WA
b
Cytoxan; MeadJohnson Oncology Products, Princeton, NJ
c
Lasix; Hoechst-Roussel Pharmaceuticals, Inc., Somerville, NJ
d
Tribrissen; Mallinckrodt Veterinary, Inc., Mundelein, IL
e
Baytril; Bayer Corporation, Shawnee Mission, KS
f
Cefazolin; Apothecon, Princeton, NJ
g
Neupogen; Amgen Corporation, Thousand Oaks, CA
Acknowledgments
Mitoxantrone was graciously provided by Immunex Cor-
poration, Seattle, Washington. The authors sincerely
thank Lawrence M. Souza (Amgen, Inc., Thousand Oaks,
California) for kindly providing rhG-CSF. The authors
also wish to thank Lynn Russel and Debbie Tate for their
technical assistance.
References
1. Riggs CE. Antitumor antibiotics and related compounds. In: Perry MC, ed.
The chemotherapy sourcebook. Baltimore: Williams & Wilkins, 1992:
318–58.
2. Ogilvie GK, Obradovich JE, Elmslie RE, et al. Efficacy of mitoxantrone
against various neoplasms in dogs. J Am Vet Med Assoc 1991;198:
1618–21.
3. Ogilvie GK, Obradovich JE, Elmslie RE, et al. Toxicoses associated with
administration of mitoxantrone to dogs with malignant tumors. J Am Vet
Med Assoc 1991;198:1613–7.
4. Bezwoda WR, Hesdorffer C. The use of mitoxantrone plus cyclophospha-
mide as first-line treatment of metastatic breast cancer. Cancer 1986;58:1621–4.
5. Mulder POM, Sleijfer DT, Willemse PHB, et al. High-dose cyclophospha-
mide or melphalan with escalating doses of mitoxantrone and autologous
bone marrow transplantation for refractory solid tumors. Cancer Res
1989;49:4654–8.
6. Bennet JM, Muss HB, Doroshow JH, et al. A randomized multicenter trial
comparing mitoxantrone, cyclophosphamide, and fluorouracil with
doxorubicin, cyclophosphamide, and fluorouracil in the therapy of
metastatic breast carcinoma. J Clin Oncol 1988;6:1611–20.
7. Leonard RCG, Cornbleet MA, Kaye SB, et al. Mitoxantrone versus
doxorubicin in combination chemotherapy for advanced carcinoma of the
breast. J Clin Oncol 1987;5:1056–63.
8. Owen LN, ed. TNM classification of tumors in domestic animals. Geneva:
World Health Organization, 1980.
9. Goodman SA, McPherson J, Rossimeisel JA, et al. Recombinant human
granulocyte colony stimulating factor decreases myelosuppression in
normal dogs following repeated cycles of mitoxantrone and cyclophospha-
mide chemotherapy without development of neutralizing antibodies. J Vet
Int Med, in press.
10. Henry CJ, Buss MS, Lothrop C. Veterinary uses of recombinant human
granulocyte colony-stimulating factor. Part I: Oncology. Compendium
1998;20:728–35.
 Prolapse of the Gland of the Third Eyelid
in a Cat: A Case Report and
Literature Review
A one-year-old, castrated male, domestic shorthair was presented with a prolapse of the gland
of the third eyelid. Similar to the dog, this cat did not appear to suffer from this eye condition.
The prolapse was unilateral on the left eye. The right eye was normal. No accompanying eye
diseases were found. Treatment consisted of a surgical pocket technique similar to the
technique used in dogs. J Am Anim Hosp Assoc 1999;35:240–2.

Sabine H. Schoofs, DVM Introduction

Prolapse of the gland of the third eyelid is a common condition in the dog.
It has been reported as glandular hyperplasia or hypertrophy or “cherry
C eye.”1–3 The cause is still unknown. A congenital, possible hereditary,
weakness or laxity of the connective tissue bands within the soft tissue
surrounding the gland and the other periorbital tissues has been pro-
posed.1–3 Others suggest a mild inflammation of the gland as a possible
cause or a scrolling of the cartilage around the gland.4 Some breeds (e.g.,
English bulldog, Pekingese, basset hound, beagle, Boston terrier, shih tzu,
cocker spaniel, and Lhasa apso) are more frequently presented with a
prolapse of the gland.1,5 The author also sees a number of Mastino
napolitano, French bulldogs, and Maltese with this condition.
To the author’s knowledge, no case of prolapse of the gland of the third
eyelid in a cat without other ocular abnormalities has been reported. This
case report describes the condition in a cat and the surgical technique that
was used.

From the Department of Opthalmology
and Soft-Tissue Surgery,
Dierenartsenpraktijk “Clos Fleuri”
Kortrijksesteenweg 1089,
9051 Sint-Denijs Westrem,
Belgium, Europe.
Case Report
A 3.5-kg, one-year-old, castrated male, domestic shorthair was presented
with a prolapsed gland of the third eyelid of the left eye [Figure 1]. The
swelling had occurred suddenly, three months earlier. At first the owner
had given eyedrops containing a combination of hydrocortisone and
neomycina twice a day for some weeks. The eye didn’t improve with this
therapy. Then a chloramphenicolb ointment was applied four times a day.
Once again this therapy did not make the swelling disappear. The cat did
not seem to suffer in any way. His appetite and behavior were normal. A
physical examination was done, and nothing remarkable was found. Upon
ophthalmological examination, the conjunctiva on the left eye was slightly
swollen and reddened. No blepharospasm or ocular discharge was present.
A red gland was visible just behind the third eyelid. A Schirmer tear test
(STT) was performed. The results were 15 mm per minute on the left eye
and 13 mm per minute on the right eye. The right eye had no abnormalities.
Surgery was done the same day. The cat was anesthetized with a
combination of rompunc and ketamined (1.8 mg/kg body weight, and 20
mg/kg body weight, respectively intramuscularly). The eyelids of the left
eye were shaved. The conjunctival sac was flushed with warm, sterile
saline, and the eyelids were disinfected with a diluted solution of povi-
done iodide. Two 1-cm incisions were made in the ocular conjunctiva of
the third eyelid, one dorsal and one ventral to the gland. Both incisions

240 JOURNAL of the American Animal Hospital Association

May/June 1999, Vol. 35
Figure 1 —Photograph clearly showing the hyperplastic gland at
the inner side of the third eyelid in the young cat of this report.
were parallel to the margin of the third eyelid. The two
wound edges most remote to each other were sutured in a
continuous pattern with 6-0 polyglactin 910.e In this
manner, a pocket is created within the conjunctiva on the
ocular side of the membrana nictitans, within which the
prolapsed gland was contained, ensuring the gland is
both retained and reduced. Extreme care was taken to
have the sutures buried in the mucosa. To avoid the risk
of having an ulcer on the cornea, the first knot of the
suture material was made on the palpebral side of the
third eyelid, then the suture material was passed through
the third eyelid to start the continuous suture. At the end
of the suture, the needle again was passed through the
third eyelid to make the last knot on the palpebral side.
With this technique, only a minimal amount of suture
material could have contact with the cornea. This mini-
mized the risk of corneal ulceration.
As aftercare, an antibioticb ointment (four times a day
for 10 days) was applied topically to the left eye. The cat
recovered without any problem, and because the eye
looked normal, the owner did not pursue follow-up
examination.
Discussion
To the author’s knowledge, this is the first case of hyper-
plasia of the gland of the third eyelid as a solitary eye
problem in a cat. Only two case reports previously have
described a “cherry eye” in the cat. The first case report
describes two Burmese with folded cartilage and acces-
sory prolapsed glands.6 The second case report describes
Third Eyelid Gland 241
a series of congenital ocular and nasal disorders in Bur-
mese kittens.7 One of these kittens had a prolapsed gland
among other ocular abnormalities. In these cases, the
gland was replaced surgically by using a suture from the
cartilage to the m. obliquus ventralis, which is a modi-
fied Blogg’s technique. 8 The author was unable to find
any other information about prolapsed glands in the cat.
McLaughlin, et al. 9 looked at the consequences of re-
moving the gland of the third eyelid and the lacrimal
gland in the cat. Removing the gland of the third eyelid
lowered the STT test by 15.7%. Removing the lacrimal
gland lowered the tear production by 34.1%. In dogs,
these percentages are even higher.10 It is therefore im-
portant to try to save the prolapsed gland instead of
excising it.
Many different techniques have been described and
used in the dog to treat a prolapsed gland. Gelatt11 in-
cludes the following classification of the different surgi-
cal procedures: 1) Posterior nictitans anchoring or tacking
approach to the ventral oblique muscle,6 the ventral equa-
torial sclera,12 or to the ventral periorbital fascia;8 2)
anterior nictitans anchoring or tacking approach to the
ventral periorbital rim13 (with modifications4); and 3)
conjunctival mucosa envelope procedure14 and pocket
methods.1 The author always uses the pocket method
described by Morgan, with a personal variation. As de-
scribed earlier, the author places the first and last suture
knots in the palpebral side of the third eyelid. The author
started doing this after two cases of corneal ulceration
occurred, due to suture material irritating the cornea
after a surgical correction of a prolapsed gland. With the
two knots placed in this way, they cannot irritate the
cornea, and it becomes easy to avoid contact between the
rest of the suture and the cornea because no thick parts of
suture material remain.
a
Neocortef (hydrocortisonum acetatum 15 mg/ml-Neomycin 3.5 mg/ml);
Laboratoires Cusi, Barcelona, Spain, Europe
b
Gloveticol (chloramphenicolum 10 mg/g); Mycofarm Belga, Turnhout,
Belgium, Europe
c
Xyl-M 2% (Xylazinum 20 mg/ml); VMD, Arendonk, Belgium, Europe
d
Aescoket (Ketaminum 100 mg/ml); Aesculaap, Boxtel, the Netherlands,
Europe
e
Vicryl, polyglactine 910; Ethicon, Neuilly sur Seine, France, Europe
References
1. Morgan RV, Duddy JM, McClurg K. Prolapse of the gland of the third
eyelid in dogs: a retrospective study of 89 cases (1980 to 1990). J Am
Anim Hosp Assoc 1993;29:56–60.
2. Bromberg N. Surgical procedures of the nictitating membrane. In: Bojrab
MJ, ed. Current techniques in small animal surgery. 2nd ed. Philadelphia:
Lea & Febiger, 1983:50.
3. Helper LC. The canine nictitating membrane and conjunctiva. In: Gelatt
KN, ed. Textbook of veterinary ophthalmology. Philadelphia: Lea &
Febiger, 1981:330–1.
(Continued on next page)
242 JOURNAL of the American Animal Hospital Association May/June 1999, Vol. 35

References (cont’d) 10. Helper LC. The effect of lacrimal gland removal on the conjunctiva and
cornea of the dog. J Am Vet Med Assoc 1970;157:72–5.
4. Peterson-Jones S. Repositioning prolapsed third eyelid glands while
11. Gelatt KN, Gelatt JP. Surgical procedures for protrusion of the gland of the
preserving secretory function. In Practice 1991;13:202–3.
nictitating membrane. In: Edney ATB, ed. Handbook of small animal
5. Dugan ST, Severin GA, Hungerford LL, Whiteley HE, Roberts SM. ophthalmic surgery. 1st ed. Trowbridge: Pergamon, 1994:149–57.
Clinical and histologic evaluation of the prolapsed third eyelid gland in
12. Gross SL. Effectiveness of a modification of the Blogg technique for
dogs. J Am Vet Med Assoc 1992;12:1861–7.
replacing the prolapsed gland of the canine third eyelid. Proc Am Coll Vet
6. Albert RA, Garett PD, Whitley RD. Surgical correction of everted third Ophthal 1983;13:38–42.
eyelid in two cats. J Am Vet Med Assoc 1982;180:763–6.
13. Kaswan RL, Martin CL. Surgical correction of the third eyelid prolapse in
7. Christmas R. Surgical correction of congenital ocular and nasal dermoids dogs. J Am Vet Med Assoc 1985;186:83.
and third eyelid gland prolapse in related Burmese kittens. Can Vet J 1992;
14. Moore CP. Imbrication technique for replacement of prolapsed third eyelid
33:265–6.
gland. In: Bojrab MJ, ed. Current techniques in small animal surgery. 3rd
8. Blogg JR. Surgical replacement of a prolapsed gland of the third eyelid—a ed. Philadelphia: Lea & Febiger, 1990:126–8.
new technique. Aust Vet Pract 1979;9:75.
9. McLaughlin SA, Brightman AH, Helper LC, Primm ND, Brown MG,
Greeley S. Effect of removal of lacrimal and third eyelid glands on
Schirmer tear test results in cats. J Am Vet Med Assoc 1988;193:820–2.
 Bone Plate Fixation of Distal Radius
and Ulna Fractures in Small- and
Miniature-Breed Dogs
Bone plate fixation was reviewed in 29 distal radial fractures of small- and miniature-breed
dogs. Twenty-two fractures in 18 dogs were available for follow-up. Number of complications
and return to function were evaluated. Complications occurred in 54% of the fractures.
Catastrophic complications occurred in 18% of fracture repairs with follow-up, while minor
complications occurred in 36%. Sixteen (89%) of 18 dogs had a successful return to function.
Bone plate fixation is a successful repair method for distal radius and ulna fractures in small-
breed dogs, compared to previously reported methods. J Am Anim Hosp Assoc 1999;35:243–50.

Linda J. Larsen, DVM Introduction

James K. Roush, DVM, MS,
Diplomate ACVS
Ron M. McLaughlin, DVM, DVSc,
Diplomate ACVS
RS
From the Department of Clinical Sciences
(Larsen, Roush),
College of Veterinary Medicine,
Kansas State University,
Manhattan, Kansas 66506-5608.
Doctor McLaughlin’s current address is
Gulf Coast Veterinary Referral,
1235 Tallevast Road,
Sarasota, Florida 34243.
Fractures of the distal one-third of the radius and ulna are the third most
common fracture in dogs.1–3 The incidence of these fractures is particu-
larly high in small and miniature breeds of dogs.1–3 Small and miniature
breeds of dogs are particularly prone to nonunion of fractures of the distal
radius and ulna.1–13 Studies have shown that when distal radius and ulna
fractures in small and medium dogs were treated identically, delayed
union and nonunion complications occurred primarily in the small-breed
dogs.5
Inherent biomechanical instability,6,7,12,13 decreased intraosseous blood
supply,1,6,13–15 and limited overlying soft tissue for provision of an
extraosseous circulation6,13,15 all most likely contribute to the higher
frequency of delayed union and nonunion fractures in small- and minia-
ture-breed dogs. A marginal blood supply in the distal radius of small-
and miniature-breed dogs has been postulated as a cause of delayed union
or nonunion of these fractures.6 In small dogs, there was decreased
vascular density and arborization of the vessels in the distal metaphysis as
compared with medium-sized dogs. This paucity of vessels resulted in a
zone of reduced vascularity at the distal metaphyseal region of the radius
in small dogs.6 Biomechanical instability also may predispose to non-
union after fracture reduction. 6,7,12,13 There is minimal bone surface con-
tact resulting from the small bone size and the short oblique or transverse
orientation of many radius and ulna fractures.6,7,14 Anatomical realign-
ment of the bone is hindered by the tendency for the carpal and digital
flexor muscles to create caudolateral displacement of the distal bone
fragment.1,6,14
Fixation of distal radius and ulna fractures in small-breed dogs with
casts, intramedullary (IM) pins, external fixators, and bone plates has
been reported. Eighty-three percent of radial fractures treated with cast
fixation developed serious complications consisting of malalignment and
nonunion.1,4 Intramedullary pinning is not recommended for distal radius
and ulna fracture repair, because it fails to adequately counteract rota-
tional forces.4 Complications including angulation, distraction, rotation,
osteomyelitis, degenerative joint disease (DJD) of the carpus, delayed
union, and nonunion were seen in 80% (12/15) of distal radius and ulna
fractures repaired using IM pins. 1 Delayed union or nonunion fractures

JOURNAL of the American Animal Hospital Association 243

244 JOURNAL of the American Animal Hospital Association
Figure 1A
Figures 1A, 1B—Lateral (A) and craniocaudal (B) radiographs
of the repair of a transverse radius and ulna fracture in a 1.5-
year-old, intact female Italian greyhound. A single intramedullary
(IM) pin was placed normograde from a distal to proximal
direction.
were seen in 47% (7/15) of dogs reported, and amputa-
tion or an alternative fixation method was required in
27% (4/15) of the dogs.1
Open reduction and fracture stabilization using a bone
plate, and open or closed reduction and fracture stabili-
zation using an external fixator have been advocated for
adequate repair of distal radius and ulna fractures.4 Com-
plications associated with external skeletal fixation of
fractures include loose pins (27/28 fractures), pin tract
drainage (27/28), infection (12/28), valgus or rotational
malalignment (16/28), and delayed union and nonunion
(1/28). 16 Complications associated with bone plate fixa-
tion of radius and ulna fractures in all sized dogs include
implant failure (2/26 dogs), angulation (4/26), infection
(1/26), stress protection-induced osteopenia (2/26), and
thermal conduction and irritation (1/26).17
The purposes of this study were to evaluate the effi-
cacy of bone plate fixation of the distal radius in small-
and miniature-breed dogs and to document the type and
rate of complications associated with this fracture repair
technique in those dogs. There are no reports of the
May/June 1999, Vol. 35
Figure 1B
incidence and types of complications associated with
bone plate fixation of these fractures in small- and min-
iature-breed dogs.
Materials and Methods
The medical records of 84 dogs with distal radius and
ulna fractures presenting to the Kansas State University
Veterinary Medical Teaching Hospital (KSUVMTH) be-
tween 1985 and 1996 were reviewed. Those cases meet-
ing the following requirements were included in the
study: body weight less than 12 kg, transverse or short
oblique fracture of the distal diaphysis of the radius and
ulna, and fracture repair at the KSUVMTH with open
reduction and internal fixation utilizing a bone plate.
Twenty-nine fractures in 23 dogs fit the criteria for in-
clusion into the study.
Data pertaining to clinical history, fracture duration,
fracture description, previous repair attempts, fixation
methods utilized in previous repair attempts, plate size
and configuration, utilization of a cancellous bone graft,
postoperative fracture alignment, postoperative compli-
cations, and time from fracture fixation (i.e., plate place-
ment) to last follow-up radiographs was recorded.
Cases with incomplete previous follow-up were con-
tacted and, if possible, radiographed to assess healing of
May/June 1999, Vol. 35
Figure 2A
Figures 2A, 2B—Lateral (A) and craniocaudal (B) radiographs of
the fracture site of the 1.5-year-old, intact female Italian
greyhound (shown in Figure 1), following pin removal four
months after IM pin placement.
the fracture and any long-term complications. A tele-
phone follow-up questionnaire was devised. All ques-
tions were asked directly from the questionnaire, and
owner’s responses were recorded. Clients contacted by
telephone and without previous follow-up radiographs
were requested to return to the KSUVMTH for radio-
graphs. Clients for whom telephone follow-up was un-
successful were mailed the same questionnaire as used
for the telephone interviews.
All radiographs were reviewed by the investigators.
Criteria for evaluation of immediate postoperative and
follow-up radiographs included plate size and configura-
tion, implant placement, reduction and alignment, and
stage of healing. Reduction and alignment were defined
as excellent if there was less than 1% angulation and
adequate if there was minor angulation. Complications
noted on postoperative radiographs included angulation,
implant failure, and osteopenia. Complications were cat-
egorized as catastrophic if plate failure occurred and
minor if angulation, osteopenia, thermal conduction, or
synostosis occurred. Lameness was evaluated on a graded
Bone Plate Fixation 245
Figure 2B
scale (1–5), with grade 1 being normal and grade 5 being
complete nonweight-bearing on the limb.18
Results
Fourteen of 23 owners were contacted by either tele-
phone or mail survey. Radiographic and clinical follow-
up were both available for 13 fractures in 12 dogs.
Additionally, radiographic follow-up but no telephone
or mail contact was available for four fractures, and
telephone or mail contact without radiographic follow-
up was available for five fractures. Two owners, con-
tacted by telephone, refused to return to the KSUVMTH
for follow-up radiographs due to travel distance required.
Mean and median last radiographic follow-up for the 17
fracture radiographs were nine months and four months,
respectively.
Italian greyhounds constituted eight (35%) of the 23
dogs in this study. Other breeds represented were the toy
poodle (n=6), toy fox terrier (n=3), Pomeranian (n=3),
Chihuahua (n=1), rat terrier (n=1), and papillon (n=1).
The mean body weight was 3.7 kg (range, 1.6 kg to 10.4
kg). The mean and median ages at fracture were 1.9
years (range, five months to seven years) and 1.5 years,
respectively. There was no predilection for right versus
left limb. Bilateral fractures occurred in six of 23 dogs.
246 JOURNAL of the American Animal Hospital Association May/June 1999, Vol. 35

Table
Radiographic and Lameness Evaluation for all Fractures

Dog Weight (kg) Plate Used* Radiographic Outcome Lameness Outcome

1 2.6 2.0-mm, 9-hole T Healed Normal

2 4.2 2.0-mm, 12-hole T Healed Unknown
3 3.9 2.0-mm, 7-hole DCP Healed Carries leg
4 1.7 2.0-mm, 6-hole mini Healed Normal
5 4.4 2.0-mm, 7-hole cuttable Healed Normal
6 3.5 2.0-mm, 5-hole DCP Healed Normal
7 3.5 2.0-mm, 5-hole DCP Unknown Normal
8 4.2 2.0-mm, 8-hole T Healed Normal
9 1.4 2.0-mm, 9-hole T Healed Normal
10 1.4 2.0-mm, 9-hole T Healed Occasionally lame
11 4.5 2.0-mm, 7-hole straight Unknown Frequently limps
12 2.3 2.0-mm, 5-hole DCP Healed Normal
13 2.0 2.0-mm, 7-hole T Healed Unknown
14 2.0 2.0-mm, 7-hole T Healed Unknown
15 5.9 2.0-mm, 7-hole mini Unknown Normal
16 5.9 2.0-mm, 6-hole mini Healed Normal
17 1.6 2.0-mm, 5-hole DCP Healed Occasionally lame
18 5.5 2.0-mm, 7-hole DCP Healed Occasionally lame
19 1.5 2.0-mm, 8-hole DCP Healed Unknown
20 4.2 2.0-mm, 7-hole DCP Unknown Occasionally lame
21 5.1 2.7-mm, 5-hole T Unknown Occasionally lame
22 2.3 2.7-mm, 4-hole straight Healed Normal

* T=“T”-plate; DCP=dynamic compression plate; mini=miniplate; cuttable=veterinary cuttable plate; straight=straight

plate

There were 16 females and seven males. An equal num-
ber of females were intact versus spayed, and an ap-
proximately equal number of males were intact versus
castrated (4/7 intact). Thirteen of the 23 fractures had
been repaired previously or stabilized, and three of those
more than once. Eleven of 13 fractures had been repaired
previously with cast fixation; four had been repaired
using an IM pin; two had been repaired with an exter-
nal fixator; and one had been repaired with a bone
plate.
All fractures were repaired at the KSUVMTH with a
2.0-mm or 2.7-mm bone plate.a Eight 2.0-mm dynamic
compression (DC) plates were used. Four were five-hole
plates, two were seven-hole plates, and two were eight-
hole plates. No holes were left open on any of the DC
plates. Seven 2.0-mm “T” plates were used. Four were
modified by removal of one (n=2) or both (n=2) bars of
the “T.” One open hole remained in two repairs with “T”
plates. Three 2.0-mm miniplates were used; two were
six-hole and one was a seven-hole plate. No holes were
left open with these repairs. One 2.0-mm, seven-hole
stacked cuttable plate was applied. One 2.0-mm, seven-
hole straight plate was applied. The mean and median
weights of the eight dogs in which a 2.0-mm DC plate
was used were 3.34 kg (range, 1.5 kg to 5.5 kg) and 3.5
kg, respectively. The mean and median weights of the
seven dogs in which a 2.0-mm “T” plate was used were
2.5 kg (range, 1.4 kg to 4.2 kg) and 2.0 kg, respectively.
The mean and median weights of the three dogs in which
a 2.0-mm miniplate was used were 4.5 kg (range, 1.7 kg
to 5.9 kg) and 5.9 kg, respectively.
Two plates were 2.7-mm; one was a “T” plate, and
one was a finger plate. The “T” plate was five-hole, and
the straight plate was a six-hole (modified to four-hole)
plate. No holes were left open in these repairs.
Either a cancellous or corticocancellous bone graft
was used in 45% (10/22) of the fractures. Bone grafts
were used more frequently (9/13) in chronic (i.e., one
week or longer duration) fractures than in acute (i.e.,
one- to three-day duration) fractures (1/9). Four chronic
fractures of one to three weeks duration were not grafted,
and one acute fracture of one day’s duration was grafted.
May/June 1999, Vol. 35
Figure 3A
Figures 3A, 3B—Lateral (A) and craniocaudal (B) radiographs
immediately following the second repair attempt using a seven-
hole, 2.0-mm dynamic compression (DC) plate in the 1.5-year-
old, intact female Italian greyhound from Figure 1.
Overall complications occurred in 54% (12/22) of the
fractures with follow-up. Catastrophic complications oc-
curred in 18% (4/22), and minor complications occurred
in 36% (8/22) of fractures.
Catastrophic complications included plate breakage
in three fractures and screw pull-out with loss of bone-
plate contact in one fracture. The weights of the dogs
which suffered plate breakage were 2.0 kg (“T” plate),
4.2 kg (“T” plate), and 4.2 kg (DC plate). Of the four
fractures with resultant catastrophic complications, frac-
ture durations were one week, two weeks, two months,
and 3.5 months. All four fractures had previous repair
attempts prior to bone plating. The first repair techniques
included an IM pin in three of the dogs (of two weeks,
two months, and 3.5 months duration [Figures 1, 2]) and
cast fixation in the fourth dog. A bone graft had been
used during the initial plating of two of the three frac-
tures in which a plate broke. One fracture was replated
using a 2.0-mm miniplate; one fracture was replated
using a 2.0-mm DC plate [Figures 3–6]; and the third
was replated using a 2.0-mm “T” plate. A cancellous
Bone Plate Fixation 247
Figure 3B
bone graft was utilized in all cases at the second plating.
All fractures went on to heal following the second plating.
Screw pull-out and loss of bone-plate contact oc-
curred three months postoperatively in one fracture.
Original plating consisted of a 2.0-mm miniplate and a
graft applied to a distal radial fracture in a dog weighing
4.5 kg. The implants were removed, and a cast was
applied to the limb. This dog was lost to follow-up.
Minor complications included skin erosion, angula-
tion, stress protection-induced osteopenia, thermal con-
duction, and synostosis. Skin erosion over the plate
occurred in two dogs. Extensive erosion in one dog was
treated with a skin graft. This skin graft was unsuccess-
ful, and the limb was amputated. Skin erosion in the
second dog was minimal, and plate removal was recom-
mended. At the time of this study, the owner has de-
clined plate removal and has treated the dog with
antibiotics intermittently when drainage from the site is
present. Protrusion of a lag screw through the skin oc-
curred in one dog. The screw was removed, resolving the
condition.
The postoperative alignment and apposition of the
radii were described as adequate to excellent in all cases
as read by the attending radiologist at the time the radio-
graphs were taken. Five (21%) cases had persistent limb
248 JOURNAL of the American Animal Hospital Association
Figure 4A
Figures 4A, 4B—Lateral (A) and craniocaudal (B) radiographs
showing implant failure nine weeks after the second fracture
repair attempt in the 1.5-year-old, intact female Italian greyhound
from Figure 1.
deformity based on follow-up radiographs. Three of these
experienced catastrophic complications including the two
broken plates and the fracture with screw pull-out and
subsequent dislodgement of the plate. Carpal valgus oc-
curred in one dog due to plate placement, and one dog
had a detectably shorter radius than the contralateral
limb and mild carpal hyperextension.
Osteopenia was radiographically detectable in three
cases but did not result in any fracture needing further
treatment.
One dog was reported to experience lameness when
out in the cold. Plate removal was recommended in this
case; however, the owner was not willing for the dog to
undergo a second surgical procedure. Synostosis was
noted infrequently. Since growth was complete in these
dogs, no clinical significance was detected.
Long-term follow-up regarding lameness evaluation
by the owner consisted of responses for 18 fractures.
Sixteen of the 18 fractures had a successful return to
function as evaluated by the owner [see Table]. Of these
18 fractured limbs, 11 were considered by the owner to
have normal use; five dogs experienced occasional lame-
May/June 1999, Vol. 35
Figure 4B
ness; one dog experienced moderate lameness; and one
dog carried the limb when running. Both cases experi-
encing moderate and nonweight-bearing lameness had
experienced implant failure. Two dogs with occasional
lameness also experienced implant failure.
Discussion
In this retrospective study, 89% of fractures had a suc-
cessful return to function as evaluated by the owner. The
results of this study support plate fixation of distal radius
and ulna fractures in small-breed dogs as an alternative
to other methods of fracture fixation and a preferable
fixation method when compared to casting or IM pin
fixation. Catastrophic complications occurred in 18%
(4/22) of fractures repaired with bone plate fixation,
compared to an 83% severe complication rate with cast
fixation,1,4 a 27% catastrophic complication rate with IM
pin fixation,1 and a 4% severe complication rate with
external fixators.16
Four fractures suffered catastrophic failure. Two of
the three dogs that experienced plate breakage were
heavier than the median weight of dogs with successful
repairs using that particular implant. The 4.2-kg dog
whose fracture was repaired using a 2.0-mm “T” plate
was the heaviest dog in which that particular plate was
used. In the failure involving a 2.0-mm DC plate, that
May/June 1999, Vol. 35
Figure 5A
Figures 5A, 5B—Lateral (A) and craniocaudal (B) radiographs
immediately following the third fracture repair attempt using an
eight-hole, 2.0-mm DC plate in the 1.5-year-old, intact female
Italian greyhound from Figure 1; lateral angulation is present.
dog also was heavier than the median weight of dogs
with successful repairs using 2.0-mm DC plates. It is
extremely important in treatment of these fractures that
the optimal-size plate be chosen for proper fixation and
that every effort be made to fill all screw holes in the
plates. Published guidelines for comparison of bone plate
strength indicate that the 2.0-mm “T” plate is as strong
as the larger (six- to eight-hole) 2.0-mm DC plate, but
the 2.0-mm miniplate and shorter (four- to six-hole) 2.0-
mm DC plate are only approximately 33% as strong as
either of the former plates.19
All of the plated fractures suffering catastrophic plate
failure had undergone at least one previous repair at-
tempt prior to plate fixation. In fact, of the 29 fractures
reviewed in this report, 16 had been repaired previously
with other fixation methods. It has been suggested that
delayed union and nonunion of distal radius and ulna
fractures are the result of a decreased blood supply to the
distal metaphyseal region;6 a lack of soft tissue for provi-
sion of extraosseous blood supply for early vasculariza-
tion and healing while the nutrient artery redevelops;16
Bone Plate Fixation 249
Figure 5B
and difficulty in maintaining biomechanical stability due
to the small diameter of the radius and the pull of the
digital flexor muscles.6,7,14 Repeated surgery, with fur-
ther disruption of the blood supply, resultant trauma to
the soft-tissue structures, scar tissue development, and
difficulty in obtaining adequate alignment may play a
role in the decreased return to normal function in these
cases.
Conclusion
In this study, 89% of the fractures repaired with bone
plates had a successful return to function as evaluated by
owners. A previous study showed a similar success rate
(88%) for return to function. This was compared with
the success rate of 93% utilizing external fixators, 50%
using IM pin fixation, and 43% using cast fixation.
In this study, clinical osteopenia was not found to be
a significant factor. In one dog, osteopenia of the radius
was radiographically evident nine weeks postoperatively.
This condition resolved with return to weight-bearing
activity by 21 weeks postoperatively. In three cases with
follow-up radiographs two and three years postopera-
tively, radiographic evidence of osteopenia was not
present.
250 JOURNAL of the American Animal Hospital Association May/June 1999, Vol. 35

Figure 6A Figure 6B

Figures 6A, 6B—Lateral (A) and craniocaudal (B) radiographs 9. Sumner-Smith G. A histologic study of fracture nonunion in small dogs. J
nine weeks after the third fracture repair, showing healing of the Sm Anim Pract 1974;15:571–8.
fracture in the 1.5-year-old, intact female Italian greyhound from 10. Hunt JM, Aitken ML, Denny HR, Gibbs C. The complications of
Figure 1; lateral angulation is present. diaphyseal fractures in dogs: a review of 100 cases. J Sm Anim Pract
1980;21:103–19.
11. Campbell JR. Healing of radial fractures in miniature dogs. Vet Annual
Based on the results of this study, plate fixation pro- 1980;20:106–12.
vides a successful method of repair of distal radius and 12. Aitola MAO, Sumner-Smith G. Nonunion fractures in dogs. J Vet Orthoped
1984;3:21–4.
ulna fractures in small- and miniature-breed dogs.
13. DeAngelis MP. Causes of delayed union and nonunion of fractures. Vet
Clin N Am 1975;5:251–8.
a 14. Herron MR. Repair of distal radio-ulnar fractures in toy breed dogs. Canine
Synthes (USA); Paoli, PA
Pract 1974;1:12–7.
15. Wilson JW. Vascular supply to normal bone and healing fractures. Sem Vet
Med and Surg 1991;6:26–38.
References 16. Johnson AL, Kneller SK, Weigel RM. Radial and tibial fracture repair with
1. Lappin MR, Aron DN, Herron HL, Malnati G. Fractures of the radius and external skeletal fixation: effects of fracture type, reduction and
ulna in the dog. J Am Anim Hosp Assoc 1983;19:643–50. complications on healing. Vet Surg 1989;18:367–72.
2. Sumner-Smith G, Cawley AJ. Nonunion of fractures in the dog. J Sm Anim 17. DeAngelis MP, Olds RB, Stoll SG, Prata RG, Sinibaldi KR. Repair of
Pract 1970;11:311–25. fractures of the radius and ulna in small dogs. J Am Anim Hosp Assoc
1973;9:436–41.
3. Sumner-Smith G. Bone plating for radial fractures in small dogs. Mod Vet
Pract 1970;3:30–3. 18. Anson LW, DeYoung DJ, Richardson DC, Betts CW. Clinical evaluation
of canine acetabular fractures stabilized with an acetabular plate. Vet Surg
4. Waters DJ, Breur GJ, Toombs JP. Treatment of common forelimb fractures
1988;17:220–5.
in miniature and toy breed dogs. J Am Anim Hosp Assoc 1993;29:442–8.
19. Muir P, Johnson KA, Markel MD. Area moment of inertia for comparison
5. Egger CE. A technique for the management of radial and ulnar fractures in
of implant cross-sectional geometry and bending stiffness. Vet Comp Ortho
miniature dogs using transfixation pins. J Sm Anim Pract 1990;31:377–87.
Traumat 1995;8:146–52.
6. Welch JA, Boudrieau RJ, DeJardin LM, Spodnic GJ. The intraosseous
blood supply of the canine radius: implications for healing of distal
fractures in small dogs. Vet Surg 1997;26:57–61.
7. Vaughan LC. A clinical study of nonunion fractures in the dog. J Sm Anim
Pract 1964;5:173–7.
8. Sumner-Smith G. A comparative investigation into the healing of fractures
in miniature poodles and mongrel dogs. J Sm Anim Pract 1974;15:323–8.
 Surgical Repair of Nasopharyngeal
Stenosis in a Cat Using a Stent
An intraluminal stent was used to maintain patency of a recurrent nasopharyngeal stenosis
(NPS) in a cat. The stenotic membrane within the nasopharynx was resected, and a 2-cm long,
braided-wire endoprosthesis was placed as a stent. The patient was evaluated at one day, six
weeks, 19 weeks, and 49 weeks following surgery. The cat tolerated the stent well. The 19-
week recheck revealed granulation tissue partially obstructing the pharyngeal aspect of the
stent which was subsequently surgically resected. Complications after excision of the
granulation tissue included intermittent upper respiratory congestion and nasal discharge. The
49-week recheck showed no increased granulation tissue; however, upper respiratory
congestion was still present. This particular stent, and its use as described in this paper, is
recommended in cases of chronic recurrent NPS. J Am Anim Hosp Assoc 1999;35:251–6.

Roberto E. Novo, DVM Introduction

Betty Kramek, DVM, MS,
Diplomate ACVS
C flecked nasal discharge, and varying degrees of nasal obstruction. In
Chronic upper respiratory tract disease is common in cats. Causes include
foreign bodies, nasopharyngeal polyps, neoplasia, bacterial and fungal
rhinitis, allergic rhinitis, sinusitis, lymphocytic-plasmacytic pharyngitis
and laryngitis, and nasopharyngeal stenosis (NPS).1 Clinical signs in-
clude snuffling, sneezing, wheezing, gagging, mucopurulent or blood-
severe cases, open-mouth breathing, poor body condition, and difficulty
eating may be present.2
Nasopharyngeal stenosis in cats is rare. All of the previously reported
cases were managed by surgical resection of the stenotic membrane
through an incision in the soft palate.3 However, long-term follow-up of
these cases has not been reported.
The cat in this report was treated with a braided-wire endoprosthesis
after unsuccessful medical and surgical management for NPS. The case
was followed for 30 weeks.

From the Department of Small Animal
Clinical Sciences,
College of Veterinary Medicine,
University of Minnesota,
1365 Gortner Street,
St. Paul, Minnesota 55108.
Address reprint requests to Dr. Novo.
Case Report
An 18-month-old, male domestic shorthair that resided outdoors in a barn,
was presented to the University of Minnesota Veterinary Teaching Hospi-
tal (UMVTH) with a six-week history of wheezing and gagging and a
recent, two-day history of intermittent open-mouth breathing. Prior to
presentation, the cat had been empirically treated for an upper respiratory
tract infection with amoxicillin/clavulanic acida (62.5 mg per os [PO],
bid for 14 days). On physical examination at the UMVTH, the cat was
depressed and appeared underweight (2.9 kg). There was marked inspira-
tory dyspnea with referred upper airway sounds audible on thoracic
auscultation. Palpation of the upper trachea and larynx was resisted.
Results of a complete blood count (CBC) and serum biochemistry profile
were within normal limits for the laboratory. Feline leukemia virus (FeLV)
and feline immunodeficiency virus (FIV) tests were negative. The cat was
premedicated with midazolamb (0.1 mg/kg body weight, intramuscularly
[IM]), ketaminec (6 mg/kg body weight, IM), and oxymorphoned (0.1
mg/kg body weight, IM). General anesthesia was induced with thiopentale

JOURNAL of the American Animal Hospital Association 251

252 JOURNAL of the American Animal Hospital Association May/June 1999, Vol. 35

Figure 1 —Close-up radiographic view of the nasopharyngeal Figure 2—Stenotic opening (5 by 1 mm) of the nasopharynx
area in an 18-month-old domestic shorthair with clinical signs of (black arrow) in the cat from Figure 1.
upper respiratory disease. Close examination reveals a soft-
tissue density across the nasopharynx (black arrow), consistent
with the location of the nasopharyngeal stenosis (NPS).

(20 mg/kg body weight, intravenously [IV]) and main-

tained with isoflurane.f Abnormalities were not noted on
thoracic and skull radiographs. Polyps were not noted on
visual and digital examination of the oral cavity and soft
palate. A 2-cm region of inflamed, proliferative tissue
was found in the right caudal nasal turbinates on rhino-
scopic examination. The left nasal turbinates appeared
unremarkable. Biopsy and histopathological examina-
tion of the proliferative tissue revealed a mixed suppura-
tive and lymphoplasmacytic, subacute to chronic rhinitis.
A diagnosis of chronic rhinitis was made, and treatment
with amoxicillin/clavulanic acid (62.5 mg PO bid) was
continued for two weeks. Figure 3—Wallstent wire-braided endoprosthesis.
The cat responded to medical management with a
decrease in nasal discharge, return of appetite, and in-
creased energy and alertness. However, the inspiratory
wheezes were still present. When the antibiotics were noted. The cat was anesthetized with the same protocol
discontinued, clinical signs returned. Resolution of clini- as previously described. Skull radiographs were repeated.
cal signs occurred within two days after antibiotic therapy A soft-tissue density was present in the region of the
was reinstituted. nasopharynx, approximately 1 cm rostral to the end of
Seven weeks after initial presentation, the cat was the soft palate [Figure 1]. Rhinoscopy was repeated, and
seen again for similar clinical signs. These clinical signs bilateral, inflamed, edematous turbinates were seen.
began one week after finishing a second two-week course There was also stenosis of the nasopharynx at the level
of antibiotics. The cat had been anorexic for two days, of the choanae seen on retroflex view of the nasophar-
vomited intermittently, continued to wheeze and gag, ynx. The stenotic opening consisted of a 5 by 1-mm slit
and had a mucopurulent discharge from both nostrils. [Figure 2]. Polyps were not seen. Based on these find-
Referred upper airway sounds and severe halitosis were ings, a diagnosis of NPS was made.
noted on physical examination. A mild left shift was Surgical correction of the NPS was performed as
seen on CBC (white blood count, 7.8 x103 cells/µl [refer- previously described.3 A midline incision was made in
ence range, 3.8 to 19 x103 cells/µl]; neutrophils, 3.06 the soft palate, exposing the nasopharynx. The nasopha-
x103 cells/µl [reference range, 1.2 to 15.9 x103 cells/µl]; ryngeal opening was enlarged by resecting the stenotic
bands, 1.16 x103 cells/µl [reference range, 0 to 0.19 x103 membrane. The soft palate was then sutured with 4-0
cells/µl]). A complete serum biochemistry profile was Vicryl g in a simple continuous pattern. Postoperative
normal. Abdominal and thoracic radiographs were taken, care included amoxicillin/clavulanic acid (62.5 mg, PO
with marked gas distention of the stomach and intestines bid for 14 days) and prednisoneh (5 mg, PO sid for seven
May/June 1999, Vol. 35
Figure 4—Intraoperative view of the nasopharynx. Two stay
sutures retracting the soft palate to reveal the wire-braided
endoprosthesis within the nasopharyngeal opening.
days; then 2.5 mg, sid for seven days; then 2.5 mg, every
other day [eod] for two doses).
Immediately following extubation, the referred upper
airway sounds heard before surgery were barely audible.
The cat returned to normal activity and appetite within
three days, with no inspiratory dyspnea.
Two weeks following surgery, the cat was seen again
with a history of a gradual increase in inspiratory dysp-
nea and wheezing. On physical examination, upper air-
way sounds were present. Abnormal sounds were not
heard in the lungs on thoracic auscultation. An oral
examination under general anesthesia (same protocol as
previously described) was performed, and restricture of
the previous surgical site in the nasopharynx was visual-
ized. A 10-French Swan Gantz catheteri was used to
bougienage the stricture. The catheter was placed through
the oral cavity and retroflexed into the nasopharyngeal
stricture. Once in place, the balloon was inflated until the
stricture was opened to an 8-mm diameter. Postoperative
medications included amoxicillin/clavulanic acid (62.5
mg, PO bid for 14 days) and prednisone (12.5 mg, PO
bid for three days; then 12.5 mg, sid for three days; then
12.5 mg, eod for three doses).
Nasopharyngeal Stent 253
Figure 5—Lateral radiographic view of the skull, with placement
of the endoprosthesis within the nasopharynx.
The cat was clinically normal for five weeks postbou-
gienage; however, clinical signs of inspiratory dyspnea
and wheezing recurred. Restricture was again suspected.
Surgery was performed to place a stent in the na-
sopharynx to maintain patency. General anesthesia was
induced and maintained as previously described. Access
to the nasopharynx was made through the oral cavity.
Two stay sutures (2-0 Prolene) j were placed in the soft
palate and retracted to allow access to the stenosis. The
stenotic opening was enlarged, and excess tissue was
excised using sharp iris scissors. A 2-cm long Wallstent
wire-braided endoprosthesisk [Figure 3] was placed
within the nasopharynx [Figure 4] with curved Kelly
forceps. The stent was placed such that the distal end did
not extend beyond the soft palate. The distal end of the
stent was attached to the pharyngeal mucosa with a
circumferential simple continuous suture pattern of 6-0
Prolene. Radiographs were taken immediately postop-
erative to evaluate stent placement [Figure 5].
Following surgery the cat showed no inspiratory dysp-
nea and was able to eat and drink normally. The cat was
observed for any possible implant-related problems
(coughing, sneezing, gagging, vomiting, etc.). No prob-
lems were noted during the two-day postoperative hospi-
talization. The cat was discharged on amoxicillin/
clavulanic acid (62.5 mg, PO bid for 10 days).
Reevaluation at six weeks following surgery showed
no evidence of respiratory difficulty. Skull radiographs
revealed no implant migration.
The cat was reexamined 19 weeks postoperatively for
intermittent upper airway noise. Nasal discharge or dysp-
nea were not noted. Evaluation of skull radiographs re-
vealed no implant migration. An oral examination was
performed under anesthesia. The cat was premedicated,
as before, with midazolam and ketamine, and induced
254 JOURNAL of the American Animal Hospital Association
and maintained with isoflurane. There was proliferative
granulation tissue present at the pharyngeal end of the
wire endoprosthesis, creating a narrowing of the lumen.
The excess tissue was excised, and protruding ends of
the wire endoprosthesis were removed. The intraluminal
diameter of the stent was unchanged; however, there was
visible epithelial overgrowth of the wire. It was not
possible to evaluate the nasal aspect of the endoprosthesis
for excessive granulation tissue. The inspiratory noise
was absent when the cat recovered from anesthesia. Oral
prednisone therapy was prescribed as before (12.5 mg,
PO bid for three days; then 12.5 mg, sid for three days;
then 12.5 mg eod for three doses).
The owner was contacted by phone 11 weeks after the
granulation tissue excision. The cat continued to have
episodes of upper respiratory congestion and nasal dis-
charge every six to eight weeks. Oral antibiotic therapy
appeared to resolve the clinical signs.
Recheck examination was performed 30 weeks after
the granulation tissue excision (49 weeks after the initial
stent placement). No nasal discharge was present; how-
ever, inspiratory wheezes were present intermittently.
Intravenous propofoll (6 mg/kg body weight) was used
for anesthesia to allow for adequate oropharyngeal ex-
amination of the cat. Oral examination showed no sig-
nificant proliferation of granulation tissue in the
nasopharyngeal aspect of the stent. The nasal aspect was
not evaluated due to difficulty in visualization. Intermit-
tent oral antibiotic therapy was continued when clinical
signs of nasal discharge and upper respiratory conges-
tion were present.
Discussion
Nasopharyngeal stenosis has been previously described
in four cats.3 The cats in that report and the one presented
here had similar clinical signs: upper respiratory ob-
struction characterized by a whistling or snoring noise
and nasal discharge; signs often aggravated during eat-
ing or swallowing; chronic duration of signs; failure to
respond to medical treatment; and normal respiration
prior to onset of clinical signs. The pathophysiology,
although not definitive, is suggested to be secondary to a
chronic inflammatory stimulus. Feline herpesvirus,
coronavirus, or chlamydial infections are often causative
agents in chronic rhinitis and sinusitis. Severe mucosal
ulcerations are often present in herpesvirus
rhinotracheitis, and it may be that the initiating stimulus
for development of stenotic membranes is mucosal ul-
ceration around the nasopharynx.
The histological changes seen in the four cats with
NPS were similar to those seen in cats with chronic
rhinitis and sinusitis.3 However, neither of these condi-
tions were present to a significant extent. Other differen-
tial diagnoses to be considered in cases of rhinitis and
sinusitis include foreign bodies, nasopharyngeal polyps,
May/June 1999, Vol. 35
intranasal neoplasia, bacterial and allergic rhinitis, and
lymphoplasmacytic pharyngitis and laryngitis.1
Nasopharyngeal cicatrix, a similar condition to NPS,
has been reported in both humans and horses. Schumaker
and Hanselka reported on 47 horses with web-like strands
of scar tissue across the nasopharynx.4 Of these, only
three had signs of respiratory impairment; however, none
were severe enough to require treatment. In humans,
nasopharyngeal cicatrices are described after acute and
chronic ulcerative pharyngitis, both secondary to caustic
burns and infectious diseases.3
Choanal atresia (CA), a congenital disease of hu-
mans,5,6 horses,7 sheep, and llamas, 8,9 has clinical mani-
festations similar to NPS. This results from an
embryological failure of the bucconasal membrane to
rupture, resulting in a lack of communication between
the nasal cavity and nasopharynx. This can be unilateral
or bilateral, complete or incomplete, and bony or mem-
branous. Choanal atresia can result in a partial or com-
plete inability to breathe through the nose.
Diagnosis of NPS involves ruling out other causes
and positively identifying the stenosis by a combination
of caudal rhinoscopy and failure to pass a catheter
through the ventral nasal meatus. Once the lack of patent
nasal passages is established, NPS and CA must be dif-
ferentiated.
In this patient, the diagnosis of NPS versus that of CA
was made based on the anatomical location and appear-
ance of the stenosis. The choanae are paired openings
that lead from the nasal cavities to the nasopharynx.
They are oval in shape, oblique in position, and sepa-
rated by the nasal septum.10,11 In CA, there is a partial or
complete obstruction of one or both of the choanae.
Resection or opening of the membrane in CA results in
visualization of the individual nasal passage. In the pa-
tient of this study, the single membrane was occluding
both nasal passages at a level more caudal to the choa-
nae. There was no obvious distinction of either right or
left sides. When the membrane was resected, the indi-
vidual nasal passages were not visualized.
Nasopharyngeal stenosis has been treated previously
by surgical excision of the stenotic membrane.3 This
involved retraction or incision of the soft palate in order
to expose the stenotic membrane and then perform mem-
brane resection. Restenosis after excision of the mem-
brane was not reported; however, no comment on
long-term evaluation was made.3 In human cases of na-
sopharyngeal cicatrices that have been treated surgically,
there has been a high incidence of restenosis.12
Treatment of CA in humans has involved various
surgical procedures. These include puncture and curet-
tage, transnasal microsurgical drill-out, and laser resec-
tions.4 All have had a high incidence of postoperative
stenosis requiring further intermittent bougienage, tem-
porary intranasal stenting, or both. In animals, treatment
May/June 1999, Vol. 35
has been limited to excision or puncture of the persistent seen in this patient, this procedure can be recommended
membrane, often followed by intranasal stenting with in cases of chronic recurrent NPS.
endotracheal tubes. Restenosis has been common, re-
quiring the stents to maintain lumen patency.7,9 In large a
animals, the current recommendation has been stenting b
for a minimum of 30 days following surgery. c
The cat in this report had recurrent stenosis despite d
e
attempts to reexcise and bougienage the stenosis. Due to
f
the size and activity level of the patient, endotracheal
g
stenting was impossible, and a Wallstent braided wire
h
endoprosthesis was chosen to maintain patency.
i
Braided stents were first used as endovascular pros- j
theses. In addition, they have been used successfully in k
the treatment of urethral strictures, obstructive biliary l
disease, and obstructive airway disease in humans.13–15 m
In one study, stents were placed in the tracheas of piglets
with surgically induced tracheal collapse, resulting in
marked clinical improvement.16 In some cases of ob-
structive airway disease in humans, the braided stents
caused a mild to severe inflammation of the trachea. 16 In References
a few of those cases, tracheal obstruction occurred. In a
recent study evaluating the stents for the treatment of
canine tracheal collapse, tracheal obstruction due to
granulation tissue ingrowth also occurred. m
In the patient described in this report, the braided
stent was selected because of the ease of placement
within the nasopharynx, the ease in determining possible
displacement with radiography, its ideal unconstrained
diameter, and its reported efficacy in treatment of other
stenotic disease. The stent wire diameter is quite small,
and the unconstrained diameter of the prosthesis is 10
mm; thus the nasopharynx can be opened adequately
with minimal stent intraluminal obstruction. It was also
assumed that the small size of the stent would make it
more tolerable for the patient and that a small amount of
granulation tissue ingrowth would inhibit dislodgment. 10.
Unfortunately, as with the human and canine studies,
11.
granulation tissue obstruction also occurred in the pa-
tient of this study. Attempts to inhibit the granulation 12.
tissue with anti-inflammatory doses of glucocorticoids
were unsuccessful. Other types of stents are available 13.
which may be less reactive, thereby decreasing the risk
of postoperative obstruction. Both a woven self-expand- 14.
ing polymeric stent17,18 and a studded silicone stent19,20
have been used experimentally and clinically in dogs and 15.
humans. These stents appear to be less reactive to the
tracheal tissue when compared to the wire stents.m
16.
Nasopharyngeal stenosis is a rare cause of upper
respiratory obstruction, with surgical resection of the stenotic
membrane often being complicated by restricture. Be- 17.
cause of the development of obstructive inflammatory
disease, the use of this braided stainless steel wire
18.
endoprosthesis in the manner described in this study
cannot be recommended for primary treatment of NPS.
However, because of the marked clinical improvement
Nasopharyngeal Stent 255
Clavamox; SmithKline Beecham Animal Health, West Chester, PA
Roche Laboratories, Inc., Nutley, NJ
Phoenix Scientific, Inc., St. Joseph, MO
Endo Pharmaceuticals, Inc., Chadds Ford, PA
Abbott Laboratories, North Chicago, IL
Marsam Pharmaceuticals, Inc., Cherry Hill, NJ
Ethicon, Inc., Cincinnati, OH
Chelsea Laboratories, Inc., Monroe, NC
Edwards-Baxter American, Inc., Santa Ana, CA
Ethicon, Inc., Cincinnati, OH
Schneider U.S. Stent Division; Pfizer Hospital Products Group, Plymouth, MN
Abbott Laboratories, North Chicago, IL
Matthews KG, Caywood DD. Clinical observances on the use of braided
endoprosthesis for the treatment of canine tracheal collapse. Department of
Small Animal Clinical Sciences, College of Veterinary Medicine, University
of Minnesota, St. Paul, MN
1. Noris AM, Laing EJ. Diseases of the nose and sinuses. Vet Clin N Am Sm
Anim Pract 1985;15:865–90.
2. Lane JG. ENT and oral surgery of the dog and cat. Boston: Wright PSG,
1982:65–79.
3. Mitten RW. Nasopharyngeal stenosis in four cats. J Sm Anim Pract
1988;29:341–5.
4. Schumacher J, Hanselka DV. Nasopharyngeal cicatrices in horses: 47 cases
(1972–1985). J Am Vet Med Assoc 1987;191:239–42.
5. Brown OE, Pownell P, Manning SC. Choanal atresia: a new anatomic
classification and clinical management applications. Laryngoscope
1996;106:97–101.
6. Menasse-Palmer L, Bogdanon A, Marion RW. Choanal atresia. Pediatr Rev
1995;16:475–6.
7. Goring RL, Campbell M, Hillidge CJ. Surgical correction of congenital
bilateral choanal atresia in a foal. Vet Surg 1984;13:211–6.
8. Fenwick BW, Fowler M, Kock M, Matern L. Complete choanal atresia in a
llama. J Am Vet Med Assoc 1982;181:1409–10.
9. Kiorpes AL, Lindsay WA, Adams WM. Repair of complete choanal atresia
in a llama. J Am Vet Med Assoc 1986;189:1169–71.
Dyce KM, Sack WO, Wensing CJG, eds. The respiratory apparatus. In:
Textbook of veterinary anatomy. Philadelphia: WB Saunders, 1987:143–7.
Evans HE, Christenson GC, eds. The respiratory apparatus. In: Miller’s
anatomy of the dog. 2nd ed. Philadelphia: WB Saunders, 1979:507–26.
Stevenson EW. Cicatrical stenosis of the nasopharynx. Laryngoscope
1969;79:2035–67.
Sigwart U, Puel J, Mirkovitch V, et al. Intravascular stents to prevent
occlusion and restenosis after transluminal angioplasty. N Engl J Med
1987;316:701–6.
Milroy EJG, Chapple CR, Eldin A, Wallsten H. A new stent for the
treatment of urethral strictures—preliminary report. Br J Urol
1989;63:392–6.
Gillams A, Dick R, Dooley JS, Wallsten H, El-Din A. Self-expandable
stainless steel braided endoprosthesis for biliary strictures. Radiology
1990;174:137–40.
Mair EA, Parsons DS, Van Dellen AF, Lally KP. Endotracheal stents for
the treatment of induced tracheomalacia in piglets. Am Coll Surgeons Surg
Forum 1989;40:569–71.
Charnsangavej C, Carrasco CH, Wright KC, Richli W, Wallace S,
Gianturco C. A new expandable metallic stent for dilatation of stenotic
tubular structures: experimental and clinical evaluation. Houston Med J
1987;3:41–51.
Tsang V, Williams AM, Goldstraw P. Sequential silastic and expandable
metal stenting for tracheobronchial strictures. Ann Thorac Surg
1992;53:856–60.
(Continued on next page)
256 JOURNAL of the American Animal Hospital Association May/June 1999, Vol. 35

References (cont’d) 20. Mason RA, O’Grady MR. Preliminary evaluation of the Dumon dedicated
tracheobronchial stent in the canine airway via pulmonary function testing.
19. Dumon JF. A dedicated tracheobronchial stent. Chest 1990;97:328–32. Champaign, IL: Proceedings 11th Veterinary Respiratory Symposium, 1992.