A Comparison of Transdermal Fentanyl

Versus Epidural Morphine for Analgesia

in Dogs Undergoing Major
Orthopedic Surgery
Postoperative analgesia provided by transdermal fentanyl was compared with that provided by
epidural morphine in dogs undergoing major orthopedic surgery. Dogs randomly were assigned
to receive either a 100 µg per hour transdermal fentanyl patch 24 hours prior to surgery (n=8)
or epidural morphine (0.1 mg/kg body weight) administered following induction of anesthesia
(n=10). Temperature, heart rate, respiratory rate, and pain score were recorded prior to surgery
and zero, six, 18, 30, and 42 hours after surgery. Blood samples were collected from the dogs
in the transdermal fentanyl group beginning 24 hours preoperatively to 42 hours
postoperatively. Fentanyl concentrations were determined by radioimmunoassay. When all time
periods after surgery were combined, dogs in the transdermal fentanyl group were experiencing
significantly less pain after surgery than dogs given epidural morphine. The transdermal
fentanyl provided analgesia after major orthopedic surgery greater than or equivalent to that of
epidural morphine. J Am Anim Hosp Assoc 1999;35:95–100.

Timothy M. Robinson, DVM Introduction

Kris T. Kruse-Elliott, DVM, PhD
Mark D. Markel, DVM, PhD
G. Elizabeth Pluhar, DVM, MS
Kathy Massa, BA
Dale E. Bjorling, DVM, MS
O not been recommended for acute postoperative pain control in humans
From the Departments of Surgical Sciences
(Robinson, Kruse-Elliot, Bjorling) and
Medical Sciences (Markel, Pluhar),
School of Veterinary Medicine,
University of Wisconsin,
Madison, Wisconsin 53706-1102.
Transdermal drug delivery systems offer significant advantages over
traditional routes of drug delivery because of ease of administration,
noninvasiveness, and continuous drug delivery over an extended period
of time. 1–4 Transdermal delivery systems for a variety of compounds,
including nicotine, estrogen, testosterone, clonidine, and several other
drugs, have been utilized in humans.5,6 The transdermal fentanyl patch
has proven effective in humans for management of pain associated with
cancer and other chronic disease conditions. 1,7–11 This system is espe-
cially suited to patients with relatively stable baseline pain levels with
little short-term fluctuation in pain intensity. 1 Transdermal fentanyl has
because of concerns regarding respiratory depression, slow onset of ac-
tion, lack of control in response to rapid changes in pain, and variability
of absorption.12
Due to the short plasma half-life of fentanyl after intravenous (IV)
administration, the postoperative use of fentanyl has been limited to either
frequent IV boluses or continuous IV infusion. This causes peak and
trough variations in blood concentrations commonly associated with in-
termittent drug administration.1,2,6,11 Interest in control of postoperative
pain in animals has lead to recent studies analyzing pharmacokinetics of
transdermal fentanyl delivery in dogs13 and comparing transdermal fenta-
nyl to parenterally administered oxymorphone for control of pain follow-
ing ovariohysterectomy in dogs.14 Control of postoperative pain in animals
is often complicated by the temperament of the animal and the short
duration of analgesic action, and these, as well as other factors, may lead
to extended hospitalization. The current study was undertaken to compare
transdermal fentanyl with epidural morphine for control of postoperative

JOURNAL of the American Animal Hospital Association 95

96 JOURNAL of the American Animal Hospital Association
Figure 1—The typical construction of a fentanyl patch is shown.
pain after major orthopedic surgery in dogs. The hypoth-
esis tested was that transdermal fentanyl would provide
analgesia in the postoperative period comparable to that
obtained with a single epidural dose of morphine.
Materials and Methods
The experimental protocol was approved by the Univer-
sity of Wisconsin Research Animal Care and Use Com-
mittee. Eighteen adult mixed-breed dogs weighing 19 to
31.5 kg (mean±standard deviation [SD], 24±3.0 kg) un-
dergoing experimental unilateral total hip arthroplasty
and proximal femoral allograft implantation were in-
cluded in this study. Each animal was randomly assigned
to one of two pain management groups: dogs in the
transdermal fentanyl group received a 100 µg per hour
transdermal fentanyl patcha 24 hours prior to surgery,
and dogs in the epidural morphineb group received 0.1
mg/kg body weight of morphine in the epidural space
immediately following anesthetic induction. Dogs were
premedicated with acepromazine (0.05 mg/kg body
weight, intramuscularly [IM]) and oxymorphone (0.1
mg/kg body weight, IM); anesthesia was induced with
sodium thiopentothal (5 mg/kg body weight, IV) and
maintained with halothane in oxygen. Dogs in the
transdermal fentanyl group included eight dogs (five
males and three females) weighing 23.2 kg±2.9 kg
(mean±SD). Dogs in the epidural morphine group in-
cluded 10 dogs (two males and eight females) weighing
24.8 kg±3.1 kg. The transdermal fentanyl patch was
applied to a clipped area between the scapulae. The area
was cleaned with chlorhexidine scrub and allowed to dry
completely before transdermal fentanyl application. The
transdermal fentanyl patch was protected with a light
bandage. c The transdermal fentanyl patch used in this
study is constructed as shown in Figure 1. The patches
are rectangular and translucent, with a protective peel
backing and four functional layers. Silicone adhesive
containing an initial dose of fentanyl is adhered to the
skin directly. An ethylene-vinyl acetate co-polymer mem-
brane controls the rate of fentanyl release from the drug
reservoir where fentanyl and alcohol are gelled with
hydroxyethyl cellulose. A layer of polyester protects the
March/April 1999, Vol. 35
exterior of the patch from mechanical damage and desic-
cation. Patches are manufactured in four sizes (i.e., 25,
50, 75, and 100 µg per hr), and the rate of fentanyl
delivery per hour from each system is proportional to the
surface area of the patch itself (10, 20, 30, or 40 cm2,
respectively). Fentanyl is delivered at a rate of 25 µg per
hour per 10 cm2 regardless of patch size.12 The trans-
dermal fentanyl patch was removed 72 hours after appli-
cation. In dogs from the epidural morphine group,
morphine was administered epidurally immediately fol-
lowing induction of anesthesia and prior to surgical
preparation. Morphine (0.5 mg/ml solution) was injected
epidurally at the lumbosacral space with the dogs in
lateral recumbency. The proximal one-third of the femur
was surgically removed on one side, and each dog under-
went proximal femoral allograft/endoprosthetic surgery
with reattachment of host to allograft tissues by either
tendon to tendon, trochanteric osteotomy, or bicortical
half-shell fixation as described elsewhere.15 Dogs had ad
libitum access to water and commercial dry dog food for
the duration of the study, except during the standard 12-
hour preoperative fasting period.
Peripheral venous blood (5 ml) was collected from
dogs in the transdermal fentanyl group by cephalic veni-
puncture into a 5-ml heparinized tube 24, 18, and 12
hours prior to surgery; just prior to induction of anesthe-
sia; and six, 18, 30, and 42 hours after surgery. Samples
were centrifuged immediately at 2,000 revolutions per
minute for 10 minutes. Plasma was removed and placed
in 2-ml vials labeled with dog number, time of collec-
tion, and date, and was stored at -70˚ C until analysis.
Plasma samples were shipped on dry ice to North Caro-
lina State University for determination of fentanyl con-
centration by radioimmunoassay (RIA) as previously
described.16
Twenty-four, 12, and zero (immediately prior to anes-
thesia) hours prior to surgery and zero (immediately
after surgery), six, 18, 30, and 42 hours after surgery the
temperature, heart rate, respiratory rate, and pain score
were recorded for all dogs. Pain score was determined
for all dogs using the pain scoring system as shown in
the Appendix.17,18 Pain scoring was performed by only
two individuals (Robinson, Massa) who were familiar
with the pain scoring system. Due to the obvious differ-
ence in appearance of dogs in the fentanyl patch group,
the individuals performing the pain scoring were not
blinded to the treatment protocol. Any animal that ac-
crued a combined pain score greater than five was given
oxymorphone (0.1 mg/kg body weight, IM) to provide
supplemental analgesia. Negative controls (i.e., dogs not
receiving transdermal fentanyl or epidural morphine)
were not included for humane reasons.
Differences between the two groups were analyzed
utilizing one-way analysis of variance (ANOVA) for
repeated measures. All differences were considered to be
significant at a probability level of 95% (p of 0.05 or
March/April 1999, Vol. 35 Analgesia 97

Appendix
Pain Score Criteria17
Vocalization Movement
Quiet 0 None 0
Vocalizing, responds to calm voice/stroking 1 Frequent position changes 1
Vocalizing, no response to calm voice/stroking 2 Thrashing 2
Agitation Heart Rate (HR)
Patient asleep or calm 0 <10% above preoperative HR 0
Mild agitation 1 10%–20% above preoperative HR 1
Moderate agitation 2 20%–30% above preoperative HR 2
Hysterical 3 >30% above preoperative HR 3
Respiratory Rate (RR)
<10% above preoperative RR 0
10%–20% above preoperative RR 1
20%–30% above preoperative RR 2
>30% above preoperative RR 3

1.5
Table 1 supplemental analgesia. The dog given sedation for ra-
diographs was included in the study up to the time when
Plasma Fentanyl

1 0.95 0.92 the drug was given. One other dog in the transdermal
(ng/ml)

fentanyl group was eliminated from the study because,

0.5 Surger
0
-24 -18 -12 -6 0 6 12
Hours
Figure 2—Mean plasma fentanyl concentrations attained during
the application period.
less). All statistical analyses were performed with a com-
mercially available software program.d
Results
The mean surgical times for the transdermal fentanyl and
epidural morphine groups were 4.5 and 5.1 hours, re-
spectively. There was no difference in anesthetic re-
quirements between the two groups. Four dogs (three in
the transdermal fentanyl group and one in the epidural
morphine group) were given supplemental postoperative
analgesia. Two dogs in the transdermal fentanyl group
each were given one dose of oxymorphone during recov-
ery from anesthesia. These dogs began vocalizing and
thrashing upon extubation and were given supplemental
analgesia immediately. The other dog in the transdermal
fentanyl group inadvertently was given one dose of
acepromazine and butorphanol 24 hours after surgery.
Luxation of the coxofemoral prosthesis was suspected in
this dog, and sedation was given to take radiographs.
One dog in the epidural morphine group was given one
dose of butorphanol 12 hours after surgery. Data was
analyzed by eliminating all data collected after adminis-
tration of supplemental analgesia and by including data
collected more than 12 hours after administration of
although the transdermal fentanyl system was applied 24
hours prior to surgery, surgery on this dog was post-
18 24 30 36 42 48 poned to a later date. This left 17 dogs (seven in the
transdermal fentanyl group and 10 in the epidural mor-
phine group) for statistical analysis.
All transdermal fentanyl patches remained tightly ad-
hered to the dogs, and there was no evidence of discom-
fort or skin reaction associated with the patches. Dogs
given transdermal fentanyl exhibited no overt signs of
sedation. All dogs given epidural morphine were noted
to be moderately sedated postoperatively.
Mean plasma fentanyl concentrations increased
steadily during the first 24 hours after patch application,
and they reached steady state between 24 (time 0) and 36
hours after application and remained relatively stable
until patch removal [Figure 2]. The highest mean plasma
fentanyl concentrations (0.95 ng/ml±0.14 ng/ml and 0.92
ng/ml±0.11 ng/ml, mean±standard error of the mean
[SEM], respectively) occurred 36 and 48 hours after
patch application. Although the same size transdermal
fentanyl patch was applied to each dog, individual plasma
fentanyl concentrations varied widely (i.e., 0.40 to 1.28
ng/ml at 24 hrs). Hemolysis was observed in three
samples, which may artificially increase measured con-
centrations of fentanyl; therefore, these samples were
not included in the final data analysis.
Mean pain scores were highest for both groups during
the first 12 hours postoperatively, and supplemental an-
algesia was most often administered during this time
[Figure 3]. Mean pain scores for dogs in the transdermal
98 JOURNAL of the American Animal Hospital Association March/April 1999, Vol. 35

5 200

4
150
Pain Score

Heart rate
3

Rate
100
2

50 Respiratory
1

0 0
0 6 12 18 24 30 36 42 48 0 6 12 24 36 48
Transdermal Fentanyl
Transdermal Fentanyl
Hours Epidural Morphine
Hours Epidural Morphine Series4

Figure 3—Mean pain scores as compared between the epidural Figure 4—The comparison of mean heart and respiratory rates
morphine and transdermal fentanyl groups. between groups.

fentanyl group consistently decreased after the first 12- 110

F)
hour postoperative period, whereas a second trend to-

(degrees
105

ward increased mean pain scores occurred in dogs of the

100
epidural morphine group 36 hours postoperatively.

Temperature
No significant differences were identified between 95

mean pain scores for the two groups at any single time 90
period. When pain scores for all time intervals in each 0 6 12 24 36
Transdermal Fentanyl
48
Hours
group were compared by one-way ANOVA for repeated Epidural Morphine

measures, an overall significant difference in pain score
was detected between dogs in the transdermal fentanyl
group and those dogs given epidural morphine (p less
than 0.05). This finding was consistent whether data was
analyzed including or excluding data from dogs given
supplemental postoperative analgesics. There were no
significant differences between the groups with regard to
temperature, heart rate, or respiratory rate [Figures 4, 5].
Discussion
Data was analyzed in two ways, either including time
periods after supplemental postoperative analgesics were
given, or excluding those time periods. Eliminating those
dogs given supplemental postoperative analgesia, only
four dogs in the transdermal fentanyl group were left for
statistical analysis at several time periods, thereby low-
ering the statistical power. However, whether or not pain
scores from dogs given supplemental postoperative anal-
gesia are included, there remains an overall significant
difference between the two groups. When all time peri-
ods were combined, dogs in the transdermal fentanyl
group were experiencing significantly less pain than dogs
in the epidural morphine group (p less than 0.05), al-
though no significant differences were present at indi-
vidual time periods. Overall, dogs in the fentanyl patch
group experienced less pain than their morphine epidural
group counterparts.
The present study was performed using dogs that
were part of another study investigating various methods
of attachment of proximal femoral allografts. Therefore,
the goals of the larger study did not always coincide with
the goals of this study, and control of supplemental
analgesic administration was not optimal, especially in
two dogs which received postoperative analgesics prior
to exceeding the pre-established maximum allowable
pain score. Individuals not familiar with the method of
pain scoring and concerned about the postoperative re-
Figure 5—The comparison of mean body temperature between
groups.
covery of these dogs immediately administered analge-
sics. Dogs in the transdermal fentanyl group seemed to
show a tendency to vocalize more and appeared less
sedate after surgery than dogs in the epidural morphine
group. A similar tendency also has been noted in human
cancer patients experiencing periods of increased pain
which may require “rescue” doses of additional analge-
sics.1 As reported, the mean time of surgery for both
groups exceeded 4.5 hours. Plasma concentrations of
premedications would be negligible after surgery, and in
combination with anesthetic recovery and the degree of
surgical manipulation, all probably contributed to the
observed pain scores and need for additional analgesics
(rescue doses) six hours after induction of anesthesia.
Additionally, plasma fentanyl levels did not peak until
12 hours after surgery. A recent study comparing plasma
fentanyl concentrations with varying fentanyl patch sizes
in normal dogs found that plasma fentanyl concentra-
tions may take at least 24 hours to reach steady state and
that interindividual and intraindividual variability was
high.19 The present study found similar results which
also may explain the variability in pain score in the early
postoperative period. Based on these findings, the au-
thors recommend applying the transdermal fentanyl patch
at least 24 hours and possibly 36 hours prior to surgery.
The authors also recommend that supplemental opioids
be given prior to recovery from anesthesia in dogs that
have received transdermal fentanyl who have undergone
major orthopedic procedures. The opioid dosage should
be adjusted depending on the anesthetic protocol and the
degree of arousal at the time of administration of addi-
tional drug.
Transdermal fentanyl patches which deliver 100 µg
of fentanyl per hour were chosen for this study based on
March/April 1999, Vol. 35
the weight of the dogs, empirical clinical experience, and
previous pharmokinetic studies in dogs and cats.13,20 The
patches were applied in a uniform manner to the inter-
scapular region. No problems with patch application were
noted. In these dogs, the patches were held in place with
a bandage encircling the thorax. From clinical experi-
ence, the patches are more conveniently held in place if a
six-inch adherent bandagee is applied over them. Prior to
placing the transdermal fentanyl patch, the site of appli-
cation should be clipped closely, cleaned, and dried to
assure optimal skin adherence.
In humans, the 100 µg per hour transdermal fentanyl
patch yielded a maximum serum concentration of 1.9 to
3.8 ng/ml in 24 to 72 hours. 12 A 50 µg per hour
transdermal fentanyl patch yielded a steady state plasma
concentration of 1.6 ng/ml in dogs averaging 13.5 kg±1.9
kg.13 In the authors’ study, the mean steady state plasma
concentration was 0.95 ng/ml in dogs averaging 23.2
kg±2.9 kg. The effective analgesic plasma concentration
in humans has been reported to be 0.9 to 2.0 ng/ml.6,21
The plasma concentration of fentanyl which provides
analgesia consistently in dogs is yet to be determined.
However, this study indicates analgesia is achieved in
dogs with plasma fentanyl concentrations of 0.95 ng/ml,
which is within the range reported for humans.
In humans, a significant amount of variation exists
both in the apparent analgesic serum fentanyl concentra-
tion and in the rate of fentanyl delivery detected with the
same size transdermal fentanyl patch applied to similar
size individuals. 1,4–6,10,22 Similar variation was noted in
the present investigation. Several dogs with relatively
low plasma fentanyl concentrations had low pain scores
after surgery. On the other hand, one dog with very low
plasma fentanyl concentration (0.40 ng/ml) in the early
postoperative period did require additional analgesia.
Even though the same size transdermal fentanyl patch
was applied to each dog, the plasma fentanyl concentra-
tions varied widely. Several factors may lead to this
variation. In humans, body and skin temperature, skin
thickness and permeability, local skin circulation, sys-
tem adherence, hydration status, sweat gland function,
ethnic group, and the state and integrity of the stratum
corneum all are known to affect the rate of transdermal
fentanyl delivery.6,23 Several factors could alter the per-
formance of the transdermal fentanyl patch when applied
to dogs. The higher body temperature of dogs (relative to
humans) presumably would increase the uptake of fenta-
nyl through the skin if the skin temperature also was
higher. Skin adhesion and the absorption of fentanyl may
vary widely among individual dogs or between breeds
due to hair coat and skin thickness characteristics. The
differences noted in this study could be due to differ-
ences in patch adherence or individual variation in the
characteristics of the skin at the site of application. These
differences may account for the increased time required
to reach steady state serum fentanyl concentrations ob-
Analgesia 99
served in this study as compared to humans. At the time
of patch removal, all the patches appeared well adhered
to the skin and were not dislodged easily from the site of
application. Body temperature was monitored, though
local skin temperature was not measured. All the dogs
were housed under standard laboratory conditions with
strictly controlled temperatures, and no differences in
skin temperature were expected. During surgery, the
dogs were in lateral recumbency, and thus not lying
directly on the patch which might be expected to influ-
ence absorption. No trends indicating a correlation be-
tween increased body temperature and increased plasma
fentanyl concentrations were detected.
The development of the transdermal fentanyl patch
was stimulated by the need for an easy, noninvasive,
cost-effective way to deliver narcotic analgesia to hu-
mans suffering from low-grade, prolonged pain associ-
ated with cancer.1,4 Fentanyl was selected over morphine
and other opioid analgesics by nature of its physico-
chemical properties. Fentanyl is a synthetic opioid anal-
gesic. It is estimated to be 80 to 300 times more potent
than morphine, depending on species, and has an analge-
sic therapeutic index approximately four times that of
morphine.24 Fentanyl has a low molecular weight and is
highly lipid soluble. Additionally, fentanyl was found to
be nonirritating on direct contact with skin and did not
promote long-term hypersensitivity in humans.4 All of
these factors make fentanyl an ideal narcotic for incorpo-
ration into a transdermal delivery system.
Fentanyl interacts strongly with µ-opioid receptors
which are located throughout the central nervous system
(CNS). The primary therapeutic effects of fentanyl are
sedation and analgesia. In humans, pain tolerance and
pain perception are altered; however, pain still may be
recognized. Like other narcotics, fentanyl can cause res-
piratory depression. Approximately 4% of human pa-
tients treated for postoperative pain with transdermal
fentanyl exhibited respiratory depression; hence, trans-
dermal fentanyl is not recommended currently for con-
trol of postoperative pain in humans. 12 Although
respiratory function was not investigated specifically, no
apparent outward clinical signs of respiratory depression
were noted in the dogs treated with transdermal fentanyl.
Based on the findings of the authors’ study, the
transdermal fentanyl patch is an acceptable mode of
administration of postoperative analgesia after major or-
thopedic surgery in the dog. Its ease of administration
and comparable cost make it an excellent option for
postoperative pain control. The use of transdermal fenta-
nyl in dogs is an extralabel use of a product designed for
humans, and owners must therefore be fully aware of the
risks of its application and use in their pet. Transdermal
fentanyl patches should not be cut at any time as this can
result in accidental human exposure. For animals that
require a smaller surface area, the authors recommend
covering the release membrane to reduce the surface
100 JOURNAL of the American Animal Hospital Association March/April 1999, Vol. 35

area exposed to the skin rather than cutting the patches. 13.
Releasing pets from the hospital with transdermal fenta- 14.
nyl patches in place may not be advisable due to the
potential for intentional or accidental removal, possible
15.
exposure of children, and concerns regarding disposal of
the patches. In addition, local laws or restrictions may
prohibit releasing animals with transdermal fentanyl 16.
patches.
Within the context of this study, transdermal fentanyl 17.
provided analgesia in dogs after major orthopedic sur-
gery greater than or equivalent to that provided by a 18.
single epidural injection of morphine. Further basic and
clinical research is indicated to establish the therapeutic
19.
plasma fentanyl level in the dog, and the use of
transdermal fentanyl should be investigated in a more
diverse clinical population of orthopedic patients. 20.
21.
a pharmacokinetics and pharmacodynamics of constant-rate I.V. and
Duragesic; Janssen Pharmaceutica, Titusville, NJ
b transdermal delivery. Br J Anaesth 1988;60:608–13.
Astramorph/PF; Astra USA, Westborough, MA
c 22.
Elasticon; Johnson & Johnson, New Brunswick, NJ
d
SAS Institute, Cary, NC
e 1989;37:193–202.
Tegaderm; 3M Corporation, Irvine, CA
23.
Acknowledgments
The authors wish to thank Dr. Mark Papich for perform- 24.
ing the fentanyl RIA and the University of Wisconsin
School of Veterinary Medicine Companion Animal Fund
for funding this project.
Kyles AE, Papich M, Hardie EM. Disposition of transdermally adminis-
tered fentanyl in dogs. Am J Vet Res 1996;57(5):715–9.
Kyles AE, Hardie EM, Hansen B, Papich M. Comparison of transdermal
fentanyl and intramuscular oxymorphone after ovariohysterectomy in dogs.
Vet Surg 1996;25(5):431–2.
Markel MD, Wood SA, Bogdanske JJ, et al. Comparison of healing of
allograft/endoprosthetic composites with three types of gluteus medius
attachment. J Orthop Res 1995;13:105–14.
Michiels M, Hendriks R, Heykants J. A sensitive radioimmunoassay for
fentanyl: plasma levels in dogs and man. Eur Clin Pharmacol 1977;12:
153–8.
Conzemius MG, Brockman DJ, King LG, Perkowsky SZ. Analgesia in
dogs after intercostal thoracotomy: a clinical trial comparing intravenous
buprenorphine and interpleural bupivacaine. Vet Surg 1994;23:291–8.
Thompson SE, Johnson JM. Analgesia in dogs after intercostal thora-
cotomy: a comparison of morphine, selective intercostal nerve block, and
intrapleural regional analgesia with bupivacaine. Vet Surg 1991;20:73–7.
Egger CM, Duke T, Archer J, Cribb PH. Comparison of plasma fentanyl
concentrations by using three transdermal fentanyl patch sizes in dogs.
Vet Surg 1998;27(2):159–66.
Scherk-Nixon M. A study of the use of a transdermal fentanyl patch in cats.
J Am Anim Hosp Assoc 1996;32:19–24.
Holley FO, Van Steennis C. Postoperative analgesia with fentanyl:
Gourlay GK, Kowalski SR, Plummer JL, Cherry DA, Gaukroger P,
Cousins MJ. The transdermal administration of fentanyl in the treatment of
postoperative pain: pharmacokinetics and pharmacodynamic effects. Pain
Gupta SK, Southam M, Gale R, Hwang SS. System functionality and
physicochemical model of fentanyl transdermal system. J Pain Symptom
Manage 1992;7(3):S17–25.
Stanley TH. The history and development of the fentanyl series. J Pain
Symptom Manage 1992;7(3):S3–7.

References
1. Payne R. Transdermal fentanyl: suggested recommendations for clinical
use. J Pain Symptom Manage 1992;7(3):S40–4.
2. Duthie DJR, Rowbotham DJ, Wyld R, Henderson PD, Nimmo WS. Plasma
fentanyl concentrations during transdermal delivery of fentanyl to surgical
patients. Br J Anaesth 1988;60:614–8.
3. Shaw JE, Urquhart J. Programmed, systemic drug delivery by the
transdermal route. Trends Pharm Sci 1980;1:208–11.
4. Lehmann KA, Zech DFJ. Transdermal fentanyl: clinical pharmacology.
J Pain Symptom Manage 1992;7(3):S8–15.
5. Sandler A. Transdermal fentanyl: acute analgesic clinical studies. J Pain
Symptom Manage 1992;7(S3):S27–35.
6. Varvel JR, Shafer SL, Hwang SS, Coen PA, Stanski DR. Absorption
characteristics of transdermally administered fentanyl. Anesthesiology
1989;70:928–34.
7. Maves TJ, Barcellos WA. Management of cancer pain with transdermal
fentanyl:phase IV trial, University of Iowa. J Pain Symptom Manage
1992;7(S3):S58–62.
8. Simmonds MA, Richenbacher J. Transdermal fentanyl: long-term analgesic
studies. J Pain Symptom Manage 1992;7(3):S36–9.
9. Patt RB, Hogan LA. Transdermal fentanyl for chronic cancer pain: detailed
case reports and the influence of confounding factors. J Pain Symptom
Manage 1992;7(3):S51–4.
10. Zech DFJ, Grond SUA, Lynch J, Dauer HG, Stollenwerk B, Lehmann KA.
Transdermal fentanyl and initial dose findings with patient-controlled
analgesia in cancer pain. A pilot study with 20 terminally ill cancer
patients. Pain 1992;50:293–301.
11. Miser AW, Narang PK, Dothage JA, Young RC, Sindelar W, Miser JS.
Transdermal fentanyl for pain control in patients with cancer. Pain 1989;
37:15–21.
12. Product insert. Duragesic (fentanyl transdermal system). Jannsen
Pharmaceutica, Titusville, NJ, January 1994.
 A Retrospective Study of Canine House
Soiling: Diagnosis and Treatment
A retrospective study was conducted to determine the relative frequency and type of elimination
problem seen in dogs at a university referral practice and to evaluate the efficacy of the
suggested treatments. Cases presented to the Animal Behavior Clinic at Cornell University
between 1987 and 1996 were reviewed. Of 1,173 cases, 105 (9%) were house-soiling cases.
Of these cases, the authors obtained outcome information from 70. Within the diagnosis of
house soiling, incomplete housebreaking (n=59; 84%) were the most frequent referral cases, of
which 48 cases (81%; 95% confidence interval, 69% to 90%) improved. Separation anxiety was
considered the second most common underlying cause (n=27; 39%), of which 85% (n=23; 95%
confidence interval, 66% to 96%) improved. Behavior modification was the most often
suggested treatment (n=58), with 48 (83%) cases improving. Behavior modification consisted of
accompanying the dog to the preferred elimination area, rewarding the dog for eliminating
there, and punishing the dog only when caught in the act of house soiling. These results
suggest that correct house training, behavior modification involving positive reinforcement, and
appropriate punishment are essential to diminish house-soiling problems in dogs.
J Am Anim Hosp Assoc 1999;35:101–6.

Seong C. Yeon, DVM, PhD Introduction

Hollis N. Erb, DVM, PhD
Katherine A. Houpt, VMD, PhD,
Diplomate ACVB
RS
From the Animal Behavior Clinic (Yeon,
Houpt) and Department of Epidemiology
(Erb), College of Veterinary Medicine,
Cornell University,
Ithaca, New York 14853.
Address all correspondence to Dr. Yeon.
Dogs that eliminate in the house present a problem to their owners. It is a
serious strain on the pet-owner bond, because if the behavior cannot be
stopped, the owner may relegate the dog to an outdoor pen or surrender it
to an animal shelter. Elimination in the house (i.e., “house soiling”) is one
of the most frequent behavior problems following aggression.1 When the
owner presents a dog with a complaint of elimination behavior, the
clinician must determine if the dog has a medical problem and if it is
urinating, defecating, or both. 2 Differential diagnosis for inappropriate
elimination behavior and determination of the specific underlying etiol-
ogy are essential for successful management. Canine elimination disor-
ders are classified into 11 types: a medical condition, incomplete
housebreaking, insufficient access to appropriate elimination areas, sub-
strate preference, separation anxiety, marking behavior, submissive uri-
nation, excitement urination, elimination associated with fear,
attention-seeking behavior, and canine cognitive dysfunction. 3 Numerical
or epidemiological data has not been presented for inappropriate canine
elimination disorder.
The purposes of this retrospective study were to determine the relative
frequency of this behavior problem in a university referral behavioral
practice and to identify the type or cause of the house soiling and the
efficacy (“owner reported improved”) of the suggested treatments.
Methods
Cases presented to the Animal Behavior Clinic at Cornell University
between 1987 and 1996 were reviewed. Of 1,173 behavior-related cases,
105 (9%) were house-soiling cases. During the consultation, each owner
was provided with a diagnosis and suggestions for correcting the prob-

JOURNAL of the American Animal Hospital Association 101

102 JOURNAL of the American Animal Hospital Association
lem. Following the initial consultation, the owner re-
ceived written recommendations, a copy of which was
sent to the referring veterinarian. A self-addressed,
stamped, follow-up postcard was sent six months after
the initial consultation. The owner could check one of
four outcome categories: cured, better, the same, or
worse. This was used to determine the overall success
(i.e., improvement) rate of the consultation. In this study,
the “cured” and “better” outcomes were categorized as
the “improved” group. The “same” and “worse” out-
comes were categorized as the “not improved” group. If
the owners did not respond to the follow-up postcard,
they were called between six months and 10 years after
the initial consultation. Of these 105 cases, 35 were lost
to follow-up.
The dogs were divided into four groups by age: puppy
(one to six months of age), adolescent (seven to 12
months of age), young adult (13 to 24 months of age),
and mature adult (25 months of age or older). The sex
and neuter status of the dog was recorded. The type of
housing (house or apartment) was recorded. House soil-
ing is often part of the complex of problems seen during
the owner’s absence and termed separation anxiety; there-
fore, frequency of the separation anxiety problem was
also calculated. Before using behavior therapy, any pos-
sible medical cause for house soiling was ruled out.
Diagnoses used in this clinic are: incomplete house-
breaking, a medical problem, separation anxiety, mark-
ing behavior, excitement-induced urination, submissive
urination, fear-associated elimination, and aggression-
associated elimination.
The original suggestions consisted of one or more of
the following: taking the dog outside more frequently
and accompanying him rather than leaving him alone
outside, behavior modification, changing the environ-
ment (i.e., using a different area for restraint or walking),
changing food and water intake schedules, paper train-
ing, crate training, using a rug deodorizer, umbilical cord
training, obedience training, more frequent exercise,
medical treatment (including neutering and drug treat-
ment), using a head halter,a and use of a diary to record
the number of inappropriate eliminations.
Behavior modification was a frequently suggested
treatment. Behavior modification included positive rein-
forcement (i.e., praising the dog profusely when it uri-
nates or defecates outside and immediately giving it a
very palatable treat, such as freeze-dried liver); never
punishing the dog except when it is caught in the act;
desensitization (used in house soiling caused by separa-
tion anxiety); and counter-conditioning (used in submis-
sive urination) by teaching the dog the “stand-stay”
command. Standing is a less submissive posture than
sitting or lying down. It is taught using the food refusal
exercise: showing the food to the dog and placing it on
an upside-down bowl or other prominent surface so the
dog is focused on the treat rather than on the owner and
March/April 1999, Vol. 35
then telling the dog to “stay,” releasing his lead, and
saying “OK” after an interval. The stay intervals gradu-
ally are increased. This training can minimize submis-
sive urination. When the dog stands, praise and giving of
the treat can immediately follow.
Desensitization treatment of separation anxiety has
been discussed in detail elsewhere. 3 Briefly, the owner
should practice going through the routine of leaving
without actually exiting the house. This may involve
picking up personal bags and keys, closing drapes, turn-
ing off the lights, or whatever other things the owners do
regularly before going out of the house. The owner should
try to determine the order in which he or she usually does
these things and rearrange the order at random and re-
ward the dog for calm behavior. Eventually, short ab-
sences are added to the protocol. More exercise, including
longer walks, lots of games of fetch, and obedience
training also were suggested to reduce anxiety. In more
severe cases, drugs such as amitriptylineb (2 mg/kg body
weight, q 24 hrs) were suggested.
There were specific suggestions to facilitate house-
training. In order to use the dog’s natural inhibition to
soil where sleeping or eating occurs, crate training was
suggested. To allow the owner to supervise the dog so
that it does not misbehave, or to punish promptly if it
does, the dog is tethered to the owner by a leash; this is
referred to as umbilical cord training. Another method is
to put a bell on the dog’s collar so that he cannot sneak
away to eliminate. To regulate the elimination schedule,
food and water availability was scheduled to take advan-
tage of the gastrocolic reflex by taking the dog outside
20 to 30 minutes after the meal.
Chi-square (χ 2) tests were used to determine if there
was a difference in the proportions of improved out-
comes dependent on age, sex, main complaint, and diag-
noses. The χ2 test also was used to determine if there was
a difference in the proportion of dogs who presented
with separation anxiety based on age or between the
proportion of dogs who presented with marking based on
sex or neutering. Two-tailed tests were used at the 0.05
significance level; exact binomial 95% confidence inter-
vals were calculated using the Epitable program in
EpiInfo version 6.04b.c
Results
One-hundred five cases of house-soiling dogs were
treated at some point in time from 1987 to 1996. Of
these, complete information was obtained on 70 cases.
Within the diagnosis of house-soiling problems, incom-
plete housebreaking (n=59; 84%) was the most frequent
diagnosis, and 48 (81%; 95% confidence interval, 69%
to 90%) of these cases improved following behavioral
consultation. Separation anxiety (n=27; 39%) was con-
sidered the second most common diagnosis, and 85%
(n=23; 95% confidence interval, 66% to 96%) of these
cases improved [Figure 1]. From 1987 to 1996, a total of
March/April 1999, Vol. 35
Figure 1—Number of canine house-soiling cases that were
improved versus not improved for each diagnosis. (*Others=fear
or aggression-associated elimination)
121 cases of separation anxiety were seen at the Animal
Behavior Clinic, and 37 (31%) of these cases had house-
soiling problems. More than one diagnosis applied in
some dogs. Of the 70 cases of house soiling for which
follow-up was available, 32 (46%) cases had only house-
soiling problems. For these cases, providing a more fre-
quent opportunity to eliminate was the most frequently
suggested treatment. Twenty-seven (39%) cases had
house-soiling problems accompanied by separation anxi-
ety, and behavior modification was prescribed most fre-
quently. Seven (10%) cases had both aggression (usually
dominance aggression) and house-soiling problems. Four
(6%) cases of house soiling also had a diagnosis of lick
granuloma and excessive licking. Cocker spaniels (n=5),
beagles, and shih tzu (n=4) were the most commonly
seen purebreds; however, mixed-breed dogs (n=14) were,
overall, the most frequently seen canine in this study.
Dogs (n=45; 64%) which both urinated and defecated in
the house were referred more often than dogs which only
urinated or defecated.
More owners lived in houses (n=65) than in apart-
ments. Overall, behavior modification was the most of-
ten suggested treatment (n=58). Taking the dog outside
more frequently was the second most frequently sug-
gested treatment (n=52). Umbilical cord training (n=40)
also was suggested [Figure 2]. In the sexually intact
male, neutering was suggested to reduce marking behav-
Canine House Soiling 103
*
Figure 2—Number of canine house-soiling cases that were
improved versus not improved by suggested treatment (cases
often had more than one suggested treatment). (*Others=use of
a head collar, maturation of the dog, and the owner’s own
training method)
ior. There was no clear-cut difference between suggested
treatments and probability of improvement [see Table].
The mature adult age group was referred most fre-
quently (n=30; 43%) [Figure 3]. Of 30 mature adult
cases, 13 cases were diagnosed with separation anxiety,
whereas in the young adult group of 16 cases, 11 cases
were diagnosed with separation anxiety (χ2=2.70; degree
of freedom [d.f.]=1; p=0.10). Older age and diagnosis of
separation anxiety were significantly associated with
each other (χ 2=13.3; d.f.=2; p=0.001 after combining the
two youngest age groups because of small sampling
sizes). In cases of separation anxiety, amitriptyline was
prescribed to reduce anxiety. There was no significant
difference between males and females (without regard to
neutering, χ2 less than 0.10; d.f.=1; p greater than 0.98)
in improvement rates. Number of cases that were im-
proved versus not improved by sex of dogs are shown in
Figure 4. Male dogs had a significantly greater frequency
of marking behavior than females (χ2=11.51; d.f.=1;
p=0.001), but there was no significant difference in mark-
ing behavior between intact male and neutered dogs
(χ2=0.55; d.f.=1; p=0.46). Of the 70 total cases, 59 cases
(84%; 95% confidence interval, 74% to 92%) were con-
sidered improved. Those cases having only urine elimi-
nation behavior problems had a similar improvement
104 JOURNAL of the American Animal Hospital Association March/April 1999, Vol. 35

Table
Comparison of Treatment Modalities and Probability of Improvement for 70 Cases of
Inappropriate Elimination in Dogs Referred to Cornell University (1987–1996)

Improved* (n=59) Not Improved (n=11)

Yes No Yes No
Take out more frequently 45 14 7 4
Behavior modification 47 12 10 1
Change environment 16 43 3 8
Change feeding schedule 27 32 8 3
Paper training 15 44 2 9
Crate training 30 29 6 5
Cleaning of soiled areas 22 37 2 9
Umbilical cord training 32 27 6 5
Obedience training 36 23 7 4
Increasing exercise 12 47 2 9
Medical treatment 25 34 7 4
Others 11 48 0 11

* Yes=owner tried suggested treatment; No=owner didn’t try suggested treatment

Note: Numbers in the improved/not improved categories represent individual cases (all Chi-square=2.70 or less; all
degree of freedom=1; all p=0.10 or more).

Figure 3—Number of house-soiling cases that were improved

versus not improved by age of dogs with elimination problem
(χ2=0.04; d.f.=1; p=0.85 when adults are compared to younger
dogs; categories combined because of small sample sizes).
Puppy (1–6 mos), Adolescence (7–12 mos), Young adult (13–24
mos), Mature adult (25 mos or older).
Figure 4—Number of house-soiling cases that were improved
versus not improved by sex of dogs. M=intact male,
MC=castrated male, F=intact female, FS=spayed female.
March/April 1999, Vol. 35
Figure 5—Number of canine house-soiling cases that were
improved versus not improved by type of elimination (comparing
urination only to the other two categories combined, χ2=1.92;
d.f.=1, p=0.17).
rate to those with the other types of problem categories
(χ2=1.92; d.f.=1; p=0.17) [Figure 5].
Discussion
An improvement rate of 84% indicates that house-soil-
ing problems can be treated successfully. Although ca-
nine house soiling is easier than feline elimination
problems to manage, if untreated it can become a serious
strain between owner and dog.3 Voith and Borchelt re-
ported that the most common reasons for concurrent
urination and defecation in homes were lack of house-
breaking and separation anxiety behavior, 2 and these
were the most common diagnoses in this study. Some
dogs’ house-soiling problems had more than one etio-
logical diagnosis. Numerical or epidemiological data has
not been presented for inappropriate canine elimination
prior to this study; therefore, no comparison can be made
to other results. But the improvement proportion of 84%
of the cases is encouraging.
In older dogs (greater than 10 years of age), one of the
most frequent owner complaints relating to behavior
problems was inappropriate urination or defecation in
the house, and the most frequent behavioral diagnosis
was separation anxiety. 5 This may be a component of
canine cognitive dysfunction.10 In the authors’ study,
only one dog was older than 10 years, so no comparisons
can be made to Voith and Chapman’s results.5 In the
young and mature adult groups, separation anxiety is one
of the most frequently diagnosed reasons for house soiling.
Canine House Soiling 105
The elimination behavior of dogs involves searching
for a specific location or surface and urinating and def-
ecating to rid the body of waste. Urine and feces also can
be used as part of a communication system between
animals (i.e., marking).4 Marking behavior is considered
almost exclusively a problem involving males, and this
was true in the authors’ study.6 In a study of eliminatory
behavior of domestic dogs in an urban environment, the
dogs responsible for most urine elimination behavior
problems were males.7 In the authors’ study, males were
presented more frequently than females for marking, and
castrated males were as likely to be presented for mark-
ing as intact males. Hart found no quantitative change in
urine marking five months after castration in the dog,
and this contrasts with the rather prompt reduction in
scent marking in male rodents following castration.8
New puppy owners frequently confront misbehavior
associated with the act of elimination due to excitement,
submission, and incomplete housebreaking (which is uri-
nating or defecating in the house outside the toilet area
specified by the owner).9 In this study, puppies were
more likely to both urinate and defecate in the house.
Three owners of improved dogs never used any sug-
gested treatment, and two owners thought their dogs’
maturation solved the elimination problems; the efficacy
of any treatment used here may be overstated. These
results emphasize that understanding normal dog behav-
ior, using correct housebreaking, and correctly using
behavior modification involving positive reinforcement
and properly timed punishment11 are useful in diminish-
ing house-soiling problems in dogs. Veterinarians should
be prepared to offer advice and provide information to
owners to prevent damage of the human-pet bond and
the consequent loss of a client. In this study, the most
frequent mistake that owners made was punishing the
dog minutes to hours after the animal eliminated. In this
scenario, the dog was unlikely to associate the act of
elimination with punishment, even in the presence of
urine or feces. Another frequent mistake was to reward
the dog for returning to the house after the animal pre-
sumably eliminated. The dog was thus rewarded not for
eliminating outside but for returning to the house. That
was why taking the dog outside on a leash rather than
letting it out was emphasized.
a
Promise/Gentle Leader; Ameri-Pet, Inc., Minneapolis, MN
b
Elavil; Stuart Pharmaceutical, Wilmington, DE
c
Anonymous; Centers for Disease Control & Prevention, USDHHS, 1997
Acknowledgments
The authors are grateful to the Korea Science and Engi-
neering Foundation for financial support to the corre-
sponding author and to Anne Gillett for assistance with
follow-up cases.
(Continued on next page)
106 JOURNAL of the American Animal Hospital Association March/April 1999, Vol. 35

References 7. Reid JB, Chantrey DF, David C. Eliminatory behavior of domestic dogs in
an urban environment. Appl Anim Behav Sci 1984;12:279–87.
1. Beaver BV. Owner complaints about canine behavior. J Am Vet Med
8. Hart BL. Environmental and hormonal influences on urine marking
Assoc 1994;204:1953–5.
behavior in adult male dog. Behavioral Biology 1974;11:167–76.
2. Voith VL, Borchelt PL. Diagnosis and treatment of elimination behavior
9. Polsky RH. Elimination misbehaviors in puppies. Vet Pract Staff
problems in dogs. Vet Clin N Am Sm Anim Pract 1982;12:637–44.
1991;3:20–2.
3. Overall KL. In: Overall KL, ed. Clinical behavioral medicine for small
10. Ruehl WW, Depaoli A, Bruyette D. Pretreatment characterization of
animals. St. Louis: Mosby, 1997:195–208.
behavioral and cognitive problems in elderly dogs. J Vet Int Med
4. Voith VL, Borchelt PL. Elimination behavior and related problems in dogs. 1994;8:179.
Comp Cont Ed Pract Vet 1985;7:537–47.
11. Voith VL, Borchelt PL. Readings in companion animal behavior. Trenton:
5. Voith VL, Chapman BL. Behavioral problems in old dogs: 26 cases (1984– Veterinary Learning Systems, 1996:62–71.
1987). J Am Vet Med Assoc 1990;196:944–6.
6. Askew HR. Treatment of behavior problems in dogs and cats. Cambridge:
Blackwell Science, 1996:228–44.
 Rectal Ganglioneuroma in a Dog
An 18-month-old, spayed female Australian terrier cross was presented with a 10-month history
of chronic large bowel diarrhea. Ulceration and two proliferative masses in the rectum were
seen on colonoscopy. Surgical resection was performed to remove the masses, and the dog
recovered without complications related to surgery. Histopathology was consistent with the
diagnosis of ganglioneuroma. The dog had no clinical signs of disease within three months of
surgery and was completely normal 2.5 years after diagnosis. This is the first report providing
follow-up and successful outcome of a ganglioneuroma in the gastrointestinal tract of a dog.
J Am Anim Hosp Assoc 1999;35:107–10.

Michele E. Reimer, DVM Case Report

Michael S. Reimer, DVM, MS,
Diplomate ACVIM
Geoffrey K. Saunders, DVM, MS,
Diplomate ACVP
Spencer A. Johnston, VMD,
Diplomate ACVS
C perform a thorough digital rectal examination, since it elicited an extreme
From the Department of Small Animal
Clinical Sciences, Virginia–Maryland
Regional College of Veterinary Medicine,
Virginia Polytechnic Institute and State
University, Blacksburg, Virginia 24061-0442.
An 18.8-kg, 18-month-old, spayed female Australian terrier cross was
presented to the Virginia-Maryland Regional College of Veterinary Medi-
cine with a 10-month history of chronic large bowel diarrhea. The diar-
rhea was characterized by hematochezia, mucus, tenesmus, and an
increased frequency. Multiple fecal flotations for intestinal parasites by
the referring veterinarian were negative. The dog had been treated with
appropriate dosages of fenbendazole, sulfasalazine, metronidazole, and
prednisone, as well as a diet change; however, no improvement was seen.
The dog was not receiving any medications at the time of admission.
The only remarkable physical examination finding was the inability to
pain response. Complete blood count and biochemical profile were within
normal limits. Two fecal flotations with zinc sulfate were negative. Rectal
cytology was negative for clostridial spores, but an increased number of
neutrophils was seen. Fecal examination for clostridial enterotoxin using
a reverse latex agglutination testa was negative.
The dog was prepared for colonoscopy with two doses of a polyethyl-
ene glycol-electrolyte solutionb (66 ml/kg body weight) given by orogastric
intubation two hours apart, the day before the procedure. Metoclopramide
(0.2 mg/kg body weight) was given subcutaneously (SC) 30 minutes prior
to the first dose of the polyethylene glycol-electrolyte solution. Five
warm-water enemas (20 ml/kg body weight) were given within 24 hours
prior to colonoscopy. Food was withheld for 24 hours.
The dog was premedicated with acepromazine (0.05 mg/kg body
weight, SC) and oxymorphone (0.05 mg/kg body weight, SC) prior to
induction with intravenous thiopental (15 mg/kg body weight) and was
maintained on isoflourane and oxygen. Endoscopic examination was
performed using a flexible video endoscopec while the dog was positioned
in left lateral recumbency. The majority of the large intestine was normal,
except for the distal 2 cm, where the mucosa was ulcerated for 180˚ of the
luminal circumference. This area oozed blood and was friable. Two
proliferative masses (5 mm and 10 mm in diameter) were present in the
ulcerated region of the rectum [Figure 1]. Multiple tissue samples were
taken from the entire colon and from the ulcerated area with flexible
forceps. A section of the larger mass was obtained, and the smaller mass
was removed completely with rigid rectal biopsy forceps passed along-
side the endoscope. The dog recovered from anesthesia uneventfully and
was discharged to receive sulfasalazine (50 mg/kg body weight, per os
[PO] tid) while awaiting histopathological results.

JOURNAL of the American Animal Hospital Association 107

108 JOURNAL of the American Animal Hospital Association
Figure 1—Endoscopic appearance of the colon from an 18-
month-old Australian terrier cross. Note the two proliferative
masses and ulceration.
Histopathological diagnosis of the masses was gan-
glioneuroma. Mild, lymphocytic plasmacytic colitis was
present throughout the colon. Histopathologically, the
masses contained large aggregates of normal ganglion
cells within the lamina propria, as well as an increase in
the number of lymphoid cells. Well-differentiated gan-
glion cells formed large aggregates that replaced glands
and formed polypoid proliferations of the mucosa. Sur-
gical resection of the masses was recommended.
The dog continued to exhibit tenesmus and was re-
turned to the hospital for surgical resection 10 days
following biopsy. The same anesthetic protocol was used.
The patient was placed in a perineal position, and the
rectal mucosa was grasped and exteriorized through the
anus. The mass was identified, undermined between the
submucosa and the muscularis, and removed. Approxi-
mately 3 cm of mucosa was removed with margins ex-
tending from the anocutaneous junction orally. The dog
strained to defecate for the first 12 hours following sur-
gery, but then this spontaneously subsided. Sulfasalazine
was continued for two weeks following surgery.
The mucosa of the rectal mass was thickened and
papillary, with hyperplasia of the glands, loss of goblet
cells, and focal hemorrhage seen on histological evalua-
tion. Aggregates of well-differentiated neurons were
present in many areas of the lamina propria which dis-
torted and displaced the glands. The mucosa and submu-
cosa were replaced mostly by aggregates of mature
neurons, satellite cells, and Schwann cells and their pro-
cesses that formed small nests, larger nodules, and cords
all separated by bands of smooth muscle or fibrous tissue
[Figure 2]. Occasional glands and mucus-filled lakes
were present in the submucosa among the neurons, and
March/April 1999, Vol. 35
Figure 2—Photomicrograph of the colonic mucosa from the dog
in Figure 1. Few widely scattered colonic glands (asterisks) are
displaced by tumor cells. Large groups of mature neurons
(arrows) and nerve fibers are present in the lamina propria
between glands (Hematoxylin and eosin stain, 150X; bar=67 µ).
some mucosal glands were dilated and filled with mucus.
Neurons were large and polygonal with abundant cyto-
plasm, and they often contained Nissl substance. Nuclei
were eccentric, large, oval, and vesicular. The support-
ing satellite and Schwann cells had short, elongated, and
dense nuclei surrounded by abundant nerve fibers. The
histopathological diagnosis was ganglioneuroma.
The dog had watery diarrhea during the initial postop-
erative period. The owner had been adding psylliumd (1
tablespoon mixed with food every 12 hrs) to the diet, and
it was recommended that this be decreased. The stool
consistency improved once the psyllium was decreased
(1 tablespoon per day); however, the dog then began to
exhibit tenesmus and dyschezia. An anal stricture was
ruled out following digital rectal examination.
Sulfasalazine was discontinued. Medical treatment of
tenesmus and dyschezia was initiated with prednisone (1
mg/kg body weight, PO bid) and loperamide (0.1 mg/kg
body weight, PO bid). Within one week, the dog’s stool
was normal and both tenesmus and dyschezia resolved.
The loperamide was continued for four weeks and then
discontinued. The prednisone was decreased to 0.5
March/April 1999, Vol. 35
mg/kg body weight, PO bid after two weeks, then elimi-
nated using a tapering dose over the next month. Three
months following surgery, the dog was not receiving any
medication and was doing well. The dog has now been
clinically normal for 2.5 years following initial diagno-
sis and treatment.
Discussion
Ganglioneuromas are rare tumors of neuroblastic origin
containing combinations of ganglion cells, nerve fibers,
Schwann cells, and neuroblasts. Ganglioneuroblastomas
tend to be more differentiated than neuroblastomas but
are less differentiated than ganglioneuromas. A separate
distinction is made for ganglioneuromatosis, which is a
hyperplastic proliferation of ganglion cells and nerves.1
Tumors in the neuroblastoma-ganglioneuroma group can
contain areas that histologically resemble both tumors,
differentiated and undifferentiated. These tumors are
sometimes difficult to characterize histopathologically,
and ultrastructural and immunohistochemical features
are often helpful.15
In humans, young children are affected most com-
monly with tumors of the neuroblastoma-ganglioneuroma
group. These tumors have a predilection for the adrenal
medulla as well as the mediastinal and retroperitoneal
areas. 2 An embryonic origin, presumably from primitive
cells of the neuroectoderm, has been hypothesized for
tumors of the neuroblastoma-ganglioneuroma group.3
Their behavior is often unpredictable.
To the authors’ knowledge, ganglioneuroma in the
gastrointestinal tract has not been reported previously in
the dog.4–9 Tumors of the neuroblastoma-ganglioneuroma
group have been identified in several case reports. A 12-
month-old, female Lhasa apso was diagnosed with
colorectal ganglioneuromatosis after presenting with
clinical signs of large bowel diarrhea.1 Tumors of the
neuroblastoma-ganglioneuroma group have been de-
scribed in the adrenal medulla,10 mediastinum,2,11,12 ret-
roperitoneal space,12 trigeminal ganglion,13 bifurcation
of the common carotid artery, 14 and nasal cavity15 of the
dog. These tumors also have been described in the je-
junum of a kitten16 and the adrenal gland of a calf.3
Most previous case reports involving animals pro-
vided histopathological diagnosis at the time of necropsy.
This is the only dog reported in the literature, to the
authors’ knowledge, that has been treated successfully
and survived.
Large bowel tumors account for 36% to 60% of the
gastrointestinal neoplasms in dogs.17 Adenocarcinoma
and lymphoma are the most common malignancies in the
large bowel, representing 66% of the cases.5,17 Adenomas
represent 33% of large bowel tumors, and occasionally
smooth muscle tumors are seen.18–21 Most large bowel
neoplasms are located in the rectum or distal third of the
colon.
Ganglioneuroma 109
Clinical signs associated with large bowel tumors are
often difficult to differentiate from other large bowel
diseases, such as colitis. Diarrhea, hematochezia, tenes-
mus, and dyschezia often are reported. 5–7,9,17,18,22 Anal
bleeding that is not associated with defecation is noted
occasionally. Constipation and rectal prolapse also have
been reported.
Most large intestinal neoplasms are detectable by digi-
tal rectal examination.17 There are usually few hemato-
logical or biochemical abnormalities. Diagnostic imaging
including ultrasonography, survey radiographs, and con-
trast enemas may be useful if endoscopy is not
available.17
This case report describes a young dog with a rare
tumor in the rectum. The dog had no clinical signs of
disease within three months of surgery and was com-
pletely normal 2.5 years after diagnosis. This is the first
report of a ganglioneuroma in the gastrointestinal tract
of a dog with a successful outcome.
a
PET-RPLA Kit; Microbio, Denver, CO
b
GoLytley; Braintree Labs, Braintree, MA
c
Fujinon EVE EG7-FP2 Video Endoscope; Fujinon Inc., Wayne, NJ
d
Metamucil; Procter and Gamble, Cincinnati, OH
References
1. Fairley R, McEntee M. Colorectal ganglioneuromatosis in a young female
dog (Lhasa apso). Vet Path 1990;27:206–7.
2. Hawkins K, Summers B. Mediastinal ganglioneuroma in a puppy. Vet Path
1987;24:283–5.
3. Riley M, Forsyth W. Bilateral adrenal ganglioneuroblastoma in a premature
calf. Aust Vet J 1976;52:234–5.
4. Schaffer E, Schiefer B. Incidence and types of canine rectal carcinomas.
J Sm Anim Pract 1968;9:491–5.
5. Holt P, Lucke V. Rectal neoplasia in the dog: a clinicopathological review
of 31 cases. Vet Rec 1985;116:400–5.
6. Seiler R. Colorectal polyps of the dog: a clinicopathologic study of 17
cases. J Am Vet Med Assoc 1979;174:72–5.
7. Brodey R, Cohen D. An epizootiologic and clinicopathologic study of 95
cases of gastrointestinal neoplasms in the dog. J Am Vet Med Assoc
1964;101:167–79.
8. Patnaik A, Hurvitz A, Johnson G. Canine gastrointestinal neoplasms. Vet
Path 1977;14:547–55.
9. Hayden D, Nielsen S. Canine alimentary neoplasia. Zbl Med A
1973;20:1–22.
10. Simon J, Albert L. Two cases of neuroblastomas in dogs. J Am Vet Med
Assoc 1960;136:210–4.
11. Schultz K, Steele K, Saunders G, et al. Thoracic ganglioneuroblastoma in a
dog. Vet Path 1994;31:716–8.
12. Kelly D. Neuroblastoma in a dog. J Path 1974;116:209–12.
13. Beezley D. A trigeminal ganglioneuroma in a dog. Cornell Vet
1969;59:584–93.
14. Ferrell J, Hunt R, Nims R. Cervical ganglioneuroma in a dog. J Am Vet
Med Assoc 1964;144:508–12.
15. Mattix M, Mattix R, Williams B, et al. Olfactory ganglioneuroblastoma in a
dog: a light, ultrastructural, and immunohistochemical study. Vet Path
1994;31:262–5.
16. Patnaik A, Leiberman P, Johnson G. Intestinal ganglioneuroma in a kitten–
a case report and review of literature. J Sm Anim Pract 1978;19:735–42.
(Continued on next page)
110 JOURNAL of the American Animal Hospital Association March/April 1999, Vol. 35

References (cont’d) 20. Kapatkin A, Mullen H, Matthiesen D, et al. Leiomyosarcoma in dogs: 44

17. White R, Gorman N. The clinical diagnosis and management of rectal and
pararectal tumours in the dog. J Sm Anim Pract 1987;28:87–107.
18. Valerius K, Powers B, McPherron M, et al. Adenomatous polyps and
carcinoma in situ of the canine colon and rectum: 34 cases (1984–1994).
J Am Anim Hosp Assoc 1997;33:156–60.
cases (1983–1988). J Am Vet Med Assoc 1992;201:1077–9.
21. Bruecker K, Withrow S. Intestinal leiomyosarcomas in six dogs. J Am
Anim Hosp Assoc 1988;24:281–4.
22. Church E, Mehlhaff C, Patnaik A. Colorectal adenocarcinoma in dogs: 78
cases (1973–1984). J Am Vet Med Assoc 1987;191:727–30.

19. McPherron M, Withrow SJ, Seim H, et al. Colorectal leiomyomas in seven

dogs. J Am Hosp Assoc 1992;28:43–6.
 Pythiosis of the Digestive Tract
in Dogs From Oklahoma
Enteric pythiosis was diagnosed in nine dogs in Oklahoma. Eight dogs had anorexia and weight
loss. Two of these dogs had diarrhea; two dogs exhibited vomiting and diarrhea; and one dog
had vomiting. One dog presented with dysphagia. Seven dogs had either a palpable or
radiographically visible abdominal mass. These seven dogs had localized regions of mucosal
ulceration and thickened gastric or intestinal walls with some involvement of the adjacent
mesentery or omentum. Two dogs had enlarged regional mesenteric lymph nodes. One dog
that presented with dysphagia had an oropharyngeal mass involving the larynx and cranial
esophagus. Microscopically, there was transmural chronic sclerosing and granulomatous to
pyogranulomatous inflammation with arteritis. Pythium spp. were identified in all specimens by
immunohistochemistry. J Am Anim Hosp Assoc 1999;35:111–4.

R. Gayman Helman, DVM, PhD Introduction

Julian Oliver III, DVM, PhD
C broad term “phycomycosis.” Phycomycosis covers infections caused by a
From the Oklahoma Animal Disease
Diagnostic Laboratory (Helman),
Oklahoma State University,
Stillwater, Oklahoma 74078 and the
Department of Veterinary Pathology (Oliver),
College of Veterinary Medicine,
Louisiana State University,
Baton Rouge, Louisiana 70803.
Address all correspondence to Dr. Helman.
Localized enteric and dermal mycotic infections have been reported to
occur in dogs with nonpigmented opportunistic fungi.1–9 These infections
are seen most commonly in states along the Gulf of Mexico where the
disease is considered endemic, and they have been grouped under the
number of organisms including Pythium gracile, Hyphomyces destruens,
the Mucorales (Rhizopus spp., Mucor spp., Absidia spp., and Mortierella
spp.), and the Entomophthorales (Conidiobolus spp. and Basidiobolus
spp.).10–12 Fungal genera Rhizopus, Mucor, Absidia, and Mortierella also
were grouped as causative agents for mucormycosis, while the genera
Conidiobolus and Basidiobolus caused entomophthoromycosis.4,11 Im-
proved fungal culture and identification methods have advanced the un-
derstanding of the biology of these organisms. One result of this has been
a new classification scheme.10–12 The term “phycomycosis” has been
replaced largely by the terms “pythiosis” and “zygomycosis.”
Pythiosis is caused by the aquatic oomycete Pythium insidiosum (class:
oomycetes) and is reported to cause dermatological and enteric diseases
in horses and dogs.2,3,5,8,10,12–14 Dermatitis is seen in situations where
there is prolonged skin contact with stagnant water in which the organism
grows. The exact mechanism of enteric infection is unknown but is
suspected to involve ingestion of the organism and entry into the wall of
the digestive tract through a mucosal wound. Zygomycosis is a disease
caused by fungal genera in the orders Mucorales and Entomophthorales
(class: zygomycetes).11 Zygomycosis also may present as either primary
skin or enteric disease. The literature identifies zygomycosis primarily as
an enteric disease in dogs.1,4,11 The pathogenesis of enteric infections with
zygomycetes is believed to be similar to that described for pythiosis.11
Both diseases are reported to have localized, progressive granuloma-
tous to pyogranulomatous inflammation.1–8 Disseminated systemic dis-
ease is less common. Vasculitis is often reported with both diseases.1–9
Morphology of the tissue-phase hyphae overlaps.2–4,7,9,12 Consequently,
these two conditions are difficult to distinguish by routine histopathology.
Culture is required for definitive identification of zygomycosis. Culture is

JOURNAL of the American Animal Hospital Association 111

112 JOURNAL of the American Animal Hospital Association March/April 1999, Vol. 35

Figure 1—Map of Oklahoma showing the distribution of nine

cases of enteric pythiosis in dogs. Cases were identified in nine
counties in the eastern half of the state, with the northeastern
quadrant having seven of the nine cases.
Figure 2A
more tedious for Pythium, and there is an immunohis-
Figures 2A, 2B—Gastrointestinal lesions in dogs with pythiosis.
tochemical method for its identification in formalin-fixed (A) Stomach and cranial duodenum of a dog with pythiosis,
tissues.15 The actual individual incidences of these two opened to show a deep gastric ulcer (left) with marked thickening
organisms in veterinary medicine and their propensity of a segment of the duodenal wall. (B) Cross section of duode-
num of a dog with pythiosis, showing marked thickening of the
for producing dermatological and enteric infections are wall with narrowing of the lumen. The yellowish-tan foci are
probably not truly known, because in many earlier re- areas of necrosis which typically have hyphal elements. The
ports of these two diseases, cultures were not always grayish-white tissue is excess fibrous connective tissue.
performed and specific organism identification was not
possible.1,4,6,7,14 In these instances, the disease was re-
ferred to simply as phycomycosis.
This report describes enteric disease caused by
Pythium spp. in nine dogs in eastern Oklahoma. This
disease has not been reported previously in Oklahoma.

Materials and Methods

Case records of nine dogs with enteric pythiosis were
reviewed. Diagnosis of pythiosis was based on finding
locally extensive granulomatous to pyogranulomatous
inflammation in the tubular digestive tract, with intra-
lesional Pythium spp. Signalments, clinical presentations,
laboratory data (when available), gross appearance of
lesions, histopathological findings, and outcome of treat-
ment were compared.
Tissues submitted for pathological examination were
preserved by immersion in 10% formalin. All samples
were processed in a standard manner for paraffin embed-
ding. The tissue sections were stained with hematoxylin
and eosin stain for lesion characterization. Gomori’s
methenamine silver (GMS) or periodic acid-Schiff’s
stains were used to stain fungi for determination of hy-
phal morphology. Immunohistochemistry using an anti-
serum against antigens in the wall of Pythium spp. was
performed according to a prior technique.15
Results
Cases of enteric pythiosis were distributed in eastern
Oklahoma, with a cluster of seven cases in northeastern
Oklahoma. The remaining two cases were located in
south-central Oklahoma [Figure 1]. These areas of Okla-
homa receive 34 to 42 inches (86 to 107 cm) of precipita-
tion per year. They are comprised of forested, rolling to
hilly landscape used extensively for food animal produc-
Figure 2B
tion. As opposed to the western half of the state, these
areas have abundant lakes, rivers, and ponds. All dogs
were housed primarily outside and allowed to roam.
These dogs were reported to have access to farm ponds.
All nine dogs in this report were native to Oklahoma and
had never traveled outside the state.
The signalments and clinical features of nine dogs
with enteric pythiosis are presented in the Table. Differ-
ential clinical diagnoses included carcinoma (or
adenocarcinoma), smooth muscle tumor, foreign body
entrapment, and intussusception.
Exploratory surgery was performed on five dogs (case
nos. 1–3, 7, 9). Three dogs (case nos. 4, 6, 8) did not
receive any treatment, died shortly after presentation,
and were necropsied. One dog (case no. 5) was eutha-
nized after a diagnosis of pythiosis was made on biopsy
March/April 1999, Vol. 35 Pythiosis 113

Table
Signalment and Clinical Data for Nine Dogs With Enteric Pythiosis

Case No. Breed Sex* Age Clinical Features

1 Dalmatian M 1.5 yrs Weight loss, anorexia, abdominal mass palpated
2 Walker hound M 2 yrs Weight loss, diarrhea, pancreatic/duodenal mass suspected on
radiographs
3 Scottish terrier M 6 yrs Weight loss, intermittent vomiting and diarrhea, anorexia, mass
palpated in craniodorsal quadrant of abdomen
4 Rottweiler M 3 yrs Weight loss, diarrhea, anorexia
5 Terrier mix M 8 yrs 5-month history of dysphagia with laryngeal swelling
6 Pekingese M 10 mos 2.5 months of vomiting, weight loss, anorexia; radiographs
showed thickened gastric and duodenal walls
7 Labrador M 1.5 yrs 6-week history of weight loss, anorexia; intestinal mass palpated
retriever mix
8 Heeler mix M 11 mos 4-day history of anorexia; abdominal mass palpated; dog died
unexpectedly within 8 hours of presentation to the veterinarian
9 Retriever (not M 1 yr 2-week duration of vomiting and diarrhea with weight
specified) loss; abdominal mass palpated in cranioventral abdomen

* M=male

specimens obtained from a pharyngeal mass. Repre-
sentative examples of gross enteric lesions in dogs
with confirmed pythiosis are shown in Figures 2A
and 2B. Dogs with gastrointestinal involvement had
localized thickening of the gastric or intestinal walls.
Affected tissue was firm and discolored with numer-
ous, yellowish-tan foci. The mucosa was ulcerated
[Figure 2A]. Lesions extended to the serosal surface
at mesenteric attachments and were reported to in-
volve mesentery and omentum. Adjacent mesenteric
lymph nodes were enlarged and firm in two dogs. The
pharyngeal mass in case no. 5 extended through the
pharyngeal wall into and surrounding the larynx and
cranial esophagus. The tissue was firm, and mucosal
surfaces of the pharynx, larynx, and esophagus were
ulcerated. Case no. 8 had infarction of the segment of
small intestine with pythiosis.
Tissues submitted for microscopic examination to the
Oklahoma Animal Disease Diagnostic Laboratory
(OADDL) had similar lesions. The basic reaction was a
mixture of granulomatous to pyogranulomatous inflam-
mation with pronounced fibrosis. Inflammatory lesions
had multiple foci of necrosis, and eosinophils were a
common component of leukocytic infiltrates. Necrotiz-
ing vasculitis also was prominent. Hyphae were present
in necrotic foci and vessels. Hyphae were coarse, occa-
sionally septate, irregularly branched, and measured ap-
proximately 5 to 7 microns in diameter. Fresh tissue was
not available for culture, but formalin-fixed, paraffin-
embedded tissues were submitted to the Department of
Veterinary Pathology at Louisiana State University’s
College of Veterinary Medicine for immunohistochemi-
cal staining for Pythium spp. All cases were positive.
Prognosis was considered poor in all cases. Surgical
resection was attempted in five dogs (case nos. 1–3, 7,
and 9). Clinical signs dissipated within the first week
after surgery. Appetites of the animals improved, and
body weights increased for periods from six to 12 weeks
following surgery. Subsequently, clinical signs similar
to those at the initial presentation recurred. These five
dogs were euthanized. No necropsies were performed.
Discussion
The clinical presentation and lesions in the nine dogs of
this report closely resemble cases of enteric pythiosis
reported by others.2,5–8 Dogs with pythiosis are prone to
develop chronic wasting disease that is unresponsive to
medical management. 2,4–8 Key presenting clinical fea-
tures are weight loss that usually is accompanied by
anorexia, with variable frequency of diarrhea and vomit-
ing. Enteric pythiosis, for unknown reasons, is more
commonly seen in young male dogs.2,7,10 Affected dogs
usually are housed outside, are free to roam, and have
access to standing water. This was true for all nine dogs
of this report.
Initial sample submissions to the OADDL for histo-
pathological interpretation did not include enteric my-
cotic infection in the differential diagnosis. This disease
was not thought to occur as far north as Oklahoma where
the climate is considerably drier and colder than the Gulf
coastal region. The disease is especially prominent in
Louisiana, Florida, and southeastern Texas, presumably
114 JOURNAL of the American Animal Hospital Association March/April 1999, Vol. 35

because the climate favors ample standing water sources
and winter temperatures are mild. These conditions fa-
vor the growth of Pythium spp. which requires both
water and damaged plant or animal tissues for growth.13,16
It has been shown that growth is reduced markedly at
cold temperatures (5˚ C), and the organism is killed by
freezing.17
Prior to 1994, reports of this disease (listed as either
pythiosis or phycomycosis) have been from states bor-
dering the Gulf of Mexico.3,6,7 Including the present
report, enteric pythiosis now has been described in the
southern plains in Oklahoma and Missouri as well as in
Kentucky and Tennessee.2,5,8 Winters in these areas gen-
erally have numerous hard freezes which affect small
agricultural ponds and tanks as well as creeks and small
3. Foil CS, Short BG, Fadok VA, Kunkle GA. A report of subcutaneous
pythiosis in five dogs and a review of the etiologic agent Pythium spp.
J Am Anim Hosp Assoc 1984;20:959–66.
4. Gleiser CA. Mucormycosis in animals. A report of 3 cases. J Am Vet Med
Assoc 1953;123:441–5.
5. McLaughlin BG, Ayer AA. Gastrointestinal pythiosis in a dog from
Kentucky. Canine Pract 1995;20:17–9.
6. Miller RI, Qualls CW Jr, Turnwald GH. Gastrointestinal phycomycosis in a
dog. J Am Vet Med Assoc 1983;182:1245–6.
7. Miller RI. Gastrointestinal phycomycosis in 63 dogs. J Am Vet Med Assoc
1985;186:473–8.
8. Patton CS, Hake R, Newton J, Toal RL. Esophagitis due to Pythium
insidiosum in two dogs. J Vet Intern Med 1996;10:130–42.
9. Purcell KL. Jejunal obstruction caused by a Pythium insidiosum granuloma
in a mare. J Am Vet Med Assoc 1994;205:337–9.
10. Foil CS. Oömycosis (pythiosis). In: Greene CE, ed. Infectious diseases of
dogs and cats. Philadelphia: WB Saunders, 1990:731.
11. Foil CS. Zygomycosis. In: Greene CE, ed. Infectious diseases of dogs and
cats. Philadelphia: WB Saunders, 1990:734.

streams in wooded areas. Although there may be exten-
sive freezing of the surface in these ponds, deeper re-
gions remain unfrozen, and presumably this allows for
overwintering of Pythium. It is reasonable to expect that
pythiosis may be present in other states in this region
with similar geographical and climatic conditions.
Pythiosis should be considered in the differential diag-
12. Mendoza L, Ajello L, McGinnis MR. Infections caused by the oomycetous
pathogen Pythium insidiosum. J Mycol Med 1996;6:151–64.
13. DeCock AWAM, Mendoza L, Padhye AA, et al. Pythium insidiosum sp
nov, the etiologic agent of pythiosis. J Clin Microbiol 1987;25:344–9.
14. Miller RI, Campbell RSF. Clinical observations on equine phycomycosis.
Aust Vet J 1982;58:221–6.
15. Brown CC, McClure JJ, Triche P, Crowder C. Use of immunohistochemi-
cal methods for diagnosis of equine pythiosis. Am J Vet Res 1988;49:
1866–8.

nosis of chronic wasting, enteric disease in dogs with 16. Mendoza L, Hernandez F, Ajello L. Life cycle of the human and animal
oomycete pathogen Pythium insidiosum. J Clin Microbiol 1993;31:
obvious thickening of intestinal segments. Gross lesions, 2967–73.
although not diagnostic, should strongly suggest my- 17. Bentinck-Smith J, Padhye AA, Maslin WR, et al. Canine pythiosis—
isolation and identification of Pythium insidiosum. J Vet Diagn Invest
cotic disease to the veterinarian. Confirmation must be 1989;1:295–8.
made by histopathological examination of diseased tis-
sue. Culture methods for Pythium spp. are complex and
may go beyond the means that most conventional labora-
tories offer for mycotic culture and identification. Vet-
erinarians should check with their diagnostic laboratory
to determine if the laboratory is set up to isolate and
identify Pythium spp. Immunohistochemical identifica-
tion of hyphae is possible, and this can be performed on
formalin-fixed tissues.

Conclusion
Pythiosis can present as a progressive wasting condition
associated with anorexia, weight loss, and diarrhea or
vomiting or both, depending on the location of the lesion
(i.e., upper or lower digestive tract). Affected dogs nor-
mally have localized areas of inflammation that present
as a mass and may occur anywhere along the tubular
digestive tract. The disease now occurs in dogs of the
southern plains of the United States in addition to the
more tropical regions along the Gulf coast. Prognosis is
poor, because most cases are advanced when the disease
is recognized and the condition has a tendency to recur
after surgical resection.

References
1. Ader PL. Phycomycosis in fifteen dogs and two cats. J Am Vet Med Assoc
1979;174:1216–23.
2. Fischer JR, Pace LW, Turk JR, Kreeger JM, Miller MA, Gosser HS.
Gastrointestinal pythiosis in Missouri dogs: eleven cases. J Vet Diagn
Invest 1994;6:380–2.
March/April 1999, Vol. 35 Pythiosis 115

Ross & Company

4C ad
pull JAAHA Jan/Feb p. 25

page 115
Veterinary Orthopedic Implants
B&W ad
pull Feb Trends p. 8

page 116
 Secondary Hypoadrenocorticism
Associated With Craniocerebral
Trauma in a Dog
An 11-month-old, neutered female miniature schnauzer presented with a severe head injury.
The dog was treated for the acute effects of craniocerebral trauma and was hospitalized for just
over a week. Several weeks later, she became weak and lethargic. A diagnosis of
hypoadrenocorticism was confirmed with an adrenocorticotropin hormone (ACTH) stimulation
test. An endogenous ACTH assay confirmed secondary hypoadrenocorticism. The dog was
tested for hypopituitarism with canine thyroid-stimulating hormone and thyroxine serum assays
and an insulin-like growth factor assay. These tests could not confirm panhypopituitarism in this
dog. The hypoadrenocorticism was treated with prednisone, and the dog remains controlled
adequately three years later. J Am Anim Hosp Assoc 1999;35:117–22.

Simon R. Platt, BVM&S, MRCVS Introduction

Cheryl L. Chrisman, DVM, MS, EdS,
Diplomate ACVIM
John Graham, MVB, MSc,
Diplomate ACVR,
Diplomate ECVDI
Roger M. Clemmons, DVM, PhD
C injury is still considered rare in human medicine.2,3 This paper describes a
Many common and significant medical complications of craniocerebral
trauma exist.1 Those complications documented in humans include car-
diovascular problems, respiratory function alterations, water and electro-
lyte derangements, and hypothalamic-pituitary dysfunction.1 Endocrine
abnormalities resulting from traumatic dysfunction of the hypothalamus
and pituitary include the syndrome of inappropriate secretion of antidi-
uretic hormone, diabetes insipidus, hypogonadotrophic hypogonadism,
hypothyroidism, growth hormone deficiency, and hypocortisolism.2–4 The
pathological changes responsible for the endocrinopathies can be fatal.
This may be the reason that posttraumatic hypothalamic-pituitary axis
case of secondary hypoadrenocorticism diagnosed in a dog following
severe head injury and documents its treatment for three years after the
trauma.

From the Department of Small Animal
Clinical Sciences,
P.O. Box 100126,
College of Veterinary Medicine,
University of Florida,
Gainesville, Florida 32610-0126.
Address all correspondence and reprint
requests to Dr. Platt.
Case Report
An 11-month-old, neutered female miniature schnauzer presented as an
emergency to the Veterinary Teaching Hospital of the University of
Florida with craniocerebral trauma. The owner reported that the dog had
been healthy with no neurological signs immediately preceding the trauma.
Just prior to presentation, the owner had witnessed two generalized tonic-
clonic motor seizures.
The dog was admitted in a stuporous state with marked tetraparesis,
which was worse on the right side. The dog had miosis with a reduced,
direct pupillary light response in the right eye and a reduced, consensual
pupillary light response in the left eye. An absent menace reaction and a
ventrolateral strabismus were detected in the right eye. Positional hori-
zontal nystagmus was present in both eyes, with the fast phase to the right.
Facial sensation was bilaterally depressed as were the palpebral responses.
Conscious proprioception was absent in all four limbs. All segmental
spinal reflexes were intact. Cervical and thoracolumbar pain was present
diffusely on palpation. A strong, regular heart beat with a rate of 70 beats
per minute was unresponsive to digital ocular pressure. Systolic blood
pressure was recorded to be 140 mmHg.

JOURNAL of the American Animal Hospital Association 117

118 JOURNAL of the American Animal Hospital Association
Figure 1—Rostrocaudal radiograph of the calvarium from an 11-
month-old miniature schnauzer presented with a severe head
injury. A comminuted, depressed fracture of the left parietal bone
is noted (arrow).
Serum chemistry analysis was normal. A mature neu-
trophilia (neutrophil count, 19.2 x103/µl; reference range,
6.0 to 17.0 x103/µl) and an elevated fibrinogen level
(600 mg/dl; reference range, less than 300 mg/dl) were
the only abnormalities evident on a complete blood count.
A comminuted, slightly depressed fracture of the left
calvarium, involving the temporal and parietal bones,
was seen on radiographs of the skull. A mildly displaced
right frontal bone fracture was present, with increased
opacity within the frontal sinus [Figure 1]. No abnor-
malities were detected on thoracic, abdominal, or spinal
radiographs. All of the first three waves of the brain-
stem auditory-evoked potentials were present when
evaluated shortly after the time of admission.
The patient was administered 2.5% dextrose in half-
strength lactated Ringer’s solution intravenously (IV) at
a maintenance rate (60 ml per kg body weight over 24
hrs). Phenobarbitala (2 mg/kg body weight, IV q 12 hrs
[bid]) was initiated in conjunction with methylpredniso-
lone sodium succinateb (15 mg/kg body weight, IV q 8
hrs) after a bolus of a 30 mg/kg body weight dose. The
IV steroids were discontinued after 48 hours and re-
placed by oral prednisonec (0.25 mg/kg body weight,
bid). To avoid the potential intestinal side effects of the
high doses of the corticosteroids, ranitidined (2 mg/kg
body weight, IV q 8 hrs [tid]), an H2-histamine receptor
antagonist, was administered. A low dose of butorphanole
(0.1 mg/kg body weight, subcutaneously [SC] q 6 hrs)
was used to provide analgesia. Prophylactic antibiotic
therapy was provided by the addition of cefazolinf (20
mg/kg body weight, IV tid). The dog was placed in an
oxygen cage with an inspired oxygen concentration of
40%.
Within 24 hours the dog had become conscious, nor-
mothermic, and her cardiopulmonary status was stable.
Glucose and electrolyte serum levels had remained nor-
mal throughout this time, and after 72 hours the dog was
eating and drinking sufficiently on her own. By the fifth
March/April 1999, Vol. 35
day the patient exhibited a left head tilt and a left
adversion syndrome. The dog’s attitude had improved,
but she was still depressed and demented. Ambulation
was possible, but a marked right-sided hemiparesis was
evident. Conscious proprioception deficits remained in
all four limbs but were worse on the right side. Intrave-
nous medication was replaced by oral medication on the
seventh day. Cefadroxilg (15 mg/kg body weight, bid)
was initiated orally for a further seven days. The phe-
nobarbital was prescribed as a 20 mg per 15 ml elixir (2
mg/kg body weight, bid), and the prednisone was dis-
continued.
The dog was discharged eight days after admission
and returned four weeks later for a reevaluation. The
owner described a gradual improvement in the dog’s
mentation. Moderate hemiparesis and conscious pro-
prioception deficits were present on the right side. A
depressed menace reaction remained in the right eye, but
the pupillary light response was normal on both direct
and consensual examination. A right-sided facial paresis
remained but had improved markedly. The dog was be-
ing maintained on phenobarbital elixir (2 mg/kg body
weight, bid) orally, and the serum phenobarbital level 10
hours after the last administration was 7 µg/ml (thera-
peutic reference range, 15 to 40 µg/ml).5 The owner had
not been aware of any seizure activity occurring in the
dog since the trauma, and it was decided at this point to
continue with the same dosing regime for the phenobar-
bital. The next reevaluation was eight weeks later. The
dog had been noted to be quiet, with episodes of mild
dementia and loss of learned behavior. Again, the owner
had not observed any seizure activity in the dog. Right-
sided hemiparesis, conscious proprioception deficits, and
facial paresis were still present on neurological examina-
tion. The right eye was visual but exhibited a reduced
menace reaction.
Three weeks later, the dog represented after having
three generalized tonic-clonic seizures within 12 hours.
In the time preceding these seizures, the dog had become
lethargic and easily tired after exercise. The hair coat
had become soft and woolly with a lack of primary guard
hairs. The animal’s neurological status had not changed
since the previous reevaluation. The hair coat and atti-
tude changes that the dog exhibited raised the concern
about the possibility of pituitary hypofunction.
Hematology was normal at this time, but elevated
liver enzymes were detected on serum chemistry analy-
sis. Alkaline phosphatase serum levels were 1,186 U/L
(reference range, 10 to 73 U/L), and alanine aminotrans-
ferase serum levels were 474.3 U/L (reference range, 15
to 58 U/L). A fasting ammonia level was 46 µmol/l
(reference range, zero to 40 µmol/l). Skull radiographs
were repeated, demonstrating little change had taken
place in the 16 weeks since her trauma. Cisternal cere-
brospinal fluid (CSF) was obtained and found to be
abnormal on analysis. The protein content was 29 mg/dl
March/April 1999, Vol. 35
Table 1
Adrenocorticotrophic Hormone (ACTH)
Stimulation Test
Resting Cortisol Post-ACTH
Level Stimulation Level
(µg/dl) (µg/dl)
Test (1) 0.99 1.24
Test (2)* 0.29 0.84
Normal values 1–5 9–22
* Repeated ACTH stimulation test two weeks after
test (1)
(normal cisternal CSF protein level, less than 25 mg/dl).6
Forty white blood cells (WBCs) per µl and one red blood
cell (RBC) per µl were detected in the fluid (normal
cisternal CSF contains less than 5 WBCs/µl and zero to
30 RBCs/µl).6 Cytocentrifugation of the fluid enabled
identification of 4% neutrophils, 8% lymphocytes, and
88% mononuclear phagocytes. Possible differential di-
agnoses responsible for this abnormal CSF were
considered. Idiopathic meningoencephalitides or menin-
goencephalitis due to viral, protozoal, rickettsial, or fun-
gal infections were the primary differentials. Intracranial
neoplasia was also considered, as was a posttraumatic
inflammatory response; but the latter has not been well
documented in the literature.
Paired serum and CSF titers for canine distemper
virus immunoglobulin G (IgG) antibodies were positive
at a dilution of 1:50 and 1:10, respectively. Immunoglob-
ulin M (IgM) titers were negative for this virus. Serum
and CSF titers for Toxoplasma gondii IgG and IgM
antibodies, Neospora caninum IgG antibody, Ehrlichia
canis IgG antibody, Rickettsia rickettsii IgG antibody,
Borrelia burgdorferi antibody, Aspergillus sp. antibody,
and Cryptococcus neoformans antigen were negative.
The CSF distemper titers were considered to be due to
immunoglobulin leakage through the blood-brain bar-
rier, but confirmatory albumin quotient or an IgG index
was not performed.
An adrenocorticotropic hormone (ACTH) stimulation
test was performed using one vial (250 µg) of synthetic
aqueous ACTH. h Serum samples were taken for cortisol
evaluation immediately prior to and one hour after ACTH
was administered intravenously. The results were com-
patible with a diagnosis of hypoadrenocorticism [Table
1]. Electrolyte levels were rechecked the following day,
and these were normal. A serum phenobarbital level
taken two hours after the regular administration of phe-
nobarbital was 5.7 µg/ml, which was considered too low
to be effective. The dog was discharged three days after
admission in a stable state, with no seizures occurring
during the period of hospitalization. Potassium bromide
was prescribed for the dog at a dosage of 30 mg/kg body
Secondary Hypoadrenocorticism 119
Table 2
Endocrinological Evaluation of the Patient
Patient Results Normal Values
Canine serum TSH* 0.05 0.02–0.42
(ng/ml)
Triiodothyronine 70.0 75–200
(T3) (ng/dl)
Thyroxine (T4) 0.6 1.2–4.0
(µg/dl)
Somatomedin C 5.9 5.5–34.0
(nmol/l)
Plasma ACTH† 0.0 6.7–25.0
(pmol/l)
* Endogenous canine specific thyroid-stimulating
hormone
†
Adrenocorticotrophic hormone
weight, per os (PO) once daily. The owner agreed to
bring the dog back in two weeks to repeat the ACTH
stimulation test to confirm the diagnosis.
At the time of her reevaluation two weeks later, the
dog had not had any more seizures. Her neurological
examination again had not changed. The ACTH stimula-
tion was repeated, using the same method as that de-
scribed previously, and the results confirmed the
diagnosis of hypoadrenocorticism [Table 1]. Serum thy-
roxine (T 4), triiodothyronine (T3), and endogenous ca-
nine-specific thyroid-stimulating hormone (cTSH) levels
were evaluated [Table 2]. The T4 and T3 levels were low,
and the cTSH level was low normal. Adrenocorticotro-
pic hormone was undetectable on an endogenous plasma
ACTH assay [Table 2]. A somatomedin C plasma level
was low normal [Table 2].
The owner consented to a magnetic resonance imag-
ing (MRI) study of the dog’s brain [Figures 2, 3]. Mul-
tiple lesions were identified, including focal areas of
decreased signal intensity within the left frontal lobe on
T1-weighted images. This region demonstrated increased
signal intensity on T2-weighted images. The ventricles
were asymmetrical and dilated. Focal areas with a cystic
appearance were seen in the left midbrain at the level of
the lateral ventricles. Due to the lack of data regarding
MRI images of the canine pituitary region, no objective
evaluation could be made in this case. The left temporal
bone just caudal to the orbit was fractured, and the brain
parenchyma extended beyond the fracture margins. The
skeletal changes were compatible with a previous severe
trauma. The parenchymal changes were also thought to
be the result of trauma and secondary degenerative
changes. The dog was discharged after five days in the
hospital and given prednisone (0.25 mg/kg body weight,
q 24 hrs), with continuation of the phenobarbital and
potassium bromide administration as previously prescribed.
120 JOURNAL of the American Animal Hospital Association
Figure 2—Transverse T1-weighted image following Gd-DPTA
administration in the dog from Figure 1. A shallow concave
depression is seen in the left parietal bone. The signal intensity
of the cerebral tissue below the fracture is abnormal, with faint
wispy contrast uptake and areas of reduced signal intensity.
There is mild asymmetric dilatation of the right ventricle.
The dog returned for reevaluation five months later,
nearly nine months after the head trauma. The owner had
not noticed the dog having any further seizures. The dog
had gained 3 kg in body weight. The dog’s mentation
had improved, but the owner noticed periods of demen-
tia. A mild hemiparesis with mild proprioceptive deficits
were evident on the right side. The menace reaction was
present in both eyes but was slower on the right side.
Otherwise, the dog was normal and was discharged on a
reduced dose of prednisone (0.2 mg/kg body weight, PO
q 48 hrs). Since this time, the dog has been reevaluated
twice, at 12 and 18 months after the trauma. Both times
no physical changes could be identified on reevaluation.
The dog has been followed for three years, with no
evident deterioration in her condition.
Discussion
To the authors’ knowledge, there are no reports describ-
ing secondary hypoadrenocorticism or anterior hypopi-
tuitarism associated with craniocerebral trauma in the
dog. This report describes a case of hypoadrenocorticism
in a dog which was associated with severe traumatic
brain injury. Diagnosis of secondary hypoadreno-
corticism was based on a serum cortisol level that was
unresponsive to IV ACTH administration, an endog-
enous plasma ACTH assay indicating no detectable pro-
duction of ACTH from the pituitary, and normal serum
sodium and potassium levels. The plasma sample for the
ACTH level was handled appropriately to ensure accu-
rate results.7 Serum cortisol and plasma ACTH were
measured by radioimmunoassay as described previously.7–9
Secondary hypoadrenocorticism is due to a deficiency
of tropic hormone secretion, either corticotropin-releas-
ing hormone (CRH) or ACTH.10,11 These deficiencies
are due to disruption of the hypothalamic-pituitary axis
which has been documented to be caused by inflamma-
March/April 1999, Vol. 35
Figure 3—Sagittal T1-weighted image following Gd-DPTA
administration in the dog from Figure 1. Normal enhancement of
the pituitary is seen (arrowheads). A shallow concave defect is
seen in the parietal bone.
tion, tumors, and congenital defects in humans and ani-
mals, and trauma in humans.2,4,10 This is less common
than primary hypoadrenocorticism which is due to dis-
eases of the adrenal gland.10,11 However, iatrogenic sec-
ondary hypoadrenocorticism, which is the most common
cause for adrenal cortical nonresponsiveness in veteri-
nary practice, occurs following exogenous glucocorti-
coid administration.10,12 The ACTH stimulation test in
most dogs becomes normal within two weeks of steroid
withdrawal.10,13 The dog in this report had not received
exogenous glucocorticoids for seven weeks at the time
of the first ACTH stimulation and for nine weeks at the
time of the second test. Prednisone, prednisolone, hydro-
cortisone, and cortisone all can cross-react with serum
cortisol assays, especially if administered 24 hours be-
fore the test.11 This dog also had normal electrolyte
levels throughout her evaluations. Reports in the litera-
ture of hypoadrenocorticism have demonstrated that 11%
to 26% of affected dogs had serum concentrations of
sodium, potassium, or both that were within the refer-
ence ranges.14 It also has been noted that some dogs with
secondary hypoadrenocorticism have hyponatremia.11
Therefore, it can be seen that the differentiation of pri-
mary from secondary disease must rely on endogenous
plasma ACTH concentrations.10 Sample handling for this
hormone is critical, as it is more labile than cortisol.
Generally, samples must be drawn and centrifuged im-
mediately and stored frozen in plastic tubes.7,10 In sec-
ondary hypoadrenocorticism, a lack of stimulation of the
adrenal cortex by ACTH causes severe adrenal cortical
atrophy. This leads to low ACTH concentrations in the
plasma,10 as was seen in the dog described in this report.
The normal values for plasma ACTH range from 20 to
100 pg/ml.15 In primary disease, the plasma ACTH lev-
els are elevated, as no negative feedback from cortisol
occurs.10 This serves to confirm that the pituitary is
functioning.
March/April 1999, Vol. 35
Clinical signs exhibited by dogs with hypoadre-
nocorticism can be nonspecific and similar to manifesta-
tions of other more common diseases.16 In one study of
100 cases of hypoadrenocorticism, 77% of the dogs were
anorexic, 64% were lethargic or depressed, and 38%
demonstrated weakness.15 The clinician should consider
this diagnosis in any patient that presents with a waxing
and waning illness that responds to nonspecific treat-
ment such as cage rest and fluid administration.17 Inter-
mittent anorexia, weakness, and lethargy were features
of this case which could have been ascribed to the effects
of her trauma. The signs displayed generally will reflect
the deficiencies of mineralocorticoid and glucocorti-
coids.10 In this case, the only deficiencies were those of
glucocorticoids, the functions of which have been well
documented.10
The dog described here also had elevated hepatic
enzyme activity. Alanine aminotransferase elevations
were reported in approximately 30% of 225 dogs with
hypoadrenocorticism in one study.16 Twenty percent to
30% of these dogs also had high serum alkaline phos-
phatase activity. The precise mechanism for these abnor-
malities was unclear, although they likely were due to
hypovolemia causing hypoperfusion of the liver. It is not
inconceivable that they may in some cases have been due
to previous glucocorticoid treatment, which was a con-
cern in the case presented here.
Even though craniocerebral trauma disables about
50,000 humans a year, hypopituitarism associated with
brain injury is rare. 2 This may be due to the fact that two-
thirds of the anterior lobe of the pituitary must be de-
stroyed before clinical symptoms develop, and that
patients with severe anterior pituitary infarction also
have severe brain injury and usually do not survive.2 All
but two human patients out of a study of 22 with is-
chemic necrosis of the anterior lobe died within one
week of the injury.18 Valenta, et al., reported that is-
chemic or hemorrhagic lesions of the anterior hypothala-
mus occurred in as many as 42.5% of a total of 106
human cases of closed head injury and were shown to be
frequently associated with pituitary infarction.19 A delay
of weeks to months, sometimes years, from the time of
the injury to presentation exists in reported human cases;
rarely is the possibility of hypopituitarism considered at
the time of the trauma.4 One case was actually diagnosed
35 years after the patient’s head injury.4 Commonly, in
humans, the initial signs of hypopituitarism include those
of diabetes insipidus and hypogonadism (e.g., loss of
libido, amenorrhea, and impotence) and often are ac-
companied by weight loss and asthenia.4 In many cases,
these signs have been ascribed to the postconcussional
syndrome before the specific diagnosis of an endocrino-
logical abnormality. 4 In the authors’ case, the endocrin-
opathy was diagnosed approximately eight weeks after
the craniocerebral trauma. This time period could have
been considerably longer if it were not for the owner’s
Secondary Hypoadrenocorticism 121
compliance with the reevaluation schedule. It is pos-
sible, even though the dog had been historically clini-
cally normal before the trauma, that the endocrinopathy
could have been pre-existing. If serum cortisol levels
had been taken at the time of the trauma, low levels may
have been supportive of hypocortisolism. Another criti-
cism of this diagnosis could be that the endocrine tests
were not repeated at least four to six months after the
recovery in order to confirm no possible interference
from previous steroid use. Most authors, however, seem
to agree that after six weeks without steroid administra-
tion, the hypothalamic-pituitary-adrenal axis should not
be affected.10,13,14
The dog’s thyroid status was evaluated as described.
The T4 and T 3 levels were low. This was considered to be
compatible with anticonvulsant administration and pre-
vious glucocorticoid administration; but hypothyroidism
could not be ruled out, as false-negative cTSH results
occur in about 25% of hypothyroid dogs.20,21 In second-
ary hypothyroidism, levels of both T4 and cTSH would
be expected to be low. However, in some human patients
after head trauma, consistent mild elevations of thyroid-
stimulating hormone (TSH) levels have been observed in
the presence of low serum T4 levels.19 Exaggerated re-
sponses of TSH to thyrotropin-releasing hormone (TRH)
stimulation in the presence of low serum T4 levels have
also been reported, especially in a number of patients with
hypothalamic hypothyroidism.22,23 Some of the patients
in these studies demonstrated low or absent TSH bio-
logical activity due to impaired receptor binding.19
Usually the pituitary gland is well protected from
injury as it resides in the osseous sella turcica at the base
of the brain. Apart from the infundibulum or stalk, which
connects the pituitary with the median eminence of the
hypothalamus, the pituitary is covered by the fibrous
dura mater. The stalk serves as the major route for the
blood supply to the anterior pituitary.2,24 This ensures
that the anterior pituitary is extremely vascular, with its
blood supplied through the hypophyseal arteries which
communicate with the portal veins in the stalk.2 The
integrity of the stalk and its blood vessels are vital for the
stimulation of the glandular cells to produce the adeno-
hypophyseal hormones. 2 The case presented here cer-
tainly had a demonstrable cerebral lesion on MRI which
could have caused a reduction in the circulatory supply
to the anterior pituitary. One of the theories evident in
the human literature regarding hypopituitarism follow-
ing trauma is that the pituitary swells as a result of shock
and edema, but because it is confined by bone and dura,
only the stalk area can expand. This would cause com-
pression of the portal vessels and prevent blood supply
to the anterior pituitary, causing infarction.2,18 If the
pituitary stalk is sectioned, about 90% of the anterior
lobe becomes infarcted.4 The blood supply to the poste-
rior lobe is not affected by this procedure or when the
stalk is ruptured by head injury.4
122 JOURNAL of the American Animal Hospital Association
Indirect evaluation of growth hormone (GH) concen- 3.
tration can be performed by measurement of somatome-
4.
din C (insulin-like growth factor I [IGF-I]).25 Growth
hormone stimulates the secretion of IGF-I, which then 5.
mediates most of growth hormone’s anabolic effects.26
6.
Insulin-like growth factor I is synthesized in the liver, is
secreted under the direct control of GH,26,27 and disap- 7.
pears from the blood in GH-deficient humans.26,28 The
mean concentrations of IGF-I in dogs have been shown 8.
to vary with the breed, with smaller breeds having lower
concentrations than the larger breeds.26,29 A linear corre- 9.
lation exists between body size and circulating IGF-I.29
The dog had a low normal IGF-I level [Table 2]. A GH
stimulation test was not done in this case. All pituitary
10.
trauma cases in one human study had no GH response to
stimulation.4 Selective GH deficiency has also been seen
in humans after severe head trauma.30 The possibility 11.
exists that the authors’ case also exhibited physiological
12.
GH deficiency, especially considering the hair coat
changes. 13.
Treatment of secondary hypoadrenocorticism requires
only glucocorticoids.7,10,11,14,17 Previously recommended 14.
dosages have started at 0.5 mg/kg body weight per day.10
This should be tapered to the lowest level needed to
15.
control the clinical signs. The authors’ case was started
on prednisone at 0.25 mg/kg body weight per day which
was lowered to 0.2 mg/kg body weight every other day. 16.
She has been maintained on this dose since that time,
which the owner reports is the lowest dose that achieves 17.
the optimal effect. The long-term prognosis for animals
18.
with hypoadrenocorticism is usually excellent, although
high-stress situations such as illness or surgery may cause 19.
an increase in the demand for glucocorticoids.10 The
prognosis for canine cases of posttraumatic hypoadreno- 20.
corticism is unknown as yet. The dog has survived for
three years at the time of writing. 21.
22.
a
Phenobarbital sodium for injection 65 mg/ml
b
Solu-Medrol; Upjohn, Kalamazoo, MI 23.
c with central hypothyroidism. Evidence for reduced biological activity of
Meticorten; Schering Animal Health, Union, NJ
d immunoreactive thyrotropin. J Clin Endocrinol Metab 1979;48:989–98.
Zantac; Glaxo Wellcome, Research Triangle Park, NC
e 24.
Torbutrol; Fort Dodge Laboratories, Fort Dodge, IA
f
Kefzol; Lilly, Indianapolis, IN 25.
g syndromes in dogs and cats. In: Bonagura JD, Kirk RW, eds. Current
Cefa-Tabs; Fort Dodge Laboratories, Fort Dodge, IA
h veterinary therapy XI. Philadelphia: WB Saunders, 1992:322–7.
Cosyntropin; Organon Pharmaceuticals, West Orange, NJ
26.
Acknowledgments
27.
The authors wish to thank Dr. Michael Schaer for his
contributions to the case and review of the paper.
28.
References 29.
1. Chesnut RM, Marshall LF. Management of severe head injury. In: Ropper
AH, ed. Neurological and neurosurgical intensive care. New York: Raven 30.
Press, 1993:203–40.
2. Klingbeil GEG, Cline P. Anterior hypopituitarism: a consequence of head
injury. Arch Phys Med Rehabil 1985;66:44–6.
March/April 1999, Vol. 35
Clark JDA, Raggatt PR, Edwards OM. Abnormalities of the hypothalamo-
pituitary-gonadal axis after head injury. Clin Endocrinol 1992;36:481–5.
Edwards OM, Clark JDA. Post-traumatic hypopituitarism. Med 1986;65:
281–90.
Podell M. Seizures in dogs. Vet Clin N Am Sm Anim Pract 1996;26:
779–809.
Braund KG. Clinical syndromes in veterinary neurology. 2nd ed. St. Louis:
Mosby-Year Book, 1994:368–75.
Melian C, Peterson ME. Diagnosis and treatment of naturally occurring
hypoadrenocorticism in 42 dogs. J Sm Anim Pract 1996;37:268–75.
Peterson ME, Orth DN, Halmi NS, et al. Plasma immunoreactive
proopiomelanocortin peptides and cortisol in normal dogs and dogs with
Addison’s disease and Cushing’s syndrome. Endocrinol 1986;119:720–30.
Peterson ME, Kemppainen RJ, Orth DN. Effects of synthetic ovine
corticotropin-releasing hormone on plasma concentrations of immunoreac-
tive adrenocorticotropin, alpha- melanocyte-stimulating hormone, and
cortisol in dogs with naturally acquired adrenocorticol insufficiency. Am J
Vet Res 1992;53:421–5.
Hardy RM. Hypoadrenal gland disease. In: Ettinger SJ, Feldman EC, eds.
Textbook of veterinary internal medicine. Philadelphia: WB Saunders,
1995:1579–93.
Kintzer PP, Peterson ME. Primary and secondary canine hypoadreno-
corticism. Vet Clin N Am Sm Anim Pract 1997;2:349–57.
Schaer MS, Chen CL. A clinical survey of 48 dogs with adrenocortical
hypofunction. J Am Anim Hosp Assoc 1983;19:443–52.
Moore GE, Hoenig M. Duration of pituitary and adrenocortical suppression
after long-term administration of anti-inflammatory doses of prednisone in
dogs. Am J Vet Res 1992;53:716–20.
Lifton SJ, King LG, Zerbe CA. Glucocorticoid deficient hypoadreno-
corticism in dogs: 18 cases (1986–1995). J Am Vet Med Assoc 1996;209:
2076–81.
Feldman EC, Nelson RW. Hypoadrenocorticism. In: Feldman EC, Nelson
RW, eds. Canine and feline endocrinology and reproduction. Philadelphia:
WB Saunders, 1987:195–217.
Peterson ME, Kintzer PP, Kass PH. Pretreatment clinical and laboratory
findings in dogs with hypoadrenocorticism: 225 cases (1979–1993). J Am
Vet Med Assoc 1996;208:85–91.
Kintzer PP, Peterson ME. Treatment and long-term follow-up of 205 dogs
with hypoadrenocorticism. J Vet Int Med 1997;11:43–9.
Kornblum RN, Fisher RS. Pituitary lesions in craniocerebral injuries. Arch
Path 1969;88:242–8.
Valenta LJ, De Feo DR. Post-traumatic hypopituitarism due to a
hypothalamic lesion. Am J Med 1980;68:614–7.
Boothe DM. Effects of drugs on endocrine tests. In: Bonagura JD, Kirk
RW, eds. Current veterinary therapy XII. Philadelphia: WB Saunders,
1995:339–46.
Nichols R. Update: diagnostic testing for canine hypothyroidism. Proceed,
15th Am Coll Vet Int Med Forum, May 1997:243–5.
Dzur J, Winternitz WW. Posttraumatic hypopituitarism: anterior pituitary
insufficiency secondary to head trauma. So Med J 1976;69:1377–9.
Faglia G, Bitensky L, Pinchera A, et al. Thyrotropin secretion in patients
Winternitz WW, Dzur JA. Pituitary failure secondary to head trauma. Case
report. J Neurosurg 1976;44:504–5.
Randolph JF, Peterson ME. Acromegaly (growth hormone excess)
Feldman EC, Nelson RW. Disorders of growth hormone. In: Feldman EC,
Nelson RW, eds. Canine and feline endocrinology and reproduction.
Philadelphia: WB Saunders, 1996:38–66.
Daughaday WH. The anterior pituitary. In: Wilson JD, Foster DW, eds.
Williams’ textbook of endocrinology. Philadelphia: WB Saunders,
1985:568–76.
Underwood LE, D’Ercole AJ, Van-Wyk JJ. Somatomedin-C and the
assessment of growth. Pediatr Clin N Am 1980;27:771–82.
Eigenmann JE, Patterson DF. Growth hormone deficiency in the mature
dog. J Am Anim Hosp Assoc 1984;20:741.
Lopez-Guzman A, Salvador J, Albero R, et al. Selective growth hormone
deficiency of hypothalamic origin following severe head injury. Acta
Paediatr 1992;81:698–9.
March/April 1999, Vol. 35 Secondary Hypoadrenocorticism 123

Synbiotics
4C Ad
Pull from Feb Trends, cover 3

page 123
AAHA Hospital Membership,
B & W ad,
pull trends, p. 46

p.124
 Babesia gibsoni Infections in Dogs
From North Carolina
The recognition of canine babesiosis in North Carolina caused by Babesia gibsoni documents
the expansion of the previously reported endemic area of this disease. Clinical signs ranged
from severe hemolytic anemia and thrombocytopenia to subclinical infections. No infected dogs
had traveled to endemic areas. Antibabesial treatment failed to eradicate the organism from
infected dogs. J Am Anim Hosp Assoc 1999;35:125–8.

Adam J. Birkenheuer, DVM Introduction

Michael G. Levy, PhD
Karine C.M. Savary, DVM
Robin B. Gager
Edward B. Breitschwerdt, DVM
C area was first reported in 1979.6 To the authors’ knowledge, this parasite
Babesia gibsoni (B. gibsoni) is a tick-transmitted, intraerythrocytic proto-
zoan of wild and domestic canids in Asia, Africa, Europe, the Middle
East, and North America. This parasite causes a potentially fatal hemolytic
anemia in dogs. The disease manifestations have been reviewed in detail
recently.1,2 Treatment of or spontaneous recovery from an acute infection
frequently fails to clear the organism from the host, resulting in a carrier
stage. Recovered animals are a reservoir for tick-transmitted infections
and are at risk for recrudescent infection.2–5 Babesia gibsoni infection in a
dog indigenous to the United States that had not traveled to an endemic
has not been reported from North Carolina or the southeastern United
States. Nine cases of B. gibsoni infection have been recently diagnosed in
dogs from North Carolina.

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.
Address all correspondence to
Dr. Birkenheuer.
Case Reports
Case Nos. 1–3
A nine-year-old intact male American Staffordshire terrier (case no. 1)
from eastern North Carolina was referred to North Carolina State Univer-
sity Veterinary Teaching Hospital (NCSUVTH) for evaluation of a rap-
idly progressing hemolytic anemia. Several days prior to presentation, the
owner noticed that the dog was weak, had pale mucous membranes, and
was not interested in breeding. The dog had a history of substantial tick
exposure and had no known history of drug or toxin exposure. Upon
physical examination, the dog was thin, weak, febrile, moderately
depressed, and had pale mucous membranes, moderate peripheral lym-
phadenopathy, and splenomegaly. Fundoscopic examination was normal.
Clinicopathological abnormalities included a macrocytic, normochromic
regenerative anemia (packed cell volume [PCV], 18%; reference range,
33% to 58%; red blood cell [RBC] count, 1.72 x10 6 cells/µl; reference
range, 4.78 to 8.26 x106 cells/µl; mean corpuscular volume [MCV], 88.8
fl; reference range, 63.9 to 72.8 fl; mean corpuscular hemoglobin concen-
tration [MCHC], 34.6 g/dl; reference range, 33.6 to 36.4
g/dl; aggregate reticulocytes, 10.9%; reference range, 0.0% to 1.5%;
nucleated RBCs, 43/100 white blood cells [WBCs]). A profound throm-
bocytopenia also was detected (platelet count, 31 x103/µl; reference range,
181 to 35 x103/µl). Total WBCs were within the laboratory reference
range; however, there was a mild left shift (WBC count, 15.1 x103 cells/
µl; reference range, 6.4 to 15.8 x103 cells/µl; segmented neutrophils, 9.9
x103 cells/µl; reference range, 3.0 to 11.5 x103 cells/µl; bands, 0.453 x103
cells/µl; reference range, 0.0 to 0.3 x103 cells/µl), mild eosinophilia

JOURNAL of the American Animal Hospital Association 125

126 JOURNAL of the American Animal Hospital Association
Figure 1—Photomicrograph of a modified Wright’s stained thin
blood smear. Babesia gibsoni organisms are located inside of
the canine erythrocytes (arrows) (800X; bar=5 µm).
(eosinophils, 0.755 x103 cells/µl; reference range, 0.1 to
0.75 x103 cells/µl), and basophilia (basophils, 0.151 x103
cells/µl; reference range, 0.0 to 0.1 x103 cells/µl). Babe-
sia gibsoni organisms were seen on peripheral blood
smears stained with a modified Wright’s stain [Figure
1]. Serum biochemical abnormalities included hypoal-
buminemia (albumin, 2.2 mg/dl; reference range, 2.8 to
4.4 mg/dl), a mild increase in alanine aminotransferase
activity (ALT, 63 IU/L; reference range, 0 to 45 IU/L),
and a mild increase in blood urea nitrogen concentration
(BUN, 27 mg/dl; reference range, 8 to 24 mg/dl). Uri-
nalysis identified a profound bilirubinuria (4+ bilirubin;
urine specific gravity 1.035). A coagulation profile was
unremarkable other than the thrombocytopenia. A heart-
worm antigen test and fecal flotation were negative.
Thoracic and abdominal radiographic abnormalities in-
cluded hepatosplenomegaly and sternal lymphadenopa-
thy. Initial therapy was started with intravenous (IV)
lactated Ringer’s solution and doxycycline hyclatea (10
mg/kg body weight, per os [PO] qid) to treat a possible
concurrent rickettsial infection. The PCV (10%) contin-
ued to decrease rapidly during the next 12 hours. Packed
RBCs from a cross match-compatible donor were ad-
ministered. The PCV after transfusion was 17%. On day
two, the antibabesial drug diminazene aceturate b
(5 mg/kg body weight, intramuscularly [IM] once) was
given. Reciprocal indirect fluorescent antibody (IFA)
titers to B. gibsoni, B. canis, and Erhlichia canis (E.
canis) antigens were 1,280, 160, and 160, respectively.
A direct Coombs’ test was positive. The PCV (15%)
continued to decrease during the next 12 hours and the
serum became profoundly icteric. Immunosuppressive
doses of prednisonec (1 mg/kg body weight, PO bid) were
administered to decrease the immune-mediated destruc-
tion of erythrocytes.3,7,8 The dog began to improve
clinically during the next several days, but remained
persistently parasitemic. A second dose of diminazine
aceturate at the same dosage was administered on day
March/April 1999, Vol. 35
eight. On day 11, the dog was discharged. The PCV was
28%, and an adequate platelet count (greater than 200
x103/µl) was estimated. At discharge, doxycycline (for
the treatment of E. canis), misoprostold (4 mcg/kg body
weight, PO bid), and prednisone were dispensed with
instructions to taper the prednisone over the next six
weeks. Several months after discharge, the dog was clini-
cally healthy and in excellent body condition. The PCV
was 45%. However, persistent parasitemia was identi-
fied on peripheral blood smears. Due to the lack of an
efficacious drug for the eradication of B. gibsoni organ-
isms from infected dogs, the owner elected not to re-treat
the dog.
Case no. 1 was from a kennel that housed a large
number (more than 50) of dogs which were mostly
American Staffordshire terriers and American pit bull
terriers. Venous blood samples were collected from 10
randomly selected dogs to examine blood smears for the
presence of B. gibsoni organisms. Babesia gibsoni or-
ganisms were observed on blood smears from two of the
10 dogs. One of these dogs (case no. 2), a 10-year-old
intact female American pit bull terrier, was anemic (PCV,
16%) and had lymphadenopathy and splenomegaly. This
dog was not treated and died at home several weeks later.
No necropsy was performed. Case no. 3 was a seven-
year-old intact male American pit bull terrier that was
clinically healthy and had a normal PCV (42%). The
owner elected to not treat this dog with antibabesial
drugs, and at the time of this report this dog remains
clinically healthy. The owner declined serological test-
ing of the entire kennel, as it was deemed cost
prohibitive.
Case Nos. 4–8
Case no. 4 was a three-year-old intact female American
pit bull terrier examined by a veterinary practitioner in
western North Carolina for an acute hemolytic crisis four
to five weeks postparturition. Blood samples were sent
to the North Carolina State University Infectious
Disease Laboratory for blood parasite screening and se-
rological testing. Babesia gibsoni piroplasms were
observed on peripheral blood smears, and a reciprocal
IFA titer to B. gibsoni antigens was 160. The reciprocal
IFA titer to B. canis was negative. At five weeks of age,
four of five puppies in this litter (case nos. 5–8) had
circulating B. gibsoni organisms on peripheral blood
smears. Antibodies to B. gibsoni antigens were not de-
tectable in sera from the puppies. Only one puppy (case
no. 5) was anemic (PCV, 16%). Case nos. 6–8 were
clinically normal and had normal PCVs. Case nos. 4–8
were lost to follow-up.
Case No. 9
A 17-year-old spayed female mixed breed dog from
central North Carolina was referred to NCSUVTH be-
cause circulating babesial organisms were noted on a
March/April 1999, Vol. 35
peripheral blood smear. This dog had a chronic history
of severe bronchitis, mitral insufficiency, cough, dysp-
nea, and obesity. Case no. 9 presented to NCSUVTH
with a two-week history of lethargy, anorexia,
vomiting, and diarrhea. Clinicopathological abnormali-
ties included thrombocytopenia (platelet count,
40 x10 3/µl) and a neutrophilic leukocytosis with a left
shift (WBC count, 34.6 x10 3 cells/µl; bands, 2.7 x103
cells/µl) and monocytosis (monocyte count, 2.0 x103
cells/µl; reference range, 0.15 to 1.35 x103cells/µl). Se-
rum biochemical abnormalities included increased alka-
line phosphatase (ALP) (ALP, 1177 IU/L; reference
range, 16 to 71 IU/L), ALT (85 IU/L; reference range,
0 to 40 IU/L), and a mild hypoalbuminemia (albumin,
2.7 mg/dl). Morphological examination of the organisms
confirmed them to be B. gibsoni. Reciprocal antibody
titers to B. gibsoni and B. canis were 640 and 160,
respectively. Case no. 9 had never traveled out of North
Carolina. Diminazine aceturate was administered (3.5
mg/kg body weight, IM once). The platelet count
returned to normal (210 x10 3/µl) after antibabesial
treatment. The other clinical signs did not resolve, and the
dog required supportive care on a daily basis. The dog
was euthanized about one month after discharge because
of acute respiratory distress suspected to be due to heart
failure. No necropsy was performed.
Discussion
Babesia gibsoni is a rapidly emerging pathogen in the
United States that recently has become endemic to the
southwestern region of the country. These cases repre-
sent the first reports of disease caused by B. gibsoni in
the eastern United States since 1979. In endemic areas
outside of the United States, B. gibsoni is transmitted by
the ixodid ticks, Haemaphysalis bispinosa and Haema-
physalis longicornis.9,10 These ticks are not known to
exist in the United States. No ticks endemic to the United
States are known to transmit B. gibsoni. Experimental
evidence supports Rhipicephalus sanguineus as a
potential vector, but transmission studies for both R.
sanguineus and Dermacentor variabilis have been un-
successful or inconclusive. In the study that reported
successful transmission, tick-proof kennels were not used
and some animals acquired incidental B. canis infec-
tions.11–13 None of the dogs in this report had traveled to
known endemic areas for B. gibsoni. The infections in
the adult dogs from North Carolina support the presence
of a competent tick vector in the United States.1,2,14 No
ticks were noted on any of the infected animals at the
time of presentation. More transmission studies are
needed to evaluate the vector competency of the known
endemic tick population. Investigations are warranted to
evaluate the species of the tick populations in the United
States, as it is possible that an exotic vector has been
introduced to North America. The route of infection for
the puppies was not determined. Previous reports have
Babesia gibsoni 127
suggested a transplacental mode of transmission.15 Since
the pre-patent period following experimental tick-trans-
mitted infections is about one to three weeks, the puppies
in this report could have been infected either transpla-
centally or as a result of tick exposure. 9,10
No drugs have been proven to be effective for the
elimination of B. gibsoni organisms from infected dogs.
Some antibabesial drugs can reduce the severity of
clinical signs and the mortality associated with the
disease. These drugs include diminazene aceturate,
imidocarb diproprionate, phenamidine isethionate, pen-
tamidine isethionate, parvaquone, and niridazone.1–3,16–
18
Unfortunately, none of these drugs are approved
currently for use in dogs in the United States. Of the
drugs that currently are available in the United States,
metronidazole may have shown some efficacy in
decreasing the mortality in infected dogs, although it
was not as effective as diminazene aceturate, and no
untreated control group was used in the study.17 Dogs
that survive an acute crisis develop a state of premuni-
tion.1,3,4 Premunition is referred to as the immunity of
infection and is a delicate balance between the host
immune response and the parasite’s ability to induce
clinical disease. Dogs existing in a state of premunition
are at risk for recrudescence and are a potential reservoir
for tick infection.1,4,5 Rigorous tick control and the tar-
geting of carrier animals are considered the best methods
of disease prevention and control.2
Babesia gibsoni infections can be misdiagnosed eas-
ily as idiopathic or autoimmune hemolytic anemia, as
organisms are not always observed at presentation.3 The
best way to definitively diagnose the disease is the visu-
alization of the parasite in erythrocytes by light micros-
copy. It has been reported that Geimsa, Romanowsky,
and Field’s stains are best for organism identification,
although in the authors’ experience commercially avail-
able modified Wright’s stains were adequate for the
visualization and identification of the organism. Babesia
gibsoni is a smaller, more pleomorphic organism than
B. canis. The trophozoites of B. gibsoni are often annu-
lar, oval, or signet ring forms, not like the classic paired
piriform-shaped organisms seen with B. canis.19 Other
methods of diagnosis include lysis centrifugation and
filtration, IFA, and enzyme-linked immunosorbent anti-
body (ELISA) tests.20–23 A recent report suggested that a
reciprocal immunofluorescent antibody titer of 320 or
greater is considered to be appropriate for serodiagnosis
if the clinical signs of the patient are consistent with
B. gibsoni infection. A higher titer of 1,280 or greater
was recommended for seroepidemiological studies.22 Ti-
ters of this magnitude will reduce the number of false
positive tests if IFA is the only diagnostic test being
evaluated. However, it is important that some dogs (e.g.,
case no. 4) may have titers below that level and that
some cases may have negative titers such as those seen
in case nos. 5–8.3 The ELISA is sensitive, but not spe-
128 JOURNAL of the American Animal Hospital Association
cific due to cross reactivity with anti-B. canis antibodies.e References
No ELISA test is commercially available for B. gibsoni. 1.
Molecular techniques such as polymerase chain reaction
2.
(PCR) amplification of B. gibsoni deoxyribonucleic acid
(DNA) are not currently available as routine diagnostic 3.
tests. Currently the combination of history, clinical signs,
evaluation of thin-stained blood smears, and IFA titer
4.
are the diagnostics of choice to evaluate dogs that are
suspected of having B. gibsoni infections. 5.
Eight of the nine cases reported from North Carolina
6.
were from kennels that maintained multiple dogs. At
least one of the kennels moved dogs on and off the 7.
premises and introduced new dogs with no quarantine
period. A significant difference in the seroreactivity
8.
against B. canis antigens associated with lifestyle, hous-
ing conditions, and climate has been reported, with dogs 9.
living in kennel situations having more exposure than
house pets.24,25 Dogs that are kept in kennel situations 10.
are presumed to have more exposure to ticks and subse-
quently more exposure to tick-transmitted pathogens such 11.
as B. gibsoni. Further studies regarding the epidemiol-
ogy of B. gibsoni infections in the United States are 12.
warranted.
The number of reports of B. gibsoni infections in the 13.
United States is increasing. This may be due to an in-
crease in the incidence of the disease or may be due to a 14.
greater awareness among clinicians with subsequent rec-
ognition of the disease. Babesia gibsoni should be con- 15.
sidered as a differential diagnosis by clinicians in the
United States as a cause of hemolytic anemia, throm- 16.
bocytopenia, and fever. Client education should address 17.
the efficacy of treatment, the possibility of recrudes-
cence, disease prevention for other pets through tick 18.
control, and the possibility of recovered animals acting
as reservoir hosts. If a competent tick vector exists in 19.
North America, B. gibsoni poses a significant health
threat to dogs on this continent. 20.
21.
a
Doxycline hyclate; Danbury Pharmacal, Inc., Florham Park, NJ
b
Berenil; Hoechst Veterinary, Munich, Germany 22.
c fluorescent antibody test for diagnosis of Babesia gibsoni infections in
Prednisone; Roxane Laboratories, Inc., Columbus, OH
d dogs. Am J Vet Res 1993;54(10):1579–84.
Cytotec; G.D. Searle & Co., Chicago, IL
e 23.
To M. 1989. Basic studies on the screening test for the development of new
drugs against canine babesiosis. Nippon Vet Anim Sci Univ, Thesis.
f 24.
Imizol; Schering-Plough Animal Health Corp., Kenilworth, NJ
25.
Addendum
Since the original submission, the authors have identi-
fied two additional dogs infected with B. gibsoni. One
dog is from North Carolina, and the other is from
Wisconsin. In addition, Imidocarb diproprionatef has
since been approved for the treatment of canine babesio-
sis in the United States.
March/April 1999, Vol. 35
Conrad PA, Thomford J, Yamane I, et al. Hemolytic anemia caused by
Babesia gibsoni infections in dogs. J Am Vet Med Assoc 1991;199:601–5.
Yamane I, Conrad PA, Gardner I. Babesia gibsoni infections in dogs.
J Protozool Res 1993;3:111–25.
Farwell GE, LeGrand EK, Cobb CC. Clinical observations on Babesia
gibsoni and Babesia canis infections in dogs. J Am Vet Med Assoc
1982;180:507–11.
Masuda T, Baba E, Arakawa A. Relapse of canine babesiosis after
prednisolone treatment. Mod Vet Pract 1983;Nov:931–2.
Groves MG, Dennis GL. Babesia gibsoni: field and laboratory studies of
canine infections. Exp Parasitol 1972;32:153–9.
Anderson JF, Magnarelli LA, Donner CS, et al. Canine Babesia new to
North America. Science 1979;204:1431–2.
Adachi K, Tateishi M, Horii Y, et al. Elevated erythrocyte-bound IgG in
dogs with clinical Babesia gibsoni infection. J Vet Med Sci 1994;56(4):
757–9.
Onishi T, Ueda K, Horie M, et al. Serum hemolytic activity in dogs
infected with Babesia gibsoni. J Parasitol 1990;76(4):564–7.
Otsuka H. Studies on transmission of Babesia gibsoni Patton by
Haemaphysalis longicornis Neumann. Bull Fac Agric Miyazaki Univ
1974;21:359–67.
Swaminath CS, Shortt HE. The arthropod vector of Babesia gibsoni. Ind J
Med Res 1937;25:499–503.
Higuchi S, Fujimori M, Hoshi F, et al. Development of Babesia gibsoni in
the salivary glands of the larval tick, Rhipicephalus sanguineus. J Vet Med
Sci 1995;57(1):117–9.
Sen SK. The vector of canine piroplasmosis due to Piroplasma gibsoni.
Ind J Vet Sci Anim Husb 1933;3:356–63.
Yamane I, Gardner I, Telford S, et al. Vector competence of Rhipicephalus
sanguineus and Dermacentor variabilis for American isolates of Babesia
gibsoni. Exp Appl Acarol 1993;17(12):913–9.
Yamane I, Gardner I, Ryan C, et al. Serosurvey of Babesia canis, Babesia
gibsoni and Ehrlichia canis in pound dogs in California, USA. Prev Vet
Med 1994;18:293–304.
Abu M, Hara I, Naito I, et al. Babesia infections in puppies: probably due
to transplacental transmission. J Vet Med 1973;609:203–6.
Groves M, Vanniasingham J. Treatment of Babesia gibsoni infections with
phenamidine isethionate. Vet Rec 1970;86:8–10.
Fowler J, Ruff M, Fernau R, Furusho Y. Babesia gibsoni: chemotherapy in
dogs. Am J Vet Res 1972;33(6):1109–14.
Ruff M, Fowler J, Fernau R, Matsuda K. Action of certain antiprotozoal
compounds against Babesia gibsoni in dogs. Am J Vet Res 1973;34(5):
641–5.
Soulsby E. In: Soulsby E, ed. Helminths, arthropods and protozoa of
domesticated animals. 7th ed. Philadelphia: Lea & Febiger, 1982:723–8.
Adachi K, Watanabe T, Yamane S, et al. Isolation of Babesia gibsoni
piroplams from infected erythrocytes of dogs. J Vet Med Sci 1993;55(3):
487–90.
Anderson J, Magnarelli L, Sulzer A. Canine babesiosis: indirect fluorescent
antibody test for a North American isolate of Babesia gibsoni. Am J Vet
Res 1980;41:2102–5.
Yamane I, Thomford J, Gardner I, et al. Evaluation of the indirect
Chang G, Tu C. A serological survey of canine babesiasis in Taiwan.
J Chin Soc Vet Sci 1992;18(3):125–31.
Levy M, Breitschwerdt E, Moncol D. Antibody activity to Babesia canis in
dogs in North Carolina. Am J Vet Res 1987;48:339–41.
Taboada J, Harvey J, Levy M. Seroprevalence of babesiosis in greyhounds
in Florida. J Am Vet Med Assoc 1992;200(1):47–50.
 Interlocking Box Jejunostomy:
A New Technique for Enteral Feeding
A new jejunostomy tube placement technique is described and compared to traditional
methods. The interlocking box technique was compared to a simple purse-string with
jejunopexy and an inverting serosal tunnel (Witzel technique) with jejunopexy. Procedures
performed in fresh canine cadavers were fluid pressure tested following tube removal.
Mean±standard deviation (SD) intraluminal pressure necessary to induce leakage at the
jejunopexy site of the interlocking box group (87.63±40.56 cm H2O) was significantly greater
(p less than 0.001) than the simple purse-string (43.17±31.69 cm H2O) and serosal tunnel
(46.33±23.60 cm H2O) groups. Significant differences were not identified between the latter
groups. The interlocking box technique resisted leakage following acute removal better than
conventional techniques and should be tested clinically. J Am Anim Hosp Assoc 1999;35:129–34.

R. Mark Daye, DVM Introduction

Michael L. Huber, DVM, MS
Ralph A. Henderson, DVM, MS
O major organ failure.1,11 Reported indications for enteral feeding include
From the Department of Small Animal
Surgery and Medicine,
Small Animal Clinic,
Auburn University,
Auburn, Alabama 36849-5523.
Enteral feeding, including appetite stimulation, force feeding, or tube
feeding, has become a mainstay of critical care management.1–10 Malnu-
trition in the critically ill or injured patient has been associated with
immunosuppression, delayed wound healing, wound dehiscence, muscu-
lar weakness, and increased susceptibility to shock, hospital stress, and
interruption of oral feeding for three to five days; acute weight loss
greater than 5%; chronic weight loss greater than 10%; generalized muscle
wasting; and disease conditions known to increase protein loss, induce
hypermetabolism, suppress appetite, or interfere with nutrient ingestion,
transit, or absorption.5,6,10
Infusion of nutrient monomeric or polymeric solutions directly into the
proximal small bowel is termed enterostomy feeding. Jejunostomy feed-
ing is more common than duodenostomy feeding and is indicated when
other forms of enteral nutrition are ineffective or contraindicated.7,10 In
small animal patients, jejunostomy feeding has been utilized in the therapy
of severe but tractable vomiting, hepatic lipidosis, severe pancreatitis,
inflammatory bowel disease, abdominal trauma, hiatal hernia, prostatic
abscessation, metritis, and following gastric, small intestinal, or
hepatobiliary surgery.6,8,9
Complications associated with jejunostomy tube feeding include ani-
mals chewing at their tubes, tube obstruction, premature removal or
dislodgment of tubes, surgical dehiscence of the jejunopexy stoma site,
stoma site discharge, vomiting, and diarrhea.2,6,8,9,12 Dislodgment and
stoma site dehiscence can permit infusion/leakage of nutrient solutions or
intestinal contents into the abdominal cavity, resulting in peritonitis.6,7,9
Peritonitis, the most severe complication of jejunostomy feeding, can lead
to serious morbidity and mortality and may necessitate further surgical
intervention.9
Placement of jejunostomy tubes during routine intra-abdominal proce-
dures is not performed because current recommendations are to keep the
tube in situ for a minimum of five days. 1,8,9,12 This period allows the
formation of firm adhesions at the jejunopexy site and is intended to

JOURNAL of the American Animal Hospital Association 129

130 JOURNAL of the American Animal Hospital Association
reduce the risk of intraperitoneal leakage following tube
removal. In patients with compromised healing, even
longer periods may be prudent. Increased complications
and duration of treatment associated with current jeju-
nostomy techniques limit placement to only animals that
clearly require enteral nutrition postoperatively. Unfor-
tunately, it can be difficult to identify this subgroup of
patients. These factors lead to the underutilization of
jejunostomy tubes in animals that may benefit from nu-
tritional support.
An improved jejunostomy tube placement technique,
which allows early (before five days) removal without
risk of ingesta leakage, would encourage more wide-
spread use of jejunostomy tubes in a variety of patients
undergoing laparotomy. Postoperatively, those patients
which returned to oral intake could have their tubes
pulled soon following surgery. Those patients unable or
unwilling to return to oral intake would benefit from
enteral feeding via the surgically placed jejunostomy
tube. Such a technique should prevent nutrient peritoni-
tis secondary to jejunostomy tube dislodgment and might
encourage common use of immediate (within hours) post-
operative enteral nutrition, which was beneficial to in-
testinal wound healing in experimental animals and
beneficial in the treatment of pediatric human surgical
patients.13,14
A jejunostomy tube technique that may permit early
removal was conceived by incorporating the jejunos-
tomy tube through a jejunopexy taking the form of an
interlocking box suture pattern (interlocking box). This
method was tested in fresh cadavers against two conven-
tional jejunostomy tube techniques by generating in-
traluminal fluid pressures in bowel segments following
acute tube removal and measuring leakage resistance at
the jejunopexy site. The authors hypothesized that the
interlocking box would resist significantly higher in-
traluminal pressures than conventional techniques.
Materials and Methods
Three jejunostomy techniques were performed in ran-
dom order in each of 12 fresh canine cadavers. Groups
included the interlocking box, the simple purse-string,
and the inverting serosal tunnel. Conventional techniques
were chosen based on common usage by surgeons at
educational institutions and private referral practices (i.e.,
authors’ personal communications). Cadavers were ob-
tained following euthanasia for reasons not related to
this study. Animals were clinically healthy antemortem
based on physical examination. No gross evidence of
disease of either the body wall or intestine was noted
during specimen preparation. Jejunostomy tube place-
ment and testing were completed within two hours of
euthanasia. Two surgeons performed an equal number
and distribution of procedures.
Via a ventral midline celiotomy, an 8-French cathetera
was pulled through a stab incision in the body wall with
March/April 1999, Vol. 35
curved Kelly forceps. It then was advanced 20 to 25 cm
aborad in the jejunal lumen through a 3-mm, anti-
mesenteric stab incision. A portion of proximal jejunum
that was apposed to the body wall was easily employed.
The first jejunostomy tube site in each animal was su-
tured to the body wall approximately 5 cm caudal to the
costal arch and 3 to 5 cm lateral to the ventral midline
celiotomy incision. Once the placement and jejunopexy
were complete, the jejunostomy tube was removed im-
mediately and the next technique was performed. The
subsequent techniques in each subject were performed
with an adjacent loop of proximal jejunum and sutured to
the body wall 3 to 5 cm caudal to the preceding tech-
nique. All suturing was performed with polydioxanone
sutureb on a tapered needle, with four throws on all knots
and suture tags cut at 3 to 4 mm.
Following completion of all three techniques, each
jejunopexy site (including several centimeters of proxi-
mal and distal jejunum and full-thickness body wall) was
removed en bloc. The bowel was ligated proximal and
distal to each jejunopexy site with umbilical tape. A fluid
administration extension set was placed in the lumen at
the proximal end to permit infusion of methylene blue-
tinted saline. No attempt was made to prevent leakage of
infusate along the tract left by the removed jejunostomy
tube.
The prepared specimen was immersed in a colorless
saline bath, and while infused with saline tinted with
methylene blue at a constant rate (8.32 ml/min) using a
commercial intravenous (IV) fluid pump,c one investiga-
tor observed for evidence of tinted infusate escape while
another monitored intraluminal pressure via an inline
water manometer. Leakage pressure was recorded the
instant that tinted saline escaped from the jejunopexy
site into the colorless surrounding saline. The site was
disassembled for positive identification of the jejunos-
tomy tube technique employed and the pressure assigned
to the appropriate group. Analysis of data was performed
using a commercial computer software package.d Data
was presented as a mean±standard deviation (SD) for
each technique. Mean intraluminal leakage pressure was
compared statistically using one-way analysis of vari-
ance (ANOVA). Differences between the means were
identified using posthoc application of the Tukey test.
Jejunostomy Techniques
Interlocking Box
The interlocking box technique jejunopexy consisted of
two complete suture boxes, each 1 to 1.25 cm to a side,
which extended from the jejunum to the body wall and
surrounded the jejunostomy tube. Each box required four
suture passes to complete. Close attention was paid to
avoiding entanglement of the first and second boxes.
The first box was started from cranioventral with a
suture pass through the transversus abdominus muscle in
a dorsal direction. The second pass was performed in an
March/April 1999, Vol. 35
Figure 1A
Figures 1A, 1B, 1C—Interlocking box jejunopexy. (A) First box
placement. (B) Second box placement. (C) Apposition of jejunum
to body wall.
Figure 1B
aborad direction on the dorsal (when apposed to the body
wall) antimesenteric aspect of the jejunum through all
tissue layers except mucosa. The third pass was directed
ventrally in the transversus abdominus muscle parallel
and 1 to 1.25 cm ventral to the first pass. The fourth pass
completed the first box and was directed orad in the
jejunal wall parallel and 1 to 1.25 cm ventral to the
second pass [Figure 1A]. These suture ends were clamped
with hemostatic forceps, leaving sufficient suture mate-
rial to allow a space between the body wall and jejunum
for placement of the second box.
Interlocking Box Jejunostomy 131
Figure 1C
The second box was started with a pass in the trans-
versus abdominus muscle from the caudoventral corner
of the first box and directed to the cranioventral corner.
This was followed by a pass through the jejunal wall
dorsally from the cranioventral to the craniodorsal cor-
ner of the first box. A pass in the transversus abdominus
muscle from the craniodorsal to the caudodorsal corner
was followed by a pass through the jejunal wall from the
caudodorsal to the caudoventral corner, which completed
the second box [Figure 1B]. The ends of the second box
suture also were clamped.
A stab incision was made in the jejunal wall at the
center of the box, and the feeding tube was introduced
and advanced aborally within the lumen. A simple purse-
string suture was placed in the jejunal wall, excluding
the mucosal layer, around the tube using a simple hori-
zontal mattress pattern. The excess suture material in
each box then was removed carefully by pulling both
hemostats simultaneously while retracting any excess
feeding tube back through the body wall. This resulted in
close apposition of the jejunum and body wall, with the
interlocking box jejunopexy completely surrounding the
jejunostomy site [Figure 1C]. Each of the suture boxes
was tied separately, completing the jejunopexy.
Simple Purse-String
A stab incision was made on the antimesenteric border of
the jejunum, and the feeding tube was advanced aboral-
ly. A simple horizontal mattress pattern purse-string was
performed around the jejunostomy tube. Two interrupted
jejunopexy sutures were placed, one on each side of the
jejunostomy tube, from the jejunum to the body wall
[Figure 2].
132 JOURNAL of the American Animal Hospital Association March/April 1999, Vol. 35

Figure 3—Inverting serosal tunnel jejunostomy (Witzel tech-

nique) with two simple interrupted jejunopexy sutures. Inset: Cut-
Figure 2—Simple purse-string jejunostomy with two simple section appearance midtunnel.
interrupted jejunopexy sutures.

Inverting Serosal Tunnel

A stab incision was made on the antimesenteric border of
the jejunum, and the feeding tube was advanced aboral-
ly. A horizontal mattress suture was placed immediately
aborad to the jejunostomy tube entry site to initiate in-
version of the jejunal serosa. This suture line was contin-
ued in an orad direction as a continuous Lembert pattern,
inverting the intestinal wall around the jejunostomy tube
and forming a serosal tunnel. The length of the tunnel
was 2.5 cm in all cases. Two interrupted jejunopexy
sutures were placed from the jejunum to the body wall,
one on each side of the jejunostomy tube as it emerged
from the serosal tunnel [Figure 3].

Results
Leakage occurred in all preparations at the jejunopexy
site and not at either the skin surface or transected ends
of the prepared jejunal segments. Mean intraluminal leak-
age pressure±SD of the interlocking box technique
jejunopexy (87.63±40.56 cm H 2O) was significantly
greater (p less than 0.001) than the simple purse-string
plus jejunopexy (43.17±31.69 cm H2O) and inverting
serosal tunnel plus jejunopexy (46.33±23.60 cm H2O).
Significant difference between the latter techniques was
not detected [Figure 4].
Subjectively, the interlocking box technique was tech-
nically more demanding to perform than either the simple
purse-string or the inverting serosal tunnel.
Discussion
This evaluation of jejunostomy tube placement tech-
niques compared the intraluminal pressures necessary to
produce leakage at the jejunopexy site following acute
tube removal. The interlocking box technique resisted
Figure 4—Mean±standard deviation (SD) intraluminal pressure
(cm H2O) at leakage following acute jejunostomy tube removal in
isolated canine cadaver intestinal segments.
greater intraluminal pressure before leaking than did
conventional techniques. Whether pressure resistance
alone determines the effectiveness of a particular
technique is speculative. However, it seems logical that
pressure resistance is an important determinant of jeju-
nostomy tube leakage resistance.
Pressure resistance of the interlocking box jejunopexy
was clearly superior to conventional techniques in ca-
daver intestinal segments. The optimal pressure resis-
tance in living intestine undergoing regular peristaltic
contractions is unknown. Physiological studies in nor-
mal dogs have shown that peak intraluminal pressures
reach upwards of 85 cm H2O during segmental contrac-
tions and migrating motor complexes.15 Documentation
of pressures generated during disease conditions is not
reported in the veterinary literature. Common sense sug-
March/April 1999, Vol. 35
gests that a technique which resists at least the normal
maximal intraluminal pressure (85 cm H 2O) of the je-
junum would be favorable. With a mean of 87.63 cm
H2O, the interlocking box technique exceeds these physi-
ological pressures, at least ex vivo, and should be favored
over the conventional techniques tested.
All procedures were performed in postmortem tis-
sues, and therefore the effect of peristalsis on leakage
resistance could not be determined. In the presence of
normal contraction, pressure resistance may differ from
the authors’ findings. It also may be possible that post-
mortem tissue resists intraluminal pressure differently
than antemortem tissue, independent of contraction. Pro-
cedures were performed in the immediate postmortem
period to minimize postmortem tissue changes, but di-
rect comparisons to live tissue could not be made.
Reported indications for enteral feeding include inter-
ruption of oral feeding for three to five days; acute
weight loss greater than 5%; chronic weight loss greater
than 10%; generalized muscle wasting; and disease con-
ditions known to increase protein loss, induce hyperme-
tabolism, suppress appetite, or interfere with nutrient
ingestion, transit, or absorption.5,6,10 The authors con-
sider supplemental feeding important for any veterinary
patient that is stressed or receiving less than 85% of its
total caloric requirement in any 24-hour period.
In disease conditions where enterostomy feeding is
indicated, the only nutritionally complete alternative is
total parenteral nutrition (TPN). Total parenteral nutri-
tion is relatively expensive and requires a dedicated cen-
tral venous catheter and strict aseptic handling of
nutrients, catheter, and administration lines.2,6 Compli-
cations associated with TPN administration include cath-
eter kinking and displacement, phlebitis and thrombosis,
sepsis, hyperglycemia, hypoglycemia, hyperlipidemia,
azotemia, and electrolyte imbalances.1,2,6 Duration of
TPN administration is limited by the ability to maintain
an aseptic and functional central venous catheter.1,2 Fur-
thermore, TPN incompletely nourishes enterocytes and
resident intestinal microflora which have been associ-
ated with transmucosal bacterial and endotoxin translo-
cation as well as malassimilation following return to
enteral forms of nutrition.4,16–19 Additionally, jejunos-
tomy feeding is more physiological than TPN and can be
continued indefinitely.7,12 Total parenteral nutrition has
an important role in clinical nutrition, but whenever
possible, enteral routes should be utilized. 2–4
As a potentially less invasive alternative to operative
jejunostomies, a percutaneous gastroduodenostomy cath-
eter has been described in the veterinary literature.20
This placement technique was not always successful and
resulted in frequent retrograde migration of the tube.
Surgically placed gastro-jejunal tubes have also been
described experimentally in dogs.21 Retrograde tube mi-
gration occurred in one of four animals. Based on these
Interlocking Box Jejunostomy 133
complications, laparotomy-placed jejunostomy tubes re-
main the most viable approach to enterostomy feeding.
Subjectively, the interlocking box technique was tech-
nically more demanding to perform than either the simple
purse-string or the inverting serosal tunnel. Entangling
the first and second boxes resulted in less than ideal
apposition of jejunum and body wall in one of the inter-
locking box preparations. This resulted in an unusually
low pressure resistance of 22 cm H2O. Had this techni-
cally incorrect preparation been discarded, the mean in-
terlocking box leakage pressure±SD would have been
93.59±11.04 cm H2O, underscoring the importance of
proper application.
This study did not address whether the interlocking
box jejunopexy technique effectively prevents leakage
in vivo. To answer this question, experimental live ani-
mal trials or prospective trials with clinical patients
should be employed to evaluate the acute and chronic
complications associated with early removal of jejunos-
tomy tubes placed via the interlocking box technique.
Additionally, the effect of box size on leakage resis-
tance was not addressed. The authors produced uniform
boxes that were 1 to 1.25 cm in size and can recommend
no other size at this time. However, it seems possible that
larger or smaller boxes also could be employed success-
fully. One concern would be that a larger box size would
result in a larger potential cavity between the apposed
jejunum and body wall, which might predispose to accu-
mulation of ingesta.
Finally, the applicability of the interlocking box tech-
nique to safe placement of other tubes such as cys-
tostomy or surgically placed gastrostomy tubes was not
addressed. The authors have placed both operative gas-
trostomy tubes and cystostomy tubes using this tech-
nique. No serious complications have been noted with
this approach, but planned early removal has not been
attempted. Anecdotally, the interlocking box cystopexy
has allowed replacement of a prematurely removed cys-
tostomy tube directly through the external stoma (with-
out anesthesia or sedation). The tube was removed by the
patient and replaced within several hours of removal by
the attending clinician. Proper replacement was con-
firmed by contrast radiography. No attempts have been
made to replace prematurely removed jejunostomy tubes,
but, at least in theory, this would be possible and a major
benefit to the interlocking box jejunopexy technique.
Conclusion
The interlocking box technique for jejunostomy tube
placement was compared to two conventional techniques.
The technique allowed the generation of significantly
greater intraluminal pressure before leakage was noted
at the jejunopexy site. The results of this ex vivo study
validate the technique and suggest further evaluation is
warranted.
134 JOURNAL of the American Animal Hospital Association March/April 1999, Vol. 35

Addendum
During the preparation of this manuscript, a new jeju-
nostomy tube placement technique and its rate of com-
plications in 47 small animal patients was reported.9 A
continuous circumferential suture pattern was utilized to
appose the jejunostomy site to the body wall. Twelve
patients removed or had their tubes removed before 10
days, a time period those investigators considered early
removal. Of the four animals with tubes removed before
five days, two died or were euthanized due to tube-
related complications and one had serious complications
but survived with additional surgical treatment. The
fourth animal removed its own tube on the first day
jejunostomy tubes in dogs and cats: 40 cases (1989–1994). J Am Vet Med
Assoc 1997;210:1764–7.
9. Crowe DT, Devey JJ. Clinical experience with jejunostomy feeding tubes
in 47 small animal patients. J Vet Emerg Crit Care 1997;7:7–19.
10. Crow Jr DT. Enteral nutrition for critically ill or injured patients—Part I.
Comp Cont Ed Pract Vet 1986;8:603–12.
11. Goode AW. The scientific basis of nutritional assessment. Br J Anaesth
1981;53:161–82.
12. Crowe Jr DT. Enteral nutrition for critically ill or injured patients—Part II.
Comp Cont Ed Pract Vet 1986;8:719–32.
13. Moss G, Greenstein A, Levy S, et al. Maintenance of GI function after
bowel surgery and immediate enteral full nutrition. I. Doubling of canine
colorectal anastomotic bursting pressure and intestinal wound mature
collagen content. J Parenter Enter Nutr 1980;4:535–8.
14. Andrassy RH, Mahour GH, Harrison MR, et al. The role and safety of early
postoperative feeding in the pediatric surgery patient. J Pediatr Surg
1979;14:381–5.

following surgery and had only mild complications, 15.
which resolved within 36 hours. None of the patients had 16.
intentional early jejunostomy tube removal (i.e., before 17.
five days). The most common mild complications re-
ported were discharge from the ostomy site or cellulitis.
18.
Surgical breakdown and serious complications occurred
at the jejunostomy tube site in three animals.
Otterson MF, Sarr MG. Normal physiology of small intestinal motility.
Surg Clin N Am 1993;7:1173–89.
Deitch EA. Bacterial translocation of the gut flora. J Trauma 1990;30:s184.
Alverdy J, Chi HS, Sheldon GF. The effect of parenteral nutrition on
gastrointestinal immunity. The importance of enteral stimulation. Ann Surg
1985;205:681–4.
Nirgotis JG, Andrassy RJ. Preserving the gut and enhancing the immune
response: the role of enteral nutrition in decreasing sepsis. Contemp Surg
1992;41:17–26.

One of the authors of the present study elected to 19. Wilmore WD, Smith RJ, O’Dwyer ST, et al. The gut: a central organ after
surgical stress. Surg 1988;104:917–23.
informally compare intraluminal leakage pressure resis- 20. McCrackin MA, DeNovo RC, Bright RM, Toal RL. Endoscopic placement
tance of this new technique to the interlocking box. In of a percutaneous gastroduodenostomy feeding tube in dogs. J Am Vet
Med Assoc 1993;203:792–7.
two groups (n=4), intraluminal leakage pressure of the
21. Hardie EM, Armstrong PJ. Development of a gastro-jejunal tube for the
continuous circumferential technique was very similar to dog. J Nutr 1991;121:S154.
the interlocking box using methods consistent with the
present study.e These newly described techniques resist
leakage similarly, and both seem superior to conven-
tional techniques. Evaluation is needed regarding whether
the interlocking box can improve upon the complication
rates of the continuous circumferential jejunopexy tech-
nique. Further studies are planned.

a
Sovereign, feeding tube and urethral catheter; Sherwood Medical, St. Louis,
MO
b
3-0 PDS* II, polydioxanone; Ethicon Inc., Somerville, NJ
c
Flow-gard 6100, volumetric infusion pump; Travenol Laboratories, Inc.,
Deerfield, IL
d
SigmaStat; Jandel Scientific, Corte Madeira, CA
e
Daye RM. Unpublished data

References
1. Crowe DT. Nutritional support for the hospitalized patient: an introduction
to tube feeding. Comp Cont Ed Pract Vet 1990;12:1711–20.
2. Ray PA, Swecker Jr WS. Nutritional management of dogs and cats with
cancer. Vet Med December 1992;1185–94.
3. Lewis LD, Morris Jr ML, Hand MS. Anorexia, inanition, and critical care
nutrition. In: Lewis LD, Morris Jr ML, Hand MS, eds. Small animal
clinical nutrition III. Topeka: Mark Morris Associates, 1987:5.1–5.43.
4. Donoghue S. Nutritional support of hospitalized patients. Vet Clin N Am
Sm Anim Pract 1989;19:475–95.
5. Lipert AC. Enteral and parenteral nutritional support in dogs and cats with
gastrointestinal disease. Sem Vet Med Surg 1989;4:232–40.
6. Armstrong PJ, Lippert AC. Selected aspects of enteral and parenteral
nutritional support. Sem Vet Med Surg 1988;3:216–26.
7. Armstrong PJ, Hand MS, Frederick GS. Enteral nutrition by tube. Vet Clin
N Am Sm Anim Pract 1990;20:237–75.
8. Swann HM, Sweet DC, Michel K. Complications associated with use of
 Cervical Vertebral Fractures in 56 Dogs:
A Retrospective Study
The clinicopathological features of cervical fractures in 56 dogs were reviewed. “Hit by car”
(HBC) was the most common inciting cause, and the axis and atlas were the vertebrae most
frequently affected. Surgical treatment was associated with high (36%) perioperative mortality.
However, all dogs that survived the perioperative period achieved functional recovery.
Functional recovery was achieved in 25 (89%) of 28 nonsurgically treated dogs with adequate
follow-up. Overall, severity of neurological deficits (nonambulatory status) and prolonged
interval (five days or longer) from trauma to referral were associated with poorer outcome.
Nonsurgical treatment is a viable therapeutic approach for many dogs with cervical fractures.
Early neck immobilization and prompt referral are recommended, because delay in referral
decreases the likelihood of functional recovery. J Am Anim Hosp Assoc 1999;35:135–46.

Julia C. Hawthorne, DVM Introduction

William E. Blevins, DVM, MS,
Diplomate ACVR
Larry J. Wallace, DVM,
Diplomate ACVS
Nita Glickman, MPH, MS
David J. Waters, DVM, PhD,
Diplomate ACVS
RS
From the Department of Veterinary Clinical
Sciences (Hawthorne, Blevins, Waters) and
The Center for the Human-Animal Bond
(Glickman),
School of Veterinary Medicine,
Purdue University,
1248 Lynn Hall,
West Lafayette, Indiana 47907
and the Department of Small Animal Clinical
Sciences (Wallace),
College of Veterinary Medicine,
University of Minnesota,
C339 1352 Boyd Avenue,
St. Paul, Minnesota 55108.
Address all correspondence to Dr. Waters.
In dogs, cervical vertebral fractures occur less frequently than fractures of
the thoracolumbar spine. Thoracolumbar fractures are well characterized
with respect to epidemiology, anatomic distribution, and prognosis after
surgical or nonsurgical treatment.1–3 In contrast, cervical vertebral frac-
tures of dogs are poorly characterized. The largest reported case series,4
published nearly 20 years ago, provided information on 12 dogs and
reviewed 15 previously reported cases.5–10 Since then, seven reports have
contributed data on an additional 21 cases.11–17 The paucity of published
information on cervical fractures in dogs, in particular the factors that
influence functional outcome after nonsurgical treatment, prompted the
authors to conduct this retrospective study.
The specific aims of this study were: 1) to characterize the clinico-
pathological features of 56 dogs with cervical vertebral fractures, and 2)
to identify factors that were predictive of successful outcome in 46 dogs
with adequate follow-up, including the subset of 28 dogs that underwent
nonsurgical treatment. The results indicate that functional recovery often
can be achieved with nonsurgical treatment. Surgical management of
cervical vertebral fractures is associated with high perioperative mortal-
ity. Nonambulatory status and delayed referral are associated with de-
creased likelihood for a functional recovery.
Materials and Methods
Case Selection
Dogs with cervical vertebral fractures admitted to the Veterinary Teach-
ing Hospitals of Purdue University and the University of Minnesota (1969
through 1992) were studied. Medical records were reviewed, and cases
were selected for study if cervical radiographs, neurological examination
findings, and treatment information were available. Fifty-six dogs satis-
fied these inclusion criteria. In all cases, diagnosis of cervical vertebral
fracture was based upon radiographs, necropsy findings, or both. The
following data was obtained from the medical record: signalment, inciting
cause of fracture, onset and severity of neurological signs, concurrent
injuries, and radiographic interpretation. The time interval from trauma to

JOURNAL of the American Animal Hospital Association 135

136 JOURNAL of the American Animal Hospital Association
Figure 1—Lateral radiograph of a six-month-old chow chow with
a second cervical (C2) vertebral fracture (arrows) in the area of
the synostosis of the dens and the body of the axis. There is
dorsal displacement of the main portion of C2 relative to the
fractured fragment. The axis is the most common vertebra
affected in dogs with cervical fracture.
presentation to the referral institution was recorded. Spe-
cific location of each fracture, treatment (surgical versus
nonsurgical), and follow-up information also were re-
corded. Follow-up data was obtained from medical
records or telephone questionnaires from the pet owner
or referring veterinarian.
Analysis of Possible Prognostic Factors
Data from 46 dogs with adequate follow-up was ana-
lyzed to determine if particular factors were associated
with functional recovery. Ten dogs that were lost to
follow-up after nonsurgical treatment were excluded
from this analysis. A separate analysis was performed to
identify factors predictive of outcome in the subset of 28
dogs that underwent nonsurgical treatment. Functional
recovery was defined as pain-free ambulation with uri-
nary and fecal continence. The following potential prog-
nostic factors were evaluated: inciting cause (HBC versus
other); level of affected vertebra (first cervical [C1] and
second cervical [C2] versus other); multiplicity (single
versus multiple vertebrae); time interval from trauma to
presentation to referral institution (less than five days
versus five or more days); severity of neurological defi-
cits on presentation (nonambulatory versus ambulatory);
concurrent head trauma; and surgical versus nonsurgical
treatment. Two by two tables were constructed, and chi-
square (χ2) or Fisher’s exact tests were used to determine
the association between these factors and functional re-
covery. Whenever possible, odds ratio (OR) and 95%
confidence intervals were calculated.18 Odds ratio pro-
vides an estimate of the relative risk that a patient with a
March/April 1999, Vol. 35
particular characteristic will have a certain outcome (e.g.,
functional recovery in this study). Odds ratios were con-
sidered statistically significant if the 95% confidence
interval did not include 1.0.
Results
Signalment and Inciting Cause
The clinicopathological features of 56 dogs with cervical
fractures are summarized in Table 1. Median age was
two years (range, 4 mos to 14 yrs), and there was an
equal sex distribution. Median body weight was 18 kg
(range, 1 to 50 kg). Twenty-six (46%) of 56 dogs with
cervical fractures were HBC. Other inciting causes in-
cluded a big dog/little dog fight (n=8); collision/rough
play (n=6); a door slam (n=4); a gunshot injury (n=2); a
fall into a hole or down stairs (n=3); blunt trauma (n=1);
a leash injury (n=1); and a tumor-associated pathological
fracture (n=1). In four cases, there was no history of
trauma.
Concurrent Injuries
Concurrent injuries were noted in 27 (48%) of 56 dogs.
Hit by car was the inciting cause of vertebral fracture in
17 (63%) of 27 dogs that had concurrent injuries. These
included fractures of long bones (n=3), mandible (n=2),
scapula (n=2), rib (n=1), occiput (n=2), and thoracolum-
bar vertebrae (n=2). Soft-tissue wounds, usually associ-
ated with gunshot injuries or animal bites, were found in
eight dogs. Three dogs had thoracic trauma (e.g., pneu-
mothorax, lung contusions). Clinical signs of head trauma
(e.g., hemorrhage from the nose or mouth, seizures, loss
of consciousness) were reported in nine dogs.
Severity and Progression of Neurological Deficits
Severity of neurological deficits upon presentation to the
referral institution was variable [Table 1]. Thirty-two
(57%) of 56 dogs were nonambulatory. However, loss of
voluntary motor function (i.e., tetraplegia) was noted in
only four dogs, and none of these dogs had complete
sensorimotor loss. Eight (14%) of 56 dogs had cervical
pain as the only abnormality on neurological examination.
In many cases, diagnosis of cervical vertebral fracture
was not made immediately, and a decline in neurological
status noted days to weeks after the traumatic event
prompted reevaluation. Median interval from trauma to
presentation at the referral institution was five days
(range, 2 hrs to 2 mos). Deterioration in neurological
status was documented in 13 (48%) of 27 dogs for which
detailed histories were available.
Anatomic Distribution
The axis (C2) was the most commonly affected vertebra
[Figure 1]. Twenty-nine (52%) of 56 dogs had C2 frac-
tures, which accounted for 29 (36%) of 81 fractures.
Twenty-five percent of dogs had C1 fractures. Nineteen
(34%) of 56 dogs had multiple cervical vertebrae af-
 Table 1
Summary of Clinicopathological Features of 56 Dogs With Cervical Fractures

Trauma to Neurological
Case Age Weight Referral Findings and
No. Breed (yrs) (kg) Site* Cause† Interval Diagnostics‡ Treatment¶ Other Injuries# Outcome**
March/April 1999, Vol. 35

1 Shepherd 0.5 14 C3 BDLD <1 day NAT Tracheostomy Dyspnea, tracheal Arrested after presenta-
tear tion; necropsy (hemotho-
rax, lung contusions)
2 Beagle 5 22 C2, C6 HBC 5 days NAT; cervical None Bilateral mandibular Euthanized; no necropsy
pain fx
3 Newfound- 7 50 C2, C3 HBC 12 days CP deficits (LF, None Rib, femoral fx; Euthanized; no necropsy
land RF); nonweight- partial brachial
bearing plexus avulsion
lameness (RH)
4 Labrador 8 34 C7 Unknown 2 mos NAT; progressive None None Euthanized; necropsy
retriever ataxia; myelogram, (osteosarcoma, no
C7 pathologic fx evidence of metastasis)
with compression
5 Doberman 9 41 C4, C6 No known 2 days NAT; weak VM None None Euthanized; necropsy
pinscher trauma (all limbs); cervical (focal axonal compres-
pain; myelogram, sion C4-C7)
C5-C6 compression
6 Cocker 3 8 C2 Fell (bath) <1 day NAT; torticollis; None None Euthanized; necropsy
spaniel intact deep pain; (fx dens, meningitis,
myelogram malacia C1-C2)
7 German 0.5 40 C1, C3, HBC 1 day NAT; cervical pain None Epistaxis Arrested; necropsy
shepherd C4, C5 (epidural hemorrhage)
dog
8 Chow chow 0.5 9 C2 Door slam 6 days NAT; cervical pain Dorsal suture None 7 days: ambulatory;
stabilization, 5 yrs: normal
postop splint;
4 days postop,
suture broke;
wire stabilization
9 Miniature 1.5 NA§ C2 HBC 8 days NAT; extreme Dorsal wire Skull fracture 4 days: died; necropsy
pinscher cervical pain; spastic stabilization; 3 (cerebral necrosis,
cervical musculature days postop, cervical cord
seizures demyelination)
Cervical Vertebral Fractures

10 Mixed- 1 3.5 C5, C7 BDLD 6 days NAT; absent VM Dorsal laminec- Puncture wounds; Perioperative arrest;
breed dog (forelimbs) tomy/facet wiring T13-L1 disk herniation necropsy (demyelination
(C6- C7); hemi- of cervical spine)
137

laminectomy (T13-L1)
(Continued on next page)
 138
Table 1 (Cont’d)

Trauma to Neurological
Case Age Weight Referral Findings and
No. Breed (yrs) (kg) Site* Cause† Interval Diagnostics‡ Treatment¶ Other Injuries# Outcome**
11 American 1 9 C1, C2 HBC 5 days Severe cervical pain; Dorsal wiring None 2 wks: ambulatory
eskimo minimal ambulation; and neck brace
hyporeflexia
(forelimbs)
12 Labrador 1 27 C2 Fell (10- 5 days NAT; intact deep pain Dorsal suture None 6 wks: ambulating, ataxic;
retriever foot deep (all limbs); absent stabilization; 14 mos: normal
hole) VM in forelimbs whirlpool
13 Mixed- 6 14 C2 HBC 8 days NAT; intact VM and Dorsal None Perioperative arrest;
breed dog CP deficits (all stabilization necropsy (contusion/
limbs); with wire of C1-C2; laceration of dura/spinal
opisthotonos decompression cord beneath wires
of caudal aspect around atlas, focal
of skull malacia of brain stem)
14 Shetland 0.5 11 C1 HBC 9 days NAT; hyporeflexia Dorsal Scapular and 24 hrs postop: arrested;
JOURNAL of the American Animal Hospital Association

sheepdog (LH, LF); CP deficits laminectomy occipital condylar fx necropsy (myelomalacia

(all limbs); cervical of C1 at base of cerebellum)
pain; cranial nerve
deficits
15 Chihuahua 6 2.5 C5, C6 BDLD <1 day NAT; hyperreflexia Dorsal Horner’s syndrome 10 mos: CP deficit (RF)
(all limbs) laminectomy
C4-C7
16 Samoyed 1.5 15 C2 HBC <1 day Cervical pain; Screws/methyl- None 5 days: mild ataxia;
decreased VM methacrylate; 22 mos: normal
(forelimbs); CP neck cast
deficits (all limbs)
17 Pit bull 1.5 20 C2 Collision 1 day NAT; cervical pain; Reduction/ None 8 days: mild ataxia
(truck bed) absent VM (all limbs); stabilization;
superficial pain methylmeth-
present acrylate implant
around 3 bone
screws
18 Miniature 1 10 C2 HBC 1 day NAT; intact VM Partial left None 3 wks: normal
schnauzer hemilaminectomy;
lubra plate; neck
brace
March/April 1999, Vol. 35
 Table 1 (Cont’d)

Trauma to Neurological
Case Age Weight Referral Findings and
No. Breed (yrs) (kg) Site* Cause† Interval Diagnostics‡ Treatment¶ Other Injuries# Outcome**
19 Golden 7 36 C2 Collision 5 days NAT; no deep pain Neck cast None 6 days: arrested while
March/April 1999, Vol. 35

retriever (side of following casting rolling over; no necropsy

car)
20 Golden 0.5 27 C2 HBC 5 days NAT; absent VM Cage rest None 3 mos: euthanized due to
retriever (LH, RH) residual paresis (RH, LH)
21 St. Bernard 2 48 C3, C4 HBC 5 days NAT; deep pain Cage rest T2 fracture 2 wks: euthanized due to
absent (LF); EMG - no improvement;
left median nerve necropsy (epidural
lesion hemorrhage C2-T3)
22 Great Dane 2 55 C2, C3 HBC 2 days CP deficits (all limbs) Neck cast Hemorrhage from 5 wks: slight CP deficits
nose/ears (LF, LH); 4 mos: normal
23 Doberman 1.5 32 C2 Playing 1 day Severe cervical pain; Neck brace None 2 mos: no cervical pain
pinscher hyperreflexia (all limbs)
24 Mixed-breed 0.5 18 C2 Chain 6 days NAT; CP deficits and Neck brace None 4 wks: normal
dog hyperreflexia (all
limbs); cervical pain
25 Mixed-breed 0.5 14 C2 HBC 10 days Cervical pain; CP Plaster neck Mandibular fx 5 wks: normal
dog deficits (LH, RH) brace
26 Mixed-breed 2 21 C3 Gunshot 2 days Ataxia; CP deficits Cardboard neck Bullet wound 2 wks: CP deficits (LH,
dog (RF, LH, RH) brace RH)
27 Labrador 5 32 C1 HBC 10 days Cervical pain Neck brace Temporal muscle 5 mos: normal
retriever atrophy
28 Yorkshire 1 1.5 C1 Door slam 2 days NAT; CP deficits Neck brace None 3 wks: normal
terrier (all limbs);
hyperreflexia (LH)
29 Mixed-breed 2 6 C2 Playing- 1 month Muscle tremors (RF/ Neck brace None 4 wks: mild ataxia
dog dogs neck); myelogram
(normal)
30 Australian 5 16 C1 HBC 1 day Seizures; blindness; Neck bandage Pneumothorax, 6 wks: normal except
shepherd ataxia scleral hemorrhage blind in right eye
31 Mixed- NA 3.5 C1 Unknown 1 day NAT; CP deficits Neck splint None 2 wks: weak in LR
breed dog (LH, LF)
32 Miniature 2 9 C5 HBC 2 days NAT; absent VM Neck brace Herniated disk 34 mos: slight LF
Cervical Vertebral Fractures

poodle (RF); cervical pain; C5- C6 weakness

myelogram (disk
herniation C5- C6) (Continued on next page)
139
 140
Table 1 (Cont’d)

Trauma to Neurological
Case Age Weight Referral Findings and
No. Breed (yrs) (kg) Site* Cause† Interval Diagnostics‡ Treatment¶ Other Injuries# Outcome**
33 Flat-coated 1.5 23 C2 HBC 1 day NAT; CP deficits Neck cast; bone Radius/ulna fx 3 days: standing; 3 yrs:
retriever (RH, RF); torticollis plate (left radius) normal; low head carriage
34 Toy poodle 0.5 2 C6 Fell down 1 day Ataxia; CP deficits Neck brace None 2 wks: ataxic; 4 yrs: mild
stairs (RF, LF) gait abnormality
35 Chihuahua 4 1.5 C3 Hit by door <1 day Cervical pain; ataxia Neck splint None 10 days: mild ataxia
36 Dalmatian 1 19 C2, C3 Unknown 14 days Cervical pain Neck cast None 4 wks: normal
(episodic, 3x/day)
37 Cocker 13 NA C1 BDLD 1 day NAT; minimal VM Neck bandage None 3 mos: ambulatory, weak
spaniel/ (LF, LH) (LH); 2 yrs: normal
poodle
38 Doberman 3 28 C1, C2 Blunt 1 day Cervical pain Neck cast None 7 wks: normal
pinscher object to
head
JOURNAL of the American Animal Hospital Association

39 German 0.5 27 C2 HBC 7 days NAT Neck cast None 3 days: assisted
shepherd ambulation; 6 wks:
dog minimal ataxia
40 Toy poodle 3 3 C2, C5, BDLD 3 days NAT Neck splint None 1 yr: LF weakness
C6
41 Poodle mix 5 18 C4, C5 HBC <1 day NAT; cervical pain Neck cast None 5 wks: CP deficits (LF)
42 Whippet 5 17 C6 Collision <1 day Ambulatory tetra- Neck cast None 6 wks: normal except
(racing) paresis; cervical pain hyperreflexia (LH, RH)
43 Brussel 3 3.5 C2 BDLD <1 day NAT Neck splint Puncture wounds 2 wks: assisted
Griffin ambulation; 1 yr: normal
44 German 0.5 23 C1 HBC <1 day Vestibular signs; Cage rest Epistaxis 5 days: normal
shepherd nystagmus; ataxia;
dog shock
45 Mixed- 0.5 10 C6, C7 HBC <1 day Cervical pain; RF Cage rest Scapular fx 10 mos: normal
breed dog lameness
46 Keeshond 3 18 C4, C5 HBC 5 days NAT; absent VM/ Cage rest; Femoral/T1 fx, 18 mos: normal
decreased sciatic external fixation epistaxis
reflex (LH) of femoral fx
47 Mixed- 1 18 C5, C6 HBC 4 days NAT; CP deficits Neck brace Pneumothorax, No follow-up
breed dog and hyperreflexia contusions,
(all limbs); myelogram concussion
(no compression)
March/April 1999, Vol. 35
 Table 1 (Cont’d)

Trauma to Neurological
Case Age Weight Referral Findings and
No. Breed (yrs) (kg) Site* Cause† Interval Diagnostics‡ Treatment¶ Other Injuries# Outcome**
48 Doberman 2 30 Collision 5 days Cervical pain Cardboard neck None No follow-up
March/April 1999, Vol. 35

C2
pinscher brace
49 Chihuahua 14 9 C2, C3, BDLD 2 days Ataxia; cervical pain Neck cast; wound Bite wounds No follow-up
mix C4, C5 management
50 German 1 18 C2 HBC 18 days Cervical pain Activity restriction None No follow-up
shepherd
dog
51 German 1 NA C2, C3 Unknown 6 wks Left hemiparesis Activity restriction None No follow-up
shepherd (especially LH)
dog
52 Basset 2 30 C4 HBC 9 days NAT; minimal VM Cage rest None No follow-up
hound (RF, LF)
53 Labrador 3 27 C1 Gunshot <1 day Nonambulatory left Cage rest Puncture wound No follow-up
retriever hemiplegia; extensor (neck)
rigidity (LH)
54 Beagle 0.5 NA C1 HBC <1 day Comatose/concus- Shock therapy; Shock No follow-up
sion; then cage rest
ambulatory
tetraparesis
55 Chihuahua 8 3 C1 BDLD 5 days NAT; myelogram None Puncture injury No follow-up
(normal)
56 Doberman 4 38 C2, C3, Hit by gate <1 day No neurological None Neck laceration No follow-up
pinscher C4 deficits

* C1=first cervical vertebra; C2=second cervical vertebra; C3=third cervical vertebra; C4=fourth cervical vertebra; C5=fifth cervical vertebra; C6=sixth cervical
vertebra; C7=seventh cervical vertebra
†
BDLD=big dog/little dog fight; HBC=hit by car
‡
NAT=nonambulatory tetraparesis; CP=conscious proprioception; LF=left forelimb; RF=right forelimb; RH=right hind limb; fx=fracture; VM=voluntary motor; LH=left
hind limb; EMG=electromyogram
§
NA=not available
¶
T13=thirteenth thoracic vertebra; L1=first lumbar vertebra
#
T2=second thoracic vertebra; T1=first thoracic vertebra
Cervical Vertebral Fractures

** T3=third thoracic vertebra

141
142 JOURNAL of the American Animal Hospital Association
*
* Data from these 46 dogs was used in the overall analysis of prognostic factors
Figure 2—Subclassification of 56 dogs with cervical fractures based upon treatment, follow-up data, and outcome.
fected. When the third cervical, fourth cervical, or fifth
cervical vertebra was fractured, a single vertebra was
affected in only 15% of the cases. A big dog/little dog
fight and unknown trauma were the only subgroups in
which fractures of the third through the seventh cervical
vertebrae outnumbered fractures of C1 and C2.
Treatment and Outcome
Fifty-six dogs with cervical fractures were subdivided
into three groups based upon treatment/outcome: 1) dogs
that died or were euthanized within 24 hours of referral;
2) dogs that received surgical treatment; and 3) dogs that
received nonsurgical treatment [Figure 2].
Seven dogs died or were euthanized within 24 hours
of referral. Two of these seven dogs suffered cardiopul-
monary arrest, and two dogs were euthanized within
hours after presentation due to the severity of systemic
injuries. Three dogs were euthanized after myelography
due to the severity of injuries or anesthetic complications.
Eleven dogs received surgical treatment for their cer-
vical vertebral fractures. Nine of 11 dogs had C1 or C2
lesions. Preoperative neurological status in nine of 11
dogs was nonambulatory tetraparesis. Surgical manage-
ment consisted of dorsal suturing/wiring with or without
dorsal laminectomy (n=5), dorsal laminectomy alone
(n=2), screw fixation/stabilization using methylmeth-
acrylate via a ventral approach (n=2), or hemilaminec-
tomy with plastic dorsal spinous process plating (n=1).
Surgical treatment was associated with high periopera-
tive mortality; four (36%) of 11 surgically treated dogs
died. Cardiopulmonary arrest occurred in three dogs
within 24 hours after surgery, while the fourth dog had
March/April 1999, Vol. 35
*
cardiopulmonary arrest on the fourth postoperative day.
All dogs that survived the perioperative period achieved
functional recovery.
Thirty-eight dogs received nonsurgical treatment con-
sisting of a neck brace/splint with activity restriction
(n=26) or activity restriction alone (n=12). Follow-up
data was available for 28 dogs that underwent nonsurgi-
cal treatment. Twenty (71%) of these 28 dogs had C1 or
C2 lesions, and 14 were nonambulatory at the time of
referral. In contrast to surgically treated patients, only
three (11%) of 28 dogs died or were euthanized. One dog
was euthanized after respiratory arrest and resuscitation
two days after neck cast application. Two dogs that were
treated with activity restriction were euthanized because
of residual deficits at two weeks and three months, re-
spectively. Twenty-five (89%) of the 28 nonsurgically
treated dogs achieved functional recovery. Four of 25
dogs that achieved functional recovery had mild residual
neurological deficits (median follow-up, 3.5 wks).
Predictors of Functional Recovery
Follow-up (median, 1.5 mos; range, 0.5 to 77 mos) out-
come data was available from the medical records or
telephone questionnaires for 46 dogs [Table 1]. Func-
tional recovery (i.e., pain-free ambulation, urinary and
fecal continence) was achieved in 32 (70%) of 46 dogs.
Overall, severe neurological deficits (OR, 13.00; 95%
confidence interval, 1.52 to 111.47) and delayed interval
from trauma to referral (OR, 5.50; 95% confidence inter-
val, 1.38 to 21.85) were associated with a decreased
likelihood for functional recovery [Table 2]. Thus, dogs
with nonambulatory tetraparesis were 13 times less likely
March/April 1999, Vol. 35 Cervical Vertebral Fractures 143

Table 2
Predictors of Functional Recovery (FR) in 46 Dogs With Cervical Fractures*

No. With No. Without

Predictor FR FR Odds Ratio† 95% CI‡
Inciting cause
Hit by car 15 8 0.66 0.15–2.78
Other 17 6 1.00
Level of affected vertebrae
C1 - C2 21 6 2.19 0.52–9.23
Other 8 5 1.00
Single versus multiple vertebra
Single 23 8 1.92 0.52–7.10
Multiple 9 6 1.00
Trauma to referral interval
<5 days 22 4 5.50 1.38–21.85
>5 days 10 10 1.00
Severity of neurological deficits
Ambulatory 16 1 13.00 1.52–111.47
Nonambulatory 16 13 1.00
Clinical signs of head trauma
Absent 28 11 0.52 0.10–2.73
Present 4 3 1.00
Surgical versus nonsurgical treatment
Nonsurgical 25 10 0.70 0.17–2.93
Surgical 7 4 1.00

* Ten of 56 dogs with cervical fractures were lost to follow-up after nonsurgical treatment; these dogs were excluded
from this analysis.
†
Odds ratio estimates the likelihood that an outcome is associated with a particular factor. For example, ambulatory
dogs are 13 times more likely to achieve functional recovery than nonambulatory dogs. The aim of this study was to
identify negative prognostic factors. Based upon this analysis, nonambulatory status is considered a significant
predictor of poor outcome. The broad 95% confidence interval suggests that the “true” risk for poor outcome
associated with nonambulatory tetraparesis is somewhere between 1.5 times and 111 times the risk for poor outcome
associated with an ambulatory status.
‡
95% CI=95% confidence interval. Odds ratios were considered statistically significant if 95% CI did not include 1.0.
For example, dogs with single fractures are 1.92 times more likely to have functional recovery than dogs with multiple
fractures. However, because the 95% CI of 0.52–7.10 includes 1.0, the risk for poorer outcome in dogs with multiple
fractures is not considered significant.

to have functional recovery compared to ambulatory
dogs, whereas dogs with a trauma to referral interval of
five days or longer were 5.5 times less likely to recover
than dogs with a trauma to referral interval of less than
five days. Analysis of prognostic factors for 28 non-
surgically treated dogs showed that delayed presentation
to the referral institution (i.e., five days or longer) was
significantly associated with a decreased likelihood for
functional recovery when compared to dogs with a trauma
to referral interval of less than five days (p=0.04) [Table
3]. No other factors were significantly associated with
outcome.
Discussion
In contrast to thoracolumbar fractures, the clinicopatho-
logical features of cervical fractures in dogs have been
poorly characterized. Previously reported data on func-
tional outcome after surgical or nonsurgical treatment is
limited to 33 and 12 dogs, respectively.4–17 Importantly,
no previous reports have evaluated prognostic factors
that might predict functional outcome. The lack of infor-
mation regarding patient outcome, in particular for those
dogs that underwent nonsurgical treatment, prompted
this study. The authors’ results indicate that nonsurgical
treatment (i.e., neck immobilization and activity restric-
tion) can be used successfully in many dogs with cervi-
cal fractures. Furthermore, this study provides the first
information that certain factors may predict the likeli-
hood of functional recovery in dogs with cervical fractures.
Surgical treatment of cervical fractures was associ-
ated with high perioperative mortality. The 36% periop-
erative mortality rate is consistent with a previous report4
in which 37% of surgically treated dogs with cervical
fractures did not survive the perioperative period [Table
144 JOURNAL of the American Animal Hospital Association March/April 1999, Vol. 35

Table 3
Predictors of Functional Recovery (FR) in 28 Nonsurgically Treated Dogs With Cervical Fractures

p value
No. With No. Without (Fisher’s
Predictor FR FR Odds Ratio* 95% CI Exact Test)
Inciting cause
Hit by car 10 2 0.33 0.1–5.82 p=0.56
Other 15 1 1.00
Level of affected vertebrae
C1 - C2 18 2 0.93 0.05–31.19 p=1.00
Other 7 1 1.00
Single versus multiple vertebra
Single 18 2 1.29 0.10–16.54 p=1.00
Multiple 7 1 1.00
Trauma to referral interval
<5 days 18 0 NA† NA p=0.04
>5 days 7 3
Severity of neurological deficits
Ambulatory 14 0 NA NA p=0.09
Nonambulatory 11 3
Clinical signs of head trauma
Absent 21 3 NA NA p=1.00
Present 4 0

* Odds ratio estimates the likelihood that an outcome is associated with a particular factor. For example, dogs with
single vertebral fractures are 1.29 times more likely to achieve functional recovery than dogs with multiple vertebral
fractures. Odds ratios were considered statistically significant if 95% confidence interval (95% CI) did not include 1.0.
Because the 95% CI of 0.10–16.54 includes 1.0, the risk for poorer outcome in dogs with multiple fractures is not
considered significant.
†
NA=not applicable

Table 4
Likelihood of Functional Recovery (FR) in Surgically Treated Dogs
With Cervical Fractures

Dogs That Survived

All Dogs Perioperative Period
No. of Dogs % FR No. of Dogs % FR
4
Stone, et al. 1979 8 63 5 100
Reviewed in reference 4 12 92 12 92
Veterinary literature (1980–1997)11–17 13 100 13 100
This series 11 64 7 100

Total 44 82 37 97

4]. However, the prognosis for functional recovery in
perioperative survivors is excellent; 100% of dogs in the
authors’ series that survived the perioperative period had
functional recovery. Collectively, 36 (97%) of 37 dogs
in the authors’ series and the literature that survived the
perioperative period achieved functional recovery.4–17
An explanation for the high perioperative mortality in
dogs with cervical fractures is not apparent, but this
observation is consistent with previous reports of com-
plications associated with cervical spinal surgery.19–21
Nonsurgical treatment resulted in functional recovery
in 89% of dogs in the authors’ series [Table 5]. The three
March/April 1999, Vol. 35
Table 5
Likelihood of Functional Recovery (FR) in
Dogs With Cervical Fractures
That Underwent Nonsurgical Treatment
No. of Dogs % FR
4
Stone, et al. 1979 2 100
Reviewed in reference 4 2 100
Veterinary literature (1980–1997)11–17 8 100
This series 28 89
Total 40 93
dogs that did not achieve functional recovery died or
were euthanized because of neurological deficits six days
to three months after referral. Collectively, in the au-
thors’ series and the literature, functional recovery after
nonsurgical treatment was seen in 37 (93%) of 40
cases.4–10 The authors conclude that many cervical frac-
tures are amenable to nonsurgical treatment because this
approach offers a high likelihood of success.
Definitive criteria for selection of dogs with cervical
fractures that would benefit from surgical treatment (in-
stead of nonsurgical treatment) have not been estab-
lished. General indications for surgical treatment have
been proposed, including vertebral displacement or com-
minution, deteriorating neurological status amidst
conservative management, or persistent pain.4,13 A retro-
spective case review cannot provide definitive conclu-
sions regarding the superiority (or inferiority) of surgical
treatment compared to nonsurgical treatment of cervical
fractures. Instead, the authors sought to identify a subset
of cervical fractures that was associated with a poor
outcome after nonsurgical treatment. If identified, these
cases would have provided valuable insight into fracture
characteristics that might render cases unsuitable for
nonsurgical management. However, almost 90% of
nonsurgically treated cases had functional recovery. As a
result, particular types of cervical fractures that would
mandate surgical treatment were not identified.
Several possible predictors (i.e., inciting cause, level
of affected vertebra, multiplicity, time interval from
trauma to referral, severity of neurological deficits on
presentation, concurrent head trauma) of functional out-
come were evaluated. Overall, nonambulatory dogs were
l3 times less likely to have functional recovery than
ambulatory dogs. However, severity of neurological defi-
cits was not predictive of functional recovery in
nonsurgically treated dogs. This may reflect that
nonambulatory dogs in this series were more likely to be
euthanized prior to treatment (perhaps due to concurrent
non-neurological injury). Nonambulatory dogs may also
have been more likely to undergo surgical treatment
Cervical Vertebral Fractures 145
which was associated with high perioperative mortality.
Importantly, the results indicate that nonambulatory dogs
with cervical fractures do not have a grave prognosis. In
fact, once the decision was made to pursue nonsurgical
treatment, there was no difference in prognosis for
nonambulatory and ambulatory dogs.
A prolonged interval from trauma to referral was a
consistent negative predictor of functional recovery.
Overall, dogs with a trauma to referral interval of five
days or more were 5.5 times less likely to achieve func-
tional recovery than dogs referred less than five days
after trauma. The strong negative association between
trauma to referral interval and functional recovery also
was seen in the 28 nonsurgically treated dogs. One pos-
sible explanation is that dogs with delayed referral may
have an unstable vertebral fracture which results in pro-
gressive spinal cord damage. This instability may be
related to the nature of the fracture or may reflect inad-
equate spinal stabilization during the post-traumatic pe-
riod. Alternatively, it may reflect diminished owner
enthusiasm to pursue treatment; lack of owner commit-
ment ultimately leads to a less favorable outcome.
Conclusion
Dogs that undergo nonsurgical treatment of their cervi-
cal fractures have a high likelihood of functional recov-
ery. Due to the high perioperative mortality associated
with cervical spinal surgery, it is imperative that specific
criteria be defined that justify surgical intervention. De-
lay in presentation to the referral institution decreases
the likelihood of functional recovery. Accordingly, early
neck immobilization, prompt referral, or both are recom-
mended to increase the likelihood of successful outcome
in dogs with cervical fractures.
References
1. Matthiesen DT. Thoracolumbar spinal fracture/luxations: surgical
management. Comp Cont Ed Pract Vet 1983;5:867–78.
2. Carberry CA, Flanders JA, Dietze AE, et al. Nonsurgical management of
thoracic and lumbar spinal fractures and fracture/luxations in the dog and
cat: a review of 17 cases. J Am Anim Hosp Assoc 1989;25:43–54.
3. Selcer RR, Bubb WJ, Walker TL. Management of vertebral column
fractures in dogs and cats: 211 cases (1977–1985). J Am Vet Med Assoc
1991;198:1965–8.
4. Stone EA, Betts CW, Chambers JN. Cervical fractures in the dog: a
literature and case review. J Am Anim Hosp Assoc 1979;15:463–71.
5. Gage ED. Surgical repair of fractured occiput, atlas, and axis in a dog.
J Am Vet Med Assoc 1971;158:1951–3.
6. Gage ED. Surgical repair of fractured cervical spine in the dog. J Am Vet
Med Assoc 1968;153:1407–12.
7. Geary JC. Traumatic lesions of the central nervous system. J Am Anim
Hosp Assoc 1971;7:296–317.
8. Gendreau CL, Cawley AJ. Repair of fractures of the axis. Can Vet J
1969;10:297–301.
9. Hoerlein BF. Traumatic lesions of the canine spine. Mod Vet Pract
1958;39:24–9.
10. Oliver JE, Lewis RE. Lesions of the atlas and axis in dogs. J Am Anim
Hosp Assoc 1973;9:304–13.
(Continued on next page)
146 JOURNAL of the American Animal Hospital Association March/April 1999, Vol. 35

References (cont’d)
11. Wong WT, Emms SG. Use of pins and methylmethacrylate in stabilization
of spinal fractures and luxations. J Sm Anim Pract 1992;33:415–22.
12. Spackman CJA, Caywood DD, Feeney DA. Postoperative complication of
fracture repair in a dog. J Am Vet Med Assoc 1984;185:1004–6.
13. Denny HR. Fractures of the cervical vertebrae in the dog. Vet Annual
1983;23:236–40.
14. Hurov L. Dorsal decompressive cervical laminectomy in the dog: surgical
considerations and clinical cases. J Am Anim Hosp Assoc 1979;15:301–9.
15. Blass CE, Waldron DR, van Ee RT. Cervical stabilization in three dogs
using Steinmann pins and methylmethacrylate. J Am Anim Hosp Assoc
1988;24:61–8.
16. Rouse GP. Cervical spine stabilization with methylmethacrylate. Vet Surg
1979;8:1–6.
17. Steyn DG. The use of methyl methacrylate bone cement as an internal
splint in the treatment of fractures of the canine axis. J S Afr Vet Assoc
1986;57:239–41.
18. Samuels ML. Statistics for the life sciences. Englewood Cliffs: Prentice
Hall, 1989:229–31.
19. Thomas WB, Sorjonen DC, Simpson ST. Surgical management of
atlantoaxial subluxation in 23 dogs. Vet Surg 1991;20:409–12.
20. Clark DM. An analysis of intraoperative and early postoperative mortality
associated with cervical spinal decompressive surgery in the dog. J Am
Anim Hosp Assoc 1986;22:739–44.
21. Waters DJ. Spinal surgery. In: Lipowitz AJ, Caywood DD, Newton CD,
Schwartz A, eds. Complications in small animal surgery. Baltimore:
Williams & Wilkins, 1996:541–62.
 Evaluation of Joint Stabilization for
Treatment of Shearing Injuries
of the Tarsus in 20 Dogs
Medical records of 20 dogs with 23 shearing injuries of the tarsus leading to joint instability
were reviewed. A transarticular external skeletal fixation device or prosthetic ligament was used
to stabilize the joints. The most common complications were fixator failure and implant
infection. The median times for wound healing and maximal joint function were 10 and 12
weeks, respectively. Clinical outcome was excellent in 22%, good in 56%, and poor in 22%.
Comparison of the two stabilization methods showed no statistically significant differences in
healing time, time to regain function, or clinical outcome. J Am Anim Hosp Assoc 1999;35:147–53.

David W. Diamond, VMD Introduction

Juliette Besso, DEDV
Randy J. Boudrieau, DVM,
Diplomate ACVS
RS
Shearing injuries of the hind limb are encountered commonly in veteri-
nary medicine. 1,2 These injuries are caused when an animal is hit by a
motor vehicle and dragged along the abrasive surface of the road.3,4 The
resulting damage can range from simple laceration of the skin to the loss
of large areas of skin, ligament, and bone with heavily contaminated,
devitalized soft tissue. The collateral ligament of the tarsus often is lost,
even in relatively small wounds, due to the prominent position of the
malleolus,2,5–9 resulting in either medial or lateral instability of the tibio-
tarsal joint. Primary reconstruction of the damaged ligament is generally
not possible because of the loss of large amounts of soft tissue; addition-
ally, these wounds often require open wound management techniques and
are allowed to heal by second intention.2,5–9 Immediate stabilization of
these joints using either a transarticular external skeletal fixator or syn-
thetic replacement of the collateral ligament has been recommended.2,5–9
Alternatively, wound management without surgical stabilization of the
joint has been reported to result in a good or excellent outcome in a
majority of cases.10
The purpose of this study was to evaluate the long-term clinical and
radiographic outcomes for dogs managed with early surgical stabilization
of the joint using transarticular external skeletal fixation or prosthetic
ligament placement and open wound management techniques where
indicated.

Materials and Methods

From the Department of Clinical Sciences,
Tufts University School of
Veterinary Medicine,
200 Westboro Road,
North Grafton, Massachusetts 01536.
Address reprint requests to Dr. Boudrieau.
The medical records of dogs with shearing wounds of one or both tarsi,
with collateral ligament loss and joint instability, and treatment with a
surgical stabilization procedure (other than arthrodesis) performed at the
Tufts University School of Veterinary Medicine (TUSVM) between Au-
gust 1987 and August 1996 were reviewed. Information obtained from the
medical records included signalment, etiology, other concurrent prob-
lems, extent and location of wounds, time until surgical joint stabiliza-
tion, methods of stabilization and wound management, duration of
hospitalization, cultures and antibiotic therapy, time until healing, time
until maximal joint function, and complications.

JOURNAL of the American Animal Hospital Association 147

148 JOURNAL of the American Animal Hospital Association
Clinical reevaluations were made whenever possible
at TUSVM. The clinical evaluation included complete
physical and orthopedic examinations. Any gait abnor-
malities were recorded, and the presence or absence of
pain, crepitation, and swelling of the affected joint(s)
were assessed. Animals were tranquilized, tibiotarsal
joint range of motion was measured with a goniometer,
and the joints were radiographed. Follow-up telephone
interviews with the owner or referring veterinarian or
both were performed to determine time to wound heal-
ing, time to recover functional limb use, any gait abnor-
malities, and assessment of outcome.
All preoperative, immediate postoperative, and fol-
low-up radiographs were evaluated by a single radiolo-
gist (Besso). An assessment of the degree of soft-tissue
and bony involvement was estimated from the original
preoperative radiographs. Based upon the severity of the
initial bony injury, the wounds were classified into one
of four categories: mild (no bone loss, collateral liga-
ment injury only); moderate (medial or lateral malleolus
loss); severe (malleolus and talus involvement); and very
severe (malleolus, talus plus other tarsal bone involve-
ment). The surgical method of joint stabilization was
described from the immediate postoperative radiographs.
The amount of degenerative joint disease (DJD) ob-
served on follow-up radiographs was classified into one
of four categories: none, mild, moderate, and severe.
Implant status and any other changes were also recorded.
Overall outcome was based upon reexamination at
TUSVM and telephone interviews with the owner and
referring veterinarian. Outcomes were assigned one of
the following designations: excellent (lameness never
detected); good (lameness only following exercise or
associated with inclement weather); and poor (lameness
frequently observed or persistent gait abnormality). When
the owner’s assessment and the in-hospital evaluation
were not in agreement, the case was assigned the lower
clinical outcome.
Correlations between the effects of concurrent injury
and final clinical outcome were assessed using Spear-
man’s rank correlation test (p less than 0.05). Time until
complete wound healing, time to return of maximal func-
tion, and final outcome were compared between the two
methods of stabilization (i.e., external fixator and pros-
thetic ligament). Analysis of variance (ANOVA) was
used to assess statistical significance (p less than 0.05).
Correlations between the original wound grade, the
amount of DJD that developed, and the final clinical
outcome were assessed for each method of stabilization
using Spearman’s rank correlation test (p less than 0.05).
Results
The medical records of 25 dogs with shearing injuries
and instability of one or both tarsal joints were reviewed.
Complete follow-up information was available for 20
dogs: 15 of 20 in hospital (17 tarsal injuries), five of 20
March/April 1999, Vol. 35
by telephone interview (six tarsal injuries). Five dogs
were lost to follow-up. Several breeds were represented:
14 purebreds (four of which were German shepherd dogs)
and six mixed-breed dogs. There were 12 males (nine
intact) and eight females (one intact). The age at the time
of injury ranged from six months to 10.6 years (mean,
3.8 yrs). All of the wounds were caused by trauma with a
motor vehicle. Most of the injuries occurred during the
warmer months of the year, with 75% of the injuries
happening between May and October. Concurrent inju-
ries were present in 15 of 20 dogs and included other
integumentary lesions (n=7), pulmonary injuries (n=5),
fractures (n=5), luxations (n=2), and peripheral neuro-
logical injuries (n=1). No correlation was observed be-
tween concurrent injuries and outcome for either of the
stabilization methods performed. Of the 23 tarsal shear-
ing injuries in the 20 dogs, 18 (78.3%) of 23 were me-
dial, and five (21.7%) of 23 were lateral.
Emergency medical management and supportive care
were administered until the wound(s) could be addressed
definitively. In most cases, the dogs were anesthetized
within 24 hours of presentation for the initial lavage and
debridement of the wounds. All bandages used wet-to-
dry dressings consisting of gauze sponges moistened
with lactated Ringer’s solution, subsequently covered by
dry gauze sponges, cotton padding, and an outer wrap. In
some cases, when joint instability was readily apparent
to the admitting doctor, splinting material was added to
the initial bandage; no splinting material was used once
the stabilization procedure was performed.
Radiographs of the affected limb(s) were obtained in
each dog prior to the stabilization procedure. Since in-
consistent or incomplete wound descriptions were present
in the medical records, wounds were graded from radio-
graphic assessment as described. The wounds were
graded as follows: mild in six (26.1%) of 23 cases,
moderate in eight (34.8%) of 23 cases, severe in four
(17.4%) of 23 cases, and very severe in five (21.7%) of
23 cases.
The joint stabilization procedures were performed an
average of 3.8 days (range, zero to 10 days) after the
injury. The fixators were applied an average of 4.7 days
after the injury, and the prosthetic ligaments were ap-
plied an average of 3.6 days after the injury. Whenever
possible, the skin margins were drawn toward each other
without tension, using mattress sutures, in order to mini-
mize the size of the remaining defect. Only three (13%)
of the wounds could be closed primarily (over drains) at
the time of the stabilization procedure. All of the subse-
quent bandage changes were performed at least once
daily, using sedation when required, until the need for
debridement decreased. As granulation tissue developed,
the bandages were changed every two to three days until
nonadherent dressings could be used. Bandage changes
were then performed once or twice weekly until the
wounds were healed or required skin grafting. Two
March/April 1999, Vol. 35
wounds required full-thickness skin grafting procedures
to cover the remaining defects. All of the original wounds
were healed at the time of reevaluation except for a small
draining tract on one of the wounds of the dog that had
required skin grafting. The median time to complete
wound healing for all of the cases was 10.0 weeks (range,
3.4 to 52 wks; standard error, 1.1 wk). For the trans-
articular external skeletal fixator (ESF) group, median
healing time was 10.0 weeks (range, 3.5 to 52 wks); for
the prosthetic ligament group, median healing time was
8.0 weeks (range, 3.4 to 30 wks).
Seven of the 23 joints had an ESF applied as the sole
means for stabilization. The ESFs were all unilateral-
uniplanar (Type I) configurations; rather than place the
ESF directly into the open wound, it was placed on
whichever side would allow the pins to penetrate healthy
skin. Five of the ESFs had an additional connecting bar
placed between the first and last pin, creating a triangu-
lar configuration. Fifteen of the 23 joints had prosthetic
collateral ligament reconstruction with 2.7 or 3.5 mm
screwsa and either 18 or 20 gauge orthopedic wireb (n=13)
or polyester suture materialc (n=2). Detailed methods for
application of the prosthetic ligament have been de-
scribed elsewhere.5–8 The one remaining dog had its
joint stabilized with both a bilateral-uniplanar (Type II)
ESF and a polyester prosthetic ligament.c
Hospitalization ranged from one to 23 days (mean,
9.2 days). Empirical antibiotic therapy was administered
in 17 of 20 dogs; however, only five wounds were cul-
tured. Antibiotic therapy ranged from one to 49 days
(mean, 16.1 days). Antibiotic therapy also did not corre-
late with outcomes in either treatment group.
The median time until the animal reached the peak
level of function was 12.0 weeks (range, 2.5 to 52 wks;
standard error, 1.8 wks). For the ESF group, median
functional recovery time was 10.0 weeks (range, 3.5 to
18 wks); for the prosthetic ligament group, median func-
tional recovery time was 12.0 weeks (range, 2.5 to 52
wks). Summary data for each of the methods of stabiliza-
tion (including grade of the wound, concurrent injuries,
time until ESF or prosthetic ligament removal, compli-
cations, presence of DJD on follow-up radiographs, clini-
cal outcome, and length of follow-up) is presented in the
Table.
Complications with the ESF group were seen in three
dogs and involved failure of the fixator by either implant
failure (i.e., broken connecting bar) or premature re-
moval by the animal during rough play. These complica-
tions were remedied by reapplication of another ESF.
Most of the ESFs (many of which had superficial pin
tract infections) had loose fixation pins by the time of
their final removal. There were no problems with pin
tract healing after removal.
Complications (and their time of onset) associated
with the prosthetic ligament technique, which required
implant removal, were observed in seven (43.8%) of 16
Shearing Injuries of the Tarsus 149
dogs: four infections (mean, 7.8 mos; range, 2.4 to 15.5
mos); two implant (screw) loosenings sufficient to cause
discomfort (mean, 2.1 mos; range, 1.3 to 3.0 mos); and
one broken wire causing lameness (at 36 mos). All of
these complications resolved after the implants were
removed. In one dog, the implants were removed at three
months postoperatively when the wounds were closed
with full-thickness, free skin grafts. (No problems were
associated with the implants at this time.)
Follow-up radiographs of 21 joints in 18 dogs were
obtained. Time to the final radiographic evaluation
ranged from 1.5 months to 6.4 years (mean, 2.1 yrs).
Degenerative joint disease was present in 17 (81%) of
the 21 joints evaluated. The presence of DJD was classi-
fied as follows: mild (six of 21; 28%), moderate (seven
of 21; 33%), and severe (four of 21; 19%). The average
times until the follow-up radiographic evaluation for
each of these classifications were: 7.8 months for the
joints with no evidence of degenerative changes (range,
1.5 mos to 2.1 yrs); 2.8 years for mild DJD (range, 6.4
mos to 6.4 yrs); 2.0 years for moderate DJD (range, 1.4
to 3.8 yrs); and 2.7 years for severe DJD (range, 1.4 to
5.9 yrs).
At the time of reevaluation, average angles of maxi-
mum flexion for the unaffected and affected tarsi were
found to be 44˚ (range, 28˚ to 58˚) and 80˚ (range, 45˚ to
115˚), respectively. Average angles of maximum exten-
sion were 157˚ (range, 154˚ to 162˚) and 151˚ (range,
135˚ to 168˚), respectively. Crepitation was palpable in
seven (41.1%) of the 17 joints evaluated at TUSVM. No
pain was associated with flexion or extension of any of
the joints. There was muscle atrophy present in seven
(41.1%) limbs, as compared to the contralateral limb.
Gait abnormalities were observed with four (23.5%) dogs
as follows: mild stiffness (n=2) or frequent skipping at a
walk (n=2).
Owner assessment of their dog’s clinical outcome
was found to be graded better than the in-hospital evalu-
ation in five (21.7%) of 23 cases (two excellent ratings
by the owner lowered to good, and three good ratings
lowered to poor). For the ESF group, clinical outcome
was excellent in one (14.3%) of seven cases, good in five
(71.4%) of seven cases, and poor in one (14.3%) of
seven cases. For the prosthetic ligament group, outcome
was excellent in three (20%) of 15 cases, good in eight
(53.3%) of 15 cases, and poor in four (26.7%) of 15
cases. The case that was stabilized with both an ESF and
prosthetic ligament had an excellent clinical outcome.
For all of the 23 wounds combined, the outcome was
excellent in five (21.7%) cases, good in 13 (56.5%)
cases, and poor in five (21.7%) cases. Analysis of vari-
ance showed no significant differences in time until com-
plete wound healing, time to return of maximal function,
or final outcome between the two methods of stabiliza-
tion. Spearman’s rank correlation analysis showed no
statistical correlation within and between methods of
 Table 150

Summary Data: Tarsal Shearing Injuries Treated by Joint Stabilization

(Transarticular External Skeletal Fixation or Prosthetic Replacement of Collateral Ligament)

Time of Wound§ Maximal Radiographic\ Clinical

Age Weight Wound Concurrent Removal Healing Function Follow-up Follow-up
(yrs) Sex* Breed (kg) Grade Injuries† Repair Complications‡ (wks) (wks) (wks) (yrs) DJD (yrs) Outcome
External Skeletal Fixator Group Side No. pins Support bar
1 5.4 FS Gordon 29.8 Mild Medial 4 Yes 6.6 9.0 18.0 1.6 Severe 1.6 Excellent
setter
2 4.7 FS German 31.4 Moderate Medial 5 Yes 3.6 10.0 6.5 5.9 Severe 5.9 Good
shepherd
dog
3 8.1 M Irish 35.0 Moderate iW Lateral 5 No FP at 7 days 6.3 3.5 10.0 0.1 None 3.3 Good
setter
4 2.0 M Great 57.0 Severe cW Medial 4 No CB at 30 days 8.4 12.0 18.0 2.5 Moderate 2.5 Good
Dane
5 0.8 M Labrador 30.0 Severe PT, PC, Medial 5 Yes 2.7 8.0 6.0 1.4 Moderate 1.4 Good
JOURNAL of the American Animal Hospital Association

retriever RN
6 1.1 FS Australian 25.5 Very iFxH Lateral 7 Yes FP at 8 wks 19.6 52.0 3.5 6.4 Mild 6.4 Good
shepherd severe
7 10.6 FS Mix 34.5 Very FxS Lateral 8 Yes 12.7 12.0 12.0 0.3 None 2.5 Poor
severe
Prosthetic Ligament Group Side No. screws Material
1 1.0 M Great 55.6 Mild iW Lateral 1 Wire 8.0 14.0 1.6 Mild 1.6 Excellent
Dane
2 0.5 M Springer 14.2 Mild Lateral 3 Polyester I 42.8 1° 8.0 3.8 Moderate 3.8 Excellent
spaniel closure
3 2.5 M German 33.8 Mild Medial 2 Wire 3.4 18.0 0.1 None 7.3 Good
shepherd
dog
4 3.1 MC Mix 24.6 Mild cW Medial 1 Wire 8.0 3.5 0.5 Mild 0.5 Good
5 10.6 FS Mix 34.5 Mild FxS Medial 3 2 Wires S 12.8 12.0 12.0 1.7 Moderate 2.5 Poor
6 1.2 M Basenji 11.4 Moderate iFxH, cFxH Medial 2 Wire S 5.6 26.0 2.5 2.1 None 2.1 Excellent
March/April 1999, Vol. 35
 Table (Cont’d)
Summary Data: Tarsal Shearing Injuries Treated by Joint Stabilization
(Transarticular External Skeletal Fixation or Prosthetic Replacement of Collateral Ligament)

Time of Wound§ Maximal Radiographic\ Clinical

March/April 1999, Vol. 35

Age Weight Wound Concurrent Removal Healing Function Follow-up Follow-up

(yrs) Sex* Breed (kg) Grade Injuries† Repair Complications‡ (wks) (wks) (wks) (yrs) DJD (yrs) Outcome
Prosthetic Ligament Group (cont’d) Side No. screws Material
7 0.5 M Springer 14.2 Moderate Medial 2 Wire W 169.5 1° 8.0 3.8 Mild 3.8 Good
spaniel closure
8 4.1 F German 41.8 Moderate PC Medial 3 Polyester I 14.6 10.0 20.0 2.1 Mild 2.1 Good
shepherd
dog
9 5.1 MC Mix 57.0 Moderate PC, cCF Medial 1 Wire 3.5 14.0 1.6 Severe 1.6 Good
10 0.7 M Mix 20.3 Moderate PC, iW Medial 2 Wire I 10.3 8.0 6.0 2.5 Mild 2.5 Poor
11 1.5 MC Dalmatian 25.1 Severe PC, iW Medial 2 Wire 12.0 52.0 NT NT 6.7 Good
12 10.1 M German 30.5 Severe FxF Medial 1 Wire SGF 12.8 30.0 12.0 1.4 Moderate 1.4 Poor
shepherd
dog
13 2.4 FS Mix 17.2 Very SI, cFxA Medial 1 Wire 8.0 14.0 NT NT 6.2 Good
severe
14 2.5 FS Labrador 32.7 Very cW Medial 1 Wire 12.0 8.0 1.7 Moderate 1.7 Good
retriever severe
15 10.1 M German 30.5 Very FxF Lateral 2 Wire SG 12.8 12.0 12.0 1.4 Moderate 1.4 Poor
shepherd severe
dog
External Skeletal Fixator with Prosthetic Ligament Type II 4 Pins No bar
1 6.5 FS Mix 36.3 Moderate Medial 3 screws Polyester I 66.3 1° 8.0 1.4 Severe 1.8 Excellent
closure

* FS=neutered female; M=male; MC=neutered male; F=female

†
iW=ipsilateral wound (laceration/abrasion); cW=contralateral wound (laceration/abrasion); PT=pneumothorax; PC=pulmonary contusions; RN=radial nerve neurapraxia; iFxH=ipsilateral
hind-limb fracture; FxS=sacral fracture; cFxH=contralateral hind-limb fracture; cCF=contralateral coxofemoral luxation; FxF=facial and dental fractures; SI=bilateral sacroiliac luxations;
cFxA=contralateral acetabular fracture
‡
FP=fixator pulled out; CB=broken connecting bar; I=infection; S=loose screw; W=broken wire; SGF=skin graft failure; SG=skin graft
§
1˚ closure=wounds that could be closed at the time of fracture stabilization
Shearing Injuries of the Tarsus

\
DJD=degenerative joint disease; NT=none taken
151
152 JOURNAL of the American Animal Hospital Association
stabilization for the following: original wound grade and
the amount of DJD that developed; the original wound
grade and the final clinical outcome; and the amount of
DJD present and the final clinical outcome.
Discussion
It has been reported that 72% of shearing injuries of the
tarsus result in instability of the joint.10 In the present
study, only dogs with tarsal instability that had a surgical
stabilization procedure done were evaluated; therefore,
no incidence rates were determined. As previously re-
ported, concurrent injuries (including lacerations, pul-
monary injuries, and other fractures) often occur1,10,11
and were present in 75% of the animals in this study.
These findings reinforce the importance of a thorough
physical examination of the entire animal when pre-
sented with orthopedic trauma.10
The present study confirmed that the medial aspect of
the tarsus is more commonly affected,10 as more than
75% of the wounds occurred medially. The extent of
damage to the tarsal region was variable; approximately
60% of the wounds had mild or moderate grade injuries
involving the collateral ligament and malleolus, while
the remaining 40% had more severe trauma involving
the tarsal bones.
All of the wounds were managed initially as open,
contaminated wounds. Only three (13%) of the wounds
were sufficiently minor, with adequate skin and soft
tissue present to allow for primary closure of the wound,
while four (17%) wounds took six months or longer to
heal by second intention; however, the median time for
the remaining wounds to heal by second intention was
10.0 weeks. This healing time was longer than the aver-
age 6.7 weeks reported for complete wound healing in
another study in which stabilization procedures were not
performed.10 No statistical comparison could be made
with that study, due to a lack of descriptive analysis and
reporting of standard errors.10 A number of factors could
account for the differences observed. One important fac-
tor involved in wound healing is the original size of the
wound, which could not be assessed due to the retrospec-
tive natures of both studies. The presence of an implant
within the wound, as described in the dogs that were
treated with the prosthetic ligament technique, could
have resulted in the increased healing times observed. It
also could be expected that the healing time for the ESF
group would be longer as a result of the detrimental
effects of immobilization. Passive physiotherapy of the
wound caused by motion stimulates circulation, provid-
ing for improved defense against infection and enhanced
healing of tissues.12 It has also been suggested that move-
ment may loosen adhesions and support drainage from
the wound through a “massaging” action.12 The data in
this study did not demonstrate any statistically signifi-
cant difference in the median healing times between the
ESF group and the prosthetic ligament group; how-
March/April 1999, Vol. 35
ever, the small sample size in this study must be
considered.
Many different antibiotics were administered both
enterally and parenterally to most of the dogs in this
study. Those cases that did not receive antibiotics were
not adversely affected with respect to clinical outcome
or wound healing. There was also no apparent advantage
attributable to any one of the various antibiotics that
were used. The benefit from the use of antibiotics is
probably minimal when aggressive open wound man-
agement techniques are employed. Once mature granula-
tion tissue has become established, antibiotic usage
presumably is unnecessary as this tissue is reported to be
resistant to infection.4
Complications observed in the ESF group may have
been avoided by increasing the stiffness of the ESF
frame. This could be easily accomplished by using a
Type II ESF frame rather than a Type I or by consistently
adding another connecting bar between the first and last
fixation pins, creating a triangular configuration.
All complications observed in the prosthetic ligament
group were associated with the presence of the implant.
Forty-four percent of the implants in this group required
removal an average of 10.7 months postoperatively be-
cause of infection, pain, or lameness caused by broken or
loose implants. In three of the four cases with infection,
a polyester suture material was used, and the infection
resolved after the material was removed. Polyester su-
ture material is a braided, nonabsorbable, synthetic ma-
terial which (although strong enough for use as a
replacement ligament) has the ability to harbor bacteria
within the crevices of the braids and probably should not
be used in the face of open wounds.13,14
Time to regain maximal function of the limb has not
been reported previously. For all of the animals in this
study, the median recovery time was 12.0 weeks. Analy-
sis of the data between the two groups in this study
suggests that the method of stabilization had no statisti-
cally significant effect on the overall length of the recov-
ery period. Although the method of stabilization used
did not appear to be a factor in the length of recovery, the
potential problems associated with each method need to
be considered.
Radiographic follow-up data for tarsal shearing inju-
ries has not been reported previously; therefore, the ef-
fect of stabilization on the development of DJD could
not be made. No statistical difference was observed be-
tween the frequency, severity, or time to development of
degenerative changes between the ESF and prosthetic
ligament groups. The data also indicates that the grade of
the original wound was not a reliable predictor for the
development of DJD since no statistically significant
difference was observed between wound grade and the
development of DJD.
As expected with any joint injury, range of joint mo-
tion can be affected severely by periarticular fibrosis and
March/April 1999, Vol. 35
development of degenerative changes in and around the
joint. In this study, the range of motion, primarily flex-
ion, for the injured hock decreased by 37% compared to
the unaffected leg. No statistically significant difference
was observed between the ESF and prosthetic ligament
groups.
Clinically, the outcome for limb function and degen-
erative change did not correlate with the original wound
grade for either group or when all of the wounds were
considered together. Although there was no statistically
significant correlation between the clinical outcome and
the original wound grade, there appeared to be a trend
toward a more favorable outcome with the less trauma-
tized joints. In a similar study of shearing injuries that
were not stabilized, the percentage of cases that were
assigned an excellent outcome rating (53%)10 was far
greater than the percentage in the authors’ study (22%).
This holds true whether the transarticular fixator (14%)
or prosthetic ligament technique (20%) was used to sta-
bilize the joint. There are several possible explanations
for this difference: 1) The report cited 10 does not indicate
the length of time until follow-up, which, if it occurred
prior to the development of DJD, could support better
outcomes. 2) As discovered in this study, owner evalua-
tion of pet progress often was inflated over the truth. The
cited report10 does not distinguish how many evaluations
were made through in-hospital examinations or owner
interview by telephone. In-hospital evaluation, as ob-
tained in this report, frequently resulted in a more critical
assessment of the dog’s function than that provided by
the owner, and therefore explains a lower clinical out-
come. Additionally, the in-hospital evaluation allowed
objective evaluations of joint range of motion and devel-
opment of DJD. 3) Better clinical outcomes could be
secondary to variations in the sample populations for
each of the studies. The severities of the original injuries
are not mentioned in the earlier report.10
Conclusion
The data does not indicate greater efficacy of either
transarticular ESF or prosthetic collateral ligament place-
ment for treatment of tarsal shearing injuries. The lack of
correlation between parameters relating to the original
severity of the wound, the type of treatment employed,
and the long-term effects on the joint is indicative of the
complexity of tarsal shearing wounds and the difficulty
in predicting the final outcome.
When compared with a study evaluating similar
wounds treated without joint stabilization, the results of
this study do not indicate that stabilization, either with
an ESF or a prosthetic ligament, improves healing time
or final clinical outcome.10 However, the results reported
here may reflect a more critical evaluation of cases in
this versus the previous study. Although the joint stabili-
zation methods described do not demonstrate any greater
efficacy of these techniques when compared to treatment
Shearing Injuries of the Tarsus 153
methods without joint stabilization, it must be recog-
nized that the presence of the implants may adversely
affect the results. Further studies are necessary, directly
comparing stabilization and nonstabilization methods of
treatment, to determine if these impressions can be
documented.
a
Synthes; Paoli, PA
b
IMEX Veterinary, Inc., Longview, TX
c
#2 Ethibond; Ethicon, Summerville, NJ
Acknowledgments
The authors acknowledge the assistance of WM Rand,
PhD, for the statistical analysis, and the Hill’s Research
Fund for their support of this study.
References
1. Swaim SF, Pope ER. Early management of limb degloving injuries. Semin
Vet Med Surg (Sm Anim) 1988;3:274–81.
2. Gorse MJ. Traumatic derangement of tarsal and carpal collateral
constraints. Proceed, Am Coll Vet Surg Symposium 1996:159–61.
3. Vig MM. Management of integumentary wounds of extremities in dogs: an
experimental study. J Am Anim Hosp Assoc 1985;21:187–92.
4. Swaim SF. Management and bandaging of soft tissue injuries of dog and
cat feet. J Am Anim Hosp Assoc 1985;21:329–40.
5. Matthiesen DT. Tarsal injuries in the dog and cat. Compend Cont Ed Pract
Vet 1983;5:548–55.
6. Aron DN, Purinton PT. Replacement of the collateral ligaments of the
canine tarsocrural joint. A proposed technique. Vet Surg 1985;14(3):
178–84.
7. Earley TD, Dee JF. Trauma of the carpus, tarsus, and phalanges of the dog
and cat. Vet Clin N Am 1980;10:717–47.
8. Piermattei DL. Ligamentous injuries of the tarsus. Proceed, Am Anim Hosp
Assoc 55th Annual Meeting 1988:183–5.
9. Brinker WO, Piermattei DL, Flo GL. Handbook of small animal
orthopedics and fracture treatment. Philadelphia: WB Saunders, 1990:
438–41.
10. Beardsley SL, Schrader SC. Treatment of dogs with wounds of the limbs
caused by shearing forces: 98 cases (1975–1993). J Am Vet Med Assoc
1995;207:1071–5.
11. Kolata RJ, Kraut NH, Johnston DE. Patterns of trauma in urban dogs and
cats: a study of 1,000 cases. J Am Vet Med Assoc 1974;164:499–502.
12. Swaim SF. Management of contaminated and infected wounds. In: Swaim
SF, ed. Surgery of traumatized skin: management and reconstruction in the
dog and cat. Philadelphia: WB Saunders, 1980:119–213.
13. Stashak TS, Yturraspe DJ. Considerations for the selection of suture
materials. Vet Surg 1978;7:48–55.
14. Varma S, Johnson LW, Ferguson HL, Lumb WV. Tissue reactions to suture
materials in infected surgical wounds—a histopathologic evaluation. Am J
Vet Res 1981;42:563–70.
 Tarsometatarsal Subluxation in Dogs:
Partial Arthrodesis by Plate Fixation
In a retrospective study of tarsometatarsal joint subluxation in eight dogs, secondary fractures
were identified in six dogs, particularly of the fourth tarsal bone and the proximal fifth metatarsal
bone. Common causes of tarsometatarsal joint injury included jumping or falling and direct
trauma to the foot. Partial tarsal arthrodesis, with the use of bone-plate stabilization and
cancellous bone grafting of joint spaces after removal of articular cartilage, led to progressive
bone healing in all dogs. Implant breakage did not occur in any dog.
J Am Anim Hosp Assoc 1999;35:155–62.

Peter Muir, BVSc, MVetClinStud, Introduction

PhD, Diplomate ACVS Trauma to the canine foot may result in injury to soft tissue or bone or
Jana L. Norris both.1 Common causes of foot injury include motor vehicle trauma,
jumping or falling, or crush injury.2 Foot injury often results in joint
instability because of disruption to carpal or tarsal ligaments.3 Disruption
of tarsal ligaments may result in loss of stability at various joints, includ-
RS ing the talocrural,4,5 calcaneoquartal/talocentral,6 tarsometatarsal,6,7 talo-
calcaneal,8 and metatarsophalangeal joints.9 Lateromedial, plantar, or
dorsal instabilities all may occur because of ligament disruption.3
Although successful repair of tarsal collateral ligament injury has been
described,5,10 injuries which include disruption of the plantar ligament
complex usually necessitate arthrodesis to restore tarsus stability.2,3 Ar-
throdesis of the calcaneoquartal/talocentral joint has been achieved by use
of bone-screw fixation with or without tension-band wiring.6,11 External
skeletal fixation12 and intramedullary pinning7,13,14 have been used for
tarsometatarsal joint arthrodesis.7,13,14
Although results of tarsometatarsal arthrodesis have been described
for 25 dogs,6,7,12–15 detailed information is available for only 10 dogs.7,12
Use of lateral bone plating for stabilization of both calcaneoquartal/
talocentral and tarsometatarsal6 joint arthrodesis 6,11 has been recom-
mended because of concerns about implant breakage with screws or
pins. 11 However, results of tarsometatarsal arthrodesis using bone-plate
stabilization have not been described in detail. In this study, the authors
describe results of tarsometatarsal arthrodesis using bone-plate fixation in
eight dogs.

Materials and Methods

From the Department of Small Animal
Medicine and Surgery (Muir),
The Royal Veterinary College,
University of London,
Hawkshead Lane, North Mymms,
Hatfield, Hertfordshire AL9 7TA,
United Kingdom, and the Department of
Surgical and Radiological Sciences (Norris),
School of Veterinary Medicine,
University of California-Davis,
Davis, California 95616.
Address all correspondence to Dr. Muir.
Animals
The medical and radiographic records of dogs admitted to the Queen
Mother Hospital for Animals at The Royal Veterinary College (1990
through 1996), the Veterinary Medical Teaching Hospital at the Univer-
sity of California-Davis (1987 through 1996), and the Veterinary Medical
Teaching Hospital at the University of Wisconsin-Madison (1992 through
1995), and which were treated with partial tarsal arthrodesis using bone-
plate stabilization because of injury to the tarsometatarsal joint, were
reviewed retrospectively. Complete data was available for eight dogs.
The clinical record of each dog was examined to determine signalment,
diagnosis, and surgical treatment. History and physical examination were
obtained at the time of initial diagnosis.

JOURNAL of the American Animal Hospital Association 155

156 JOURNAL of the American Animal Hospital Association March/April 1999, Vol. 35

Figure 1A

Figures 1A, 1B, 1C, 1D—Mediolateral and dorsoplantar

radiographic views of the right tarsus of a two-year-old beagle
(case no. 1) with tarsometatarsal instability after motor vehicle
trauma. (A) Flexion of the tarsus demonstrates plantar instability.
(B) Lateral stress to the foot demonstrates medial instability.
Fracture of the fourth tarsal bone also is visible (arrow). (C)
Partial tarsal arthrodesis has been performed using a medial,
2.0-mm, dynamic compression plate. (D) Progressive healing of
the partial arthrodesis has occurred five weeks postsurgery.

Figure 1C

Radiography
Radiographic views of the tarsus were made and evalu-
ated to determine the location of any instability and
whether fractures were associated with ligament disrup-
tion [Figures 1A, 1B, 2A, 3A, 3B]. Stress radiographic
views16 were also made for case nos. 1 and 8 [Figures
1B, 3B]. Radiographic views were made immediately
after surgery to assess implant position and tarsal reduc-
tion [Figures 1C, 2B, 3C].

Surgery
Each case was treated by open reduction and internal
fixation using bone plating. After general anesthesia was
induced, a lateral or a medial approach (based on sur-
geon preference) to the tarsometatarsal joint was made.17
The tarsometatarsal joint then was exposed. Articular
cartilage was removed with a pneumatic burr. A cancel-
lous bone autograft, collected from the proximal hu-
merus, was placed in the joint spaces. After reduction of
the tarsus, a dynamic compression plate was applied to
the tarsus. After surgery, external coaptation, using ei-
Figure 1B ther a full fiberglass cast or a modified Robert-Jones
March/April 1999, Vol. 35
Figure 1D
bandage with or without a lateral fiberglass splint, was
applied to the tarsus for four to six weeks.
Follow-Up
During follow-up examination, additional history was
obtained, physical examination was performed, and fur-
ther radiographic views of the tarsus were made [Figures
1D, 2C, 3D]. These views were evaluated for evidence
of implant loosening, implant failure, and progressive
healing of the arthrodesis.
Results
Signalment and history are recorded in the Table. Com-
mon etiologies for tarsometatarsal subluxation included
jumping or falling or direct trauma to the foot. Second-
ary fractures of the metatarsal or tarsal bones were iden-
tified in six of seven cases [Figures 1B, 2B]. The fourth
tarsal bone and the proximal second and fifth metatarsal
bones were fractured most often. Osteophyte formation
due to the chronicity of the injury in case no. 8 prevented
accurate determination of the presence of secondary frac-
tures [Figure 3A]. Injury included disruption of the plan-
tar or medial ligamentous support in all cases. Plantar
instability was predominantly present in three cases; three
cases had combined medial and plantar instability; and
Tarsometatarsal Subluxation 157
Figure 2A
Figures 2A, 2B, 2C—Mediolateral and dorsoplantar radiographic
views of the left tarsus of a four-year-old, neutered female
Rhodesian ridgeback cross (case no. 3) with tarsometatarsal
instability after jumping over a fence. (A) Instability at the
tarsometatarsal joint is present. (B) Partial arthrodesis has been
performed using a lateral, 2.7-mm, dynamic compression plate.
Fracture of the proximal second metatarsal also is visible
(arrow). (C) Progressive healing of the partial arthrodesis has
occurred nine weeks postsurgery.
two cases had predominantly medial instability [Figures
1A, 1B, 2A, 3A, 3B].
Bone plates were applied to the lateral aspect of the
tarsus in seven cases [Figures 2B, 2C, 3C, 3D], and a
medial plate was applied in one case [Figures 1C, 1D].
Good reduction of the tarsus was achieved in all cases,
although revision of one lateral repair was performed
(case no. 2) because of rotational malalignment of the
foot. Implant breakage did not occur in any case, and a
stable arthrodesis was achieved in all patients. Compli-
cations such as pressure sores and swelling of the foot
were observed after surgery in association with use of
external coaptation [see Table].
Discussion
Joint instability because of injury to the plantar ligament
complex, which attaches to the base of the calcaneous
and the fourth and fifth metatarsal bones,2 is a more
prevalent cause of tarsal subluxation than dorsal inju-
158 JOURNAL of the American Animal Hospital Association
Figure 2B
ries.2,3 Dorsal subluxation was not seen in the dogs of
this report.
Acute traumatic injury is the most important cause for
instability of the tarsometatarsal joint6,13–15 and was the
cause of injury for most cases in the authors’ study.
Calcaneoquartal/talocentral joint instability is a more
prevalent injury compared with tarsometatarsal joint in-
stability.6 Progressive plantar ligament tearing and fail-
ure as a result of apparent minimal trauma often leads to
calcaneoquartal/talocentral joint instability, particularly
in Shetland sheepdogs and collies.6,18 Mechanical over-
load of the plantar ligaments in middle-aged, obese dogs
may be a major factor in the development of this in-
jury.6,18 Support from the fourth tarsal bone, which spans
the centrodistal articulation, is probably an important
factor in reducing the risk of instability developing at
this articulation.6
Fracturing of tarsal bones and proximal metatarsal
bones, particularly the fourth tarsal bone and the proxi-
mal fifth metatarsal bones, occurred secondary to insta-
bility of the tarsometatarsal joints and is typical of an
March/April 1999, Vol. 35
Figure 2C
acute traumatic etiology. This pattern of secondary frac-
turing has not been described in previous studies of
tarsometatarsal instability.7,12–15 These fractures appear
to have occurred because of medial to lateral bending
loads applied to the distal limb during injury.
Partial arthrodesis is considered the treatment of
choice for tarsal instabilities with plantar ligament dis-
ruption.2,6,7,11 Distractive forces will be present during
weight-bearing because of the location of the plantar
ligaments on the tension side of the tarsus,2 and this
appears to be an important reason why conservative
treatment or repair of the plantar ligament complex has
not been successful for dogs. External coaptation, which
had been used to treat two cases before presentation,
failed to restore joint stability. Use of bone-plate fixation
is a successful method of stabilization for partial arthro-
 Table
Case Histories of Eight Dogs With Partial Tarsal Arthrodesis by Plate Fixation for Tarsometatarsal Subluxation

Case Age Body Weight

No. Breed (yrs) Sex* (kg) Injury Treatment† Complications Clinical Outcome
1 Beagle 2 M 15 Right plantaromedial Medial, 2.0-mm, DCP; None Mild lameness five weeks
instability with fracture CBG; modified RJB postsurgery after splint
of the fourth tarsal bone, with fiberglass splint removal; progressive bone
March/April 1999, Vol. 35

because of vehicular trauma healing

2 Border 2 M 25 Left plantaromedial Lateral, 2.7-mm, DCP; Rotational Moderate lameness nine
collie instability with a CBG; fiberglass cast malalignment weeks postsurgery;
comminuted fracture of the necessitating surgical progressive bone healing
fourth tarsal bone, because revision; excessive
of a kick hock swelling;
wound dehiscence
3 Rhodesian 4 FS 31 Left plantar instability Lateral, 2.7-mm, DCP; Excessive limb No lameness nine weeks
ridgeback with fracture of the CBG; fiberglass cast swelling after cast postsurgery; progressive
cross proximal second metatarsal, application bone healing
after jumping over a fence
4 Boxer 3 FS 23 Left plantaromedial Lateral, 2.7-mm, DCP; Pressure sores No lameness nine weeks
instability, fracture of CBG; fiberglass cast postsurgery; progressive
proximal fifth metatarsal bone healing
bone, after jumping over a
fence
5 Labrador 5 M 29 Left plantar instability, Lateral, 3.5-mm, broad Coaptation before Minimal lameness six
retriever fracture of proximal second DCP; CBG; modified referral failed weeks postsurgery;
metatarsal bone, after RJB progressive bone healing
jumping over a fence
6 Doberman 2 MC — Left plantar instability, Lateral, 3.5-mm, narrow None Bout of lameness four
pinscher fracture of proximal second DCP; CBG; fiberglass months postsurgery
metatarsal bone, chip fracture cast which responded to
of proximal fifth metatarsal NSAID‡ medication;
bone, unknown etiology progressive bone healing
7 German 4 M 34 Left medial instability, Lateral, 4.5-mm, narrow None Minimal lameness at six
shepherd fractures of the proximal DCP; CBG; modified weeks; progressive
dog fourth and fifth metatarsal RJB bone healing
bones, after falling while
running
8 Rottweiler 5 F 49 Left medial instability of Lateral, 3.5-mm, narrow Coaptation before Good limb use five
several months duration, DCP; CBG; modified referral failed; weeks postsurgery;
with extensive osteophyte RJB with fiberglass pressure sores progressive bone healing
formation splint after surgery
Tarsometatarsal Subluxation

* M=male; FS=spayed female; MC=castrated male; F=female

†
DCP=dynamic compression plate; CBG=cancellous bone graft; RJB=Robert-Jones bandage
159

‡
NSAID=nonsteroidal anti-inflammatory drug
160 JOURNAL of the American Animal Hospital Association
Figure 3A
Figures 3A, 3B, 3C, 3D—Mediolateral and dorsoplantar
radiographic views of the left tarsus of a five-year-old, female
rottweiler (case no. 8) with chronic tarsometatarsal instability. (A)
Osteophyte formation has occurred around the tarsometatarsal
joint after the injury. (B) Medial instability was corrected by (C)
partial arthrodesis using a lateral, 3.5-mm, dynamic compression
plate. (D) Progressive healing of the arthrodesis has occurred
five weeks postsurgery.
desis of the tarsometatarsal joint based on the results
from this study, as a stable arthrodesis and progressive
bone healing was achieved in each case. Although it may
be possible to restore tarsal stability by formation of a
stable ankylosis after closed reduction and external skel-
etal fixation, open reduction and full surgical arthrodesis
with cartilage removal and cancellous bone autografting
were performed in the cases of this report to optimize
healing.
Stabilization with Steinmann pins has been recom-
mended for arthrodesis of the tarsometatarsal joint,7
March/April 1999, Vol. 35
Figure 3B
based on results from a small number of cases. Joint
stabilization with Steinmann pins or bone screws has
also been recommended for arthrodesis of the calcaneo-
quartal/centrodistal joints.6 However, implant breakage
is an important complication associated with pin or screw
stabilization of this more common joint instability.11
Implant breakage appears less prevalent if bone plates
are used for stabilization of calcaneoquartal/talocentral
joint arthrodeses. 11 Use of dynamic compression plates
also permits compression to be applied across the joint
spaces. Medical record data was not detailed sufficiently
to accurately determine the use of interfragmentary com-
pression in the cases of this report.
Many of the complications associated with treatment
of these cases were caused by use of external coaptation
after surgery. Use of a full cast probably is not necessary
after plate stabilization, as all cases healed without com-
March/April 1999, Vol. 35
Figure 3C
plications when only modified Robert-Jones bandages,
with or without splint bandages, were used. Use of modi-
fied Robert-Jones bandaging with a fiberglass splint
should minimize the risk of external coaptation-associ-
ated soft-tissue complications.
Partial arthrodesis was used to treat two cases (case
nos. 7 and 8) with principally medial instability of the
tarsometatarsal joints. Although ligament repair may
have been a reasonable treatment option in the absence
of plantar ligament disruption, partial arthrodesis was
chosen as a treatment because of concerns regarding
undetected injury to the plantar ligament tissue and the
presence of secondary fractures or osteoarthritis. Fur-
thermore, stable arthrodesis of these low motion joints is
associated with an excellent functional outcome.7
Use of stress radiography under general anesthesia
should form an integral part of patient evaluation and
will help to determine accurately which tarsal ligaments
Tarsometatarsal Subluxation 161
Figure 3D
are disrupted,16 although it was not used consistently in
the cases of this report. Although tarsometatarsal insta-
bility had a medial component in 63% of dogs, a bone
plate was placed on the medial aspect of the tarsus in
only one case. Placement of the plate on the most un-
stable side of the tarsus may facilitate anatomic reduc-
tion and rigid fixation of the injury.
Conclusion
The results of this study suggest that bone plating is a
successful fixation method for instabilities of the tar-
sometatarsal joints and is associated with a low compli-
cation rate and a low risk of implant breakage.
(Continued on next page)
162 JOURNAL of the American Animal Hospital Association March/April 1999, Vol. 35

References 10. Sjöström L, Håkanson N. Traumatic injuries associated with the short
collateral ligaments of the talocrural joint of the dog. J Sm Anim Pract
1. Basher A. Foot injuries in the dogs and cats. Comp Cont Ed Pract Vet 1994;35:163–8.
1994;16:1159–76.
11. Allen MJ, Dyce J, Houlton JEF. Calcaneoquartal arthrodesis in the dog.
2. Earley TD, Dee JF. Trauma to the carpus, tarsus, and phalanges of dogs and J Sm Anim Pract 1993;34:205–10.
cats. Vet Clin N Am Sm Anim Pract 1980;10:717–47.
12. Ardwedsson G. Arthrodesis in traumatic plantar subluxation of the
3. Matthiesen DT. Tarsal injuries in the dog and cat. Comp Cont Ed Pract Vet metatarsal bones of the dog. J Am Vet Med Assoc 1954;124:21–4.
1983;7:548–55.
13. Le Roux PH. Treatment of luxation of the tarsometatarsal joint of the dog.
4. Stoll SG, Sinibaldi KR, DeAngelis MP, Rosen H. A technique for J S Afr Vet Med Assoc 1971;42:195.
tibiotarsal arthrodesis utilizing cancellous bone screws in small animals.
14. Le Roux PH. Treatment of luxation of the tarsometatarsal joint of the dog.
J Am Anim Hosp Assoc 1975;11:185–91.
J S Afr Vet Med Assoc 1972;43:110–1.
5. Holt PE. Treatment of tibio-tarsal instability in small animals. J Sm Anim
15. Holt PE. Ligamentous injuries to the canine hock. J Sm Anim Pract
Pract 1977;18:415–22.
1974;15:457–74.
6. Campbell JR, Bennett D, Lee R. Intertarsal and tarsometatarsal subluxation
16. Farrow CS. Stress radiography: applications in small animal practice. J Am
in the dog. J Sm Anim Pract 1976;17:427–42.
Vet Med Assoc 1982;181:777–84.
7. Penwick RC, Clark DM. A simple technique for tarsometatarsal arthrodesis
17. Piermattei DL. An atlas of approaches to the bones and joints of the dog
in small animals. J Am Anim Hosp Assoc 1988;24:183–8.
and cat. 3rd ed. Philadelphia: WB Saunders, 1993:310–5.
8. Gorse MJ, Purinton PT, Penwick RC, Aron DN, Roberts RE. Talocalcaneal
18. Vaughan LC. Disorders of the tarsus in the dog II. Brit Vet J 1987;
luxation: an anatomic and clinical study. Vet Surg 1990;19:429–34.
143:498–505.
9. Van Ee RT, Blass CE. Arthrodesis of metatarsophalangeal joints in a dog.
J Am Vet Med Assoc 1989;194:82–4.
AAHA Annual Meeting,
4C Ad,
New

pg 163
The Genetic Connection
4C Ad,
new

pg 164
 Intermuscular Lipomas of the Thigh
Region in Dogs: 11 Cases
Ten dogs with intermuscular lipomas in the thigh region were treated by surgical resection. The
masses were located predominantly between the semitendinosus and semimembranosus
muscles and involved the full length of the femur. These lipomas were not infiltrative but located
deep between the fascial planes of the associated muscles. These tumors can appear similar
to soft-tissue sarcomas in this location, but they can be differentiated by cytology and histology.
Differentiation from an infiltrative lipoma is predominantly determined at the time of surgery. No
tumors recurred in the median follow-up period of 17 months.
J Am Anim Hosp Assoc 1999;35:165–7.

Maurine J. Thomson, BVSc, Introduction

MACVSc
Stephen J. Withrow, DVM,
Diplomate ACVS,
Diplomate ACVIM
William S. Dernell, DVM, MS,
Diplomate ACVS
Barbara E. Powers, DVM, PhD,
Diplomate ACVP
RS
From the Comparative Oncology Unit,
Department of Clinical Sciences
(Thomson, Withrow, Dernell) and
Department of Radiology and Radiation
Biology (Powers),
College of Veterinary Medicine and
Biomechanical Sciences,
Colorado State University,
Fort Collins, Colorado 80523.
Lipomas are frequently recognized mesenchymal tumors consisting of
localized nodules of fat, and they commonly occur in the subcutaneous
tissues of the ventral thorax and abdomen of middle-aged dogs.1 They
have occasionally been reported to occur within the thorax and abdomen
with clinical signs of organ compression in the affected body cavity.2,3
They only rarely recur after marginal surgical resection, and patients with
multiple lesions are considered to have separate tumors, not metastatic
disease.1 It is only their macroscopic appearance as a mass that has
justified their inclusion as a neoplasm. Necrosis, hemorrhage, and fibrosis
can occur in large or traumatized lipomas.4 Infiltrative lipomas also have
been recognized in dogs, histologically appearing as benign adipocytes
that readily invade muscle, fascia, and occasionally bone.5–7 They also
have been referred to previously in the literature as lipomatosis8 or well-
differentiated liposarcomas. 9 These tumors are fluctuant on palpation and
generally appear attached to surrounding structures.5 Cytological exami-
nation of infiltrative lipomas is consistent with simple lipomas, and a
definitive diagnosis is made by direct examination during surgery or on
histology. With aggressive surgical resection, the recurrence rate was
36% of 16 cases.5 Liposarcomas are typically pleomorphic soft-tissue
sarcomas, characterized by undifferentiated spindle cells that have a
finely vacuolated cytoplasm, and are mixed with recognizable lipocytes.
They are readily distinguishable from lipomas cytologically and histo-
logically and do not appear to arise from malignant transformation of
existing lipomas.4,9–11 The purpose of this paper is to present a separate
syndrome of simple lipomas which occur intermuscularly in the caudal
thigh region of dogs, predominantly between the semitendinosus and
semimembranosus muscles. Although the lipomas may appear foreboding
clinically, these intermuscular lipomas are not infiltrative and have an
excellent prognosis after surgical excision.
Methods
The medical records of dogs presented to the Colorado State University
Veterinary Teaching Hospital between 1989 and 1997 with a cytological
or histological diagnosis of a large thigh lipoma were reviewed. Informa-
tion obtained included patient signalment, tumor size and location, dura-
tion of disease, diagnostic evaluation, and outcome following surgical

JOURNAL of the American Animal Hospital Association 165

166 JOURNAL of the American Animal Hospital Association
Figure 1—Lateral radiograph demonstrating a mass in the
caudal thigh region of a dog. Arrows indicate the extent of the
mass, almost encompassing the full length of the femur. The soft-
tissue density is consistent with fat.
excision. Surgery involved incising over the region of
obvious swelling and separating the fascia of the muscles
surrounding the lipoma. The mass was able to be re-
moved generally by blunt dissection and digital extru-
sion. When the lipoma was found in the region of the
caudal thigh musculature, identification of the sciatic
nerve was performed initially, and in three cases the
mass required careful dissection from around the nerve.
Closure was performed routinely in three layers, and a
drain was inserted prior to closure in five cases. Follow-
up consisted of reexamination of the animal or phone
consultation with the owner or referring veterinarian.
Results
Ten dogs met the criteria for inclusion in this study. One
dog presented with two separate lesions over a 16-month
time span, for a total of 11 lesions in 10 dogs. The breeds
which were presented included Labrador retriever (n=4),
miniature schnauzer (n=3), cockapoo (n=1), and mixed-
breed (n=2). There were six spayed females, one intact
female, and three castrated males. The age at presenta-
tion ranged from seven to 12 years (median, 10 yrs).
Weight ranged from 9 kg to 55 kg (median, 28 kg).
Duration of clinical signs ranged from two weeks to two
March/April 1999, Vol. 35
years (median, three mos). The lesion usually appeared
as a slow-growing, firm, fixed mass in the caudal thigh
region. The size of the mass varied, generally involving
the full length of the femur. In four cases, a mild inter-
mittent lameness was associated with the lesion, and in
two cases the limb appeared suddenly swollen and pain-
ful. In the remaining five cases, no clinical lameness was
detected. Cytological evaluation of a fine-needle aspi-
rate was performed and was consistent with a lipoma in
all cases. Radiographs were taken in eight of the 11 cases
and revealed a soft-tissue density consistent with a li-
poma in the caudal or medial thigh region [Figure 1].
Each dog had the mass surgically excised with a mar-
ginal resection. The location of the lipoma was between
the muscles of the semitendinosus and semimembrano-
sus muscles in eight cases, between the semitendinosus
and biceps femoris muscles in one case, under the sarto-
rius muscle on the medial aspect of the thigh in one case,
and between the biceps femoris and lateral vastus muscles
in one case. There was no gross evidence of invasion into
muscle or fascia during surgical exploration. Tumors
from seven cases were examined histologically, which
confirmed mature adipose tissue without evidence of a
capsule or demarcation and no infiltration into the sur-
rounding musculature. In very large tumors, areas of
necrosis, inflammation, hemorrhage, and fibrosis were
seen frequently. Postoperatively, the dogs were usually
mild to moderately lame for several days; lameness re-
solved in all cases by the time of suture removal. Seroma
formation was the most common postoperative compli-
cation, occurring in four of six dogs which did not have a
drain placed following surgery. This resolved with con-
servative management in all cases. None of the five
cases with a Penrose drain placed developed a seroma.
Follow-up of the cases ranged from three to 36 months
(median, 17 mos). No evidence of recurrence of disease
was reported in any of the 11 cases.
Discussion
This case series describes a syndrome of lipomas of the
caudal thigh region in dogs that is neither subcutaneous
nor infiltrative. The general physical appearance of these
lesions is a very large, firm to fluctuant mass associated
with the caudal thigh musculature, which can appear
similar to soft-tissue sarcomas of this region. Cytology,
radiography, and histology were consistent with a li-
poma in all cases. Differentiation from soft-tissue sarco-
mas can be made on cytology and histology, but
differentiation from infiltrative lipomas is predominantly
made at the time of surgical exploration. It is unknown
why these large, fluctuant masses have a predisposition
for the caudal thigh musculature; a similar predisposi-
tion in the forelimbs has not been recognized.
Labrador retrievers are a common breed presented for
assessment of fatty tumors, and five of the 11 lesions in
this study were found in Labrador retrievers. These dogs
March/April 1999, Vol. 35
also had multiple, small, subcutaneous lipomas present
in various locations of the trunk. Miniature schnauzers
are not considered to be a breed very commonly afflicted
with lipomas, and in one study of 175 cases, only two
schnauzers were represented.1 It is therefore surprising
that three of the 10 dogs in this study with intermuscular
lipomas were miniature schnauzers. Of these three dogs,
one was grossly overweight, and the other two had mul-
tiple lipomas present. It appears that this breed is over-
represented in this study.
Conclusion
Intermuscular lipomas were found to be located in the
caudal thigh region predominantly between the semiten-
dinosus and semimembranosus muscles of mature, me-
dium-sized, female dogs. These lesions have a menacing
clinical appearance initially but are removed easily after
incision and separation of the fascia of the thigh muscu-
lature, though care is required when dissecting around
the sciatic nerve. It is recommended that a passive or
active drain be placed prior to closure to minimize seroma
formation. Intermuscular lipomas do not clinically or
histologically invade the muscles or fascia, and cytology
and radiography are suggestive of a fatty tumor. Each
Intermuscular Lipomas 167
case of intermuscular lipoma in this study had an excel-
lent prognosis after surgical excision.
References
1. Strafuss AC, Smith JE, Kennedy GA, Dennis SM. Lipomas in dogs. J Am
Anim Hosp Assoc 1973;9:555–61.
2. McLaughlin R, Kuzma AB. Intestinal strangulation caused by intra-
abdominal lipomas in a dog. J Am Vet Med Assoc 1991;199:1610–1.
3. Wilson DS, Hawe RS. Intrathoracic lipoma in a dog. J Am Anim Hosp
Assoc 1986;22:95–7.
4. Wilcock B. Neoplastic diseases of skin and mammary glands. In: Jubb
KVF, Kennedy PC, Palmer N, eds. Pathology of domestic animals. 4th ed.
Vol 1. San Diego: Academic Press, 1993:725.
5. Bergman PJ, Withrow SJ, Straw RC, Powers BE. Infiltrative lipomas in
dogs: 16 cases (1981–1992). J Am Vet Med Assoc 1994;205:322–4.
6. Kramek BA, Spackman CJA, Hayden DW. Infiltrative lipoma in three
dogs. J Am Vet Med Assoc 1985;186:81–3.
7. Frazier KS, Herron AJ, Dee JF, Altman NH. Infiltrative lipoma in a canine
stifle joint. J Am Anim Hosp Assoc 1993;29:81–3.
8. Berzon JL, Howard PE. Lipomatosis in dogs. J Am Anim Hosp Assoc
1980;16:253–7.
9. Saik JE, Diters RW, Wortman JA. Metastasis of a well-differentiated
liposarcoma in a dog and a note on nomenclature of fatty tumours. J Comp
Path 1987;97:369–73.
10. McCarthy PE, Hedlund CS, Veazy RS, Prescott-Mathews, Doo-Youn C.
Liposarcoma associated with a glass foreign body in a dog. J Am Vet Med
Assoc 1996;209:612–4.
11. Strafuss AC, Bozarth AJ. Liposarcoma in dogs. J Am Anim Hosp Assoc
1973;9:183–7.
AAHA Press Author Recruitment,
4C Ad,
new

pg 168
 Partial Colonic Obstruction Following
Ovariohysterectomy:
A Report of Three Cases
Partial extramural obstruction of the descending colon was diagnosed in two dogs and a cat as
a complication of elective ovariohysterectomy. In each case, the obstruction was caused by
fibrous tissue that encircled or crossed the descending colon, severely restricting the organ’s
normal mobility and luminal diameter. Clinical signs secondary to obstipation were observed in
two cases, five weeks and 27 months after elective ovariohysterectomy. In one dog without
clinical signs, the adhesion was an incidental finding during a laparotomy performed nine years
after the ovariohysterectomy. The fibrous adhesions were removed surgically in all three cases
without additional complications. J Am Anim Hosp Assoc 1999;35:169–72.

Bradley R. Coolman, DVM Introduction

Sandra Manfra Marretta, DVM,
Diplomate AVDC,
Diplomate ACVS
Melissa B. Dudley, DVM,
Diplomate ACVS
Scott M. Averill, DVM
C postoperative complication). Minor complications occurring during sur-
Ovariohysterectomy (OVH) is one of the most commonly performed
surgical procedures in dogs and cats.1,2 Many clients and veterinarians
regard the surgery as “routine” because the procedure is performed so
frequently. However, numerous potential complications are associated
with elective OVH.1–13 The frequency of complications following OVH
has been reported to be between 12.2% and 31.5% by various authors.1–3
Complications following OVH may be classified as minor (i.e., self-
limiting, easily resolved) or major (i.e., life-threatening, requiring exten-
sive treatment or additional surgery). Complications also may be classified
based on the time of occurrence (intraoperative, postoperative, or delayed
gery or within the first postoperative week are most common.1–3 Ex-
amples include reversible anesthetic complications, intraoperative
hemorrhage, postoperative pyrexia, patient-inflicted incisional trauma,
seroma, suture infection, and delayed healing.1–7 Less common major
complications include anesthetic death, severe hemorrhage/exsan-
guination, hydroureter/hydronephrosis, stump pyometra, recurrent estrus
from residual ovarian tissue, ovarian or uterine stump granulomas, drain-
ing fistulous tracts, adhesion formation, and bowel obstruction.1–13 Al-
though major complications following elective OVH in dogs and cats are
uncommon, it is nevertheless important for veterinarians to be aware of
these potentially serious complications when considering this surgical
procedure. Extramural colonic obstruction secondary to adhesion forma-
tion is a rare but potentially serious delayed complication of OVH in dogs
and cats and may require surgical correction.

From the Department of
Veterinary Clinical Medicine,
College of Veterinary Medicine,
University of Illinois,
Urbana, Illinois 61802.
Case Reports
Case No. 1
A 2.8-year-old, spayed female boxer was referred to the University of
Illinois for evaluation of a suspected colonic stricture and possible spinal
disorder. The dog initially had presented to the referring veterinarian four
days prior with complaints of restlessness, anorexia, vomiting, painful
arching of the back, and difficulty defecating for 24 hours. Physical
examination by the referring veterinarian revealed a depressed dog with a

JOURNAL of the American Animal Hospital Association 169

170 JOURNAL of the American Animal Hospital Association
normal temperature, pulse, and respiratory rate. The dog
stood with her back arched and seemed to experience
pain on palpation of the lumbar spine. On abdominal
palpation, the colon was full of soft feces, and rectal
examination was normal.
A complete blood count (CBC) showed a mild throm-
bocytopenia (platelet count, 180 x103/µl; reference range,
200 to 500 x103/µl) but was otherwise unremarkable.
Serum chemistry analysis was within normal limits. Ab-
dominal radiographs showed extensive ventral spondy-
losis throughout the caudal thoracic and lumbar spine.
The descending colon was distended with fecal material
and appeared to narrow cranial to the pubis.
The dog was treated with methocarbamol a (15 mg/kg
[6.8 mg/lb] body weight, per os [PO] q 8 hrs) for spinal
pain and was placed on a commercial bland diet,b and the
owners were instructed to encourage oral fluid intake.
Twenty-four hours later, the dog presented to the refer-
ring veterinarian for reevaluation. She continued to
vomit, act painful and restless, and strained to pass small
amounts of stool. A fecal flotation was negative for
parasitic ova, and fecal cytology was normal. Abdominal
radiographs were repeated and appeared similar to those
obtained previously. Fluid deficits were corrected by
intravenous (IV) administration of lactated Ringer’s so-
lution, and the dog was referred to the University of
Illinois, Veterinary Medicine Teaching Hospital for fur-
ther evaluation.
On presentation, the patient appeared thin and de-
pressed. Temperature, pulse, and respiratory rate were
within normal limits. Neurological examination was nor-
mal. The dog experienced pain during abdominal palpa-
tion, which revealed distention of the colon with feces.
Rectal examination was normal. The dog strained ac-
tively when attempting to defecate, but only passed small
amounts of soft stool. Warm water enemas were adminis-
tered to relieve the obstipation, and lactated Ringer’s solu-
tion was administered intravenously to maintain hydration.
Thoracic and abdominal radiographs were performed.
Thoracic films revealed mild spondylosis of the midthoracic
vertebrae and a moderate bronchointerstitial pattern in
all lung fields. Abdominal films showed extensive bridg-
ing spondylosis of the caudal thoracic vertebrae, cranial
lumbar vertebrae, and lumbosacral junction. There was a
moderate amount of gas in the gastrointestinal tract, with
gas distention of the colon. The descending colon had an
irregular shape, with an apparent narrowing of the lumen
and thickening of the colonic wall 5 cm cranial to the
pelvic brim.
Abdominal ultrasonography demonstrated mild peri-
toneal effusion. A partial obstruction of the descending
colon with focal thickening of the bowel wall was noted.
An enlarged sublumbar lymph node, measuring 8 mm,
also was identified. Differential diagnoses included co-
lonic neoplasia, fungal infection with secondary stric-
ture, and extraluminal stricture of the descending colon.
March/April 1999, Vol. 35
A ventral midline laparotomy was performed. At sur-
gery, extraluminal compression of the descending colon
by a band of fibrous tissue was found. The scar tissue
extended from the caudal pole of the left kidney across
the descending colon to the right dorsal body wall. Rem-
nants of the right broad ligament as well as the uterine
stump were adhered to the fibrous band which trapped
the colon against the lumbar musculature, but was not
adhered to the colon directly. The sublumbar and mesen-
teric lymph nodes were enlarged slightly, and the co-
lonic and regional serosa were hyperemic. The fibrous,
constricting band was broken down by blunt and sharp
dissection and removed. This freed the colon from its
entrapped ventral spinal position. The area was lavaged
with physiologic saline, and normal peristaltic move-
ment of the colon was observed.
An approximately 6-cm length of the colon in the
entrapped region was thickened and hyperemic. A full-
thickness longitudinal wedge biopsy of the colon was
taken, and the incision was closed transversely in a two-
layer inverting pattern with 3-0 polydioxanone sutures.c
A wedge biopsy of the sublumbar lymph node also was
taken. The abdomen was lavaged with physiologic saline
and closed routinely in three layers.
The dog had a normal postoperative recovery and
passed feces within four hours of the completion of
surgery. By 48 hours postoperatively, the dog was bright
and alert, eating readily, and passing formed stools with-
out straining. Histopathological examination of the co-
lonic biopsy demonstrated diffuse, chronic-active
inflammation of the serosal surface and evidence of local
peritonitis. Minimal inflammatory changes were ob-
served in the mucosal and submucosal layers. The lymph
node biopsy revealed marked lymphoid hyperplasia. No
evidence of neoplasia or an etiological agent was ob-
served.
Ten months postoperatively, the owner reported that
the dog has had no recurrence of the presenting symp-
toms. The referring veterinarian reported that the dog
was spayed approximately 27 months prior to presenta-
tion, and the ovarian and uterine ligatures were made
with 2-0 chromic catgut. No abnormalities were noted
during the OVH, and no immediate complications from
the surgery were encountered.
Case No. 2
A two-year-old, spayed female domestic shorthair cat
was presented to the University of Illinois, Veterinary
Medicine Teaching Hospital for evaluation of depres-
sion, anorexia, vomiting, and passage of small-diameter
stools. The cat had undergone an elective OVH six weeks
prior to admission. She had been clinically normal until
one week before presentation.
The cat’s body temperature (103.1˚ F) and heart rate
(240 beats per min) were elevated on presenting physical
examination. The animal was approximately 5% dehy-
March/April 1999, Vol. 35
Figure 1—Lateral abdominal radiograph of a two-year-old
domestic shorthair. Note the distention of the colon with fecal
material and the apparent stricture cranial to the pelvic brim,
which was seen on repeated radiographs. The gas in the
intrapelvic colon is attributed to a recent enema.
drated. Abdominal palpation revealed impaction of the
colon with firm feces. Abdominal radiographs demon-
strated marked obstipation of the colon and an apparent
narrowing of the lumen cranial to the pelvic inlet [Figure
1]. No other abnormalities of the abdomen or pelvis were
noted. Differential diagnoses included idiopathic mega-
colon, colonic neoplasia, fungal infection with second-
ary stricture, and extraluminal stricture of the descending
colon.
Blood was drawn for a CBC, serum chemistry profile,
and retrovirus testing. The CBC was within normal lim-
its. Serum chemistry revealed mild hyperprotein-
emia (7.9 g/dl; reference range, 5.8 to 7.8 g/dl) and
hyperalbuminemia (4.3 g/dl; reference range, 2.7 to 3.8
g/dl). Serological examination for feline leukemia virus
and feline immunodeficiency virus was negative.
Colonoscopy was performed with a rigid proctoscoped
and demonstrated a stricture of the descending colon
approximately 10 cm cranial to the anus. The colonic
mucosa appeared hyperemic at the level of the stricture,
and the proctoscope could not be passed further crani-
ally. Colonic mucosal biopsies were obtained and showed
a mild, chronic, nonsuppurative colitis.
The cat was treated with warm-water enemas every
eight hours; lactulosee (1 ml PO q 8 hrs); amoxicillinf (22
mg/kg [10 mg/lb] body weight, PO q 12 hrs); cimetidineg
(10 mg/kg [4.5 mg/lb] body weight, PO q 8 hrs); and
lactated Ringer’s solution (50 ml subcutaneously q 12
hrs). Medical management failed to result in clinical
improvement, and surgical exploration was recom-
mended.
A ventral midline laparotomy was performed and
showed the descending colon to be constricted by a ring
of connective tissue that originated from the uterine
stump and adhered dorsally to the mesocolon bilaterally.
The colon itself was not adhered to the scar tissue, but
instead passed through the narrow, fibrous ring. The
Colonic Obstruction 171
colon and the regional peritoneal surfaces were hyper-
emic. Suture material was not observed in the stricture or
the adhesion. The fibrous bands were dissected free and
excised, relieving the stricture. During dissection, a small
rent was created inadvertently in the mesocolon and
subsequently closed. The abdomen was lavaged with
physiologic saline and then closed routinely in three
layers.
Within 24 hours postoperatively, the cat passed a
large amount of feces. A significant improvement was
noted in the cat’s attitude and appetite, and the cat was
discharged from the hospital 48 hours postoperatively.
The amoxicillin and cimetidine were continued for seven
days postoperatively.
Nine months after discharge, the owners reported that
the cat had no recurrence of the presenting symptoms or
complications related to surgery. The referring veteri-
narian reported that the cat was spayed six weeks prior to
presentation and that no surgical complications were
encountered. The ovarian and uterine ligatures had been
performed with 3-0 polyglactin 910 sutures.h
Case No. 3
A 10-year-old, spayed female Scottish terrier underwent
a laparotomy at the University of Illinois Veterinary
Medicine Teaching Hospital for surgical treatment of a
urinary bladder tumor. At surgery, a band of fibrous
connective tissue was noted compressing the colon to the
dorsal body wall. The scar tissue was attached to the
uterine stump ventrally and the region of the broad liga-
ments dorsolaterally. The colon was not adhered to the
fibrous band, but the lumen was compressed dorsoven-
trally by at least 50%. The cicatrix held the colon against
the lumbar musculature and severely limited the organ’s
mobility. The band of scar tissue was broken down by a
combination of blunt and sharp dissection and removed.
This allowed normal mobility of the descending colon. A
partial cystectomy then was performed to remove the
apical one-third of the urinary bladder for treatment of a
transitional cell carcinoma. The dog recovered from sur-
gery without complications.
The owner reported that the dog was spayed approxi-
mately nine years prior to presentation. Details of the
operation, including the suture material utilized, were
not available. The owner also stated that the dog had
never shown any clinical signs of tenesmus, obstipation,
or abnormally shaped stools.
Discussion
Joshua first reported on bowel obstruction in dogs as a
complication of OVH. 4 She referred to incarceration of
bowel, requiring resection and anastomosis, secondary
to adhesions within granulomas that formed around non-
absorbable suture material. Subsequently, other authors
have reported on bowel obstruction secondary to organ
entrapment in an extensive granulomatous reaction in-
172 JOURNAL of the American Animal Hospital Association
volving the ovarian pedicles or uterine stump.5,6,8 In all
cases, the origin of the granuloma was thought to be a
reaction to contaminated, multifilament, nonabsorbable
suture material. The interval between surgery and the
detection of the obstructing granuloma has been reported
to be between several months and several years.5,6,8
Unlike the previous reports of dogs with bowel ob-
struction, neither of the dogs in this report had a granulo-
matous lesion. Instead, surgical findings revealed only
organized fibrous tissue that was not adhered to the
bowel wall directly. In addition, there was no evidence
of residual suture material in these cases. The boxer
developed clinical signs of colonic obstruction more than
two years after being spayed. The Scottish terrier did not
have clinical signs, and the extramural scar tissue was
found incidentally during a laparotomy nine years after
OVH.
To the authors’ knowledge, partial colonic obstruc-
tion secondary to extraluminal scar tissue without sero-
sal adhesions has not been reported previously in the dog
as a complication of OVH. However, Kunin reported a
case of urinary incontinence in a two-year-old, mixed-
breed dog that occurred secondary to an encircling band
of scar tissue originating from the uterine stump.12 The
fibrous band wrapped around the bladder, limiting full
distention and resulted in a bizarre, cranial diverticulum.
Similar to the dogs in this report, the fibrous tissue was
not adhered to the organ directly and was removed easily
at surgery. The dog recovered uneventfully.12
Obstipation as a complication of OVH has been de-
scribed previously in three cats.9–11 In the reported cases,
the complication was diagnosed six, 10, and 24 weeks
following surgery. Similar to the cat in the authors’
report, the obstructions were caused by fibrous connec-
tive tissue at the uterine stump that encircled but was not
adhered to the descending colon directly.9–11 In two of
the prior cases, short sections of the uterine horns had
been left proximal to the uterine body, and the encircling
stricture originated from them.10,11 These two cases re-
quired additional resection of the uterus, to the level of
the cervix, after the fibrous adhesions were broken down.
In the authors’ case, additional uterine tissue was not
removed at surgery.
In all previously reported feline cases, as well as the
cat in this report, clinical signs were first observed within
10 weeks of OVH. The encircling scar tissue that re-
sulted in obstipation and partial colonic obstruction ap-
parently was due to excessive fibroplasia during the
healing period. It is harder to explain the delayed onset
of clinical signs in the boxer. This dog probably had
transcolonic scar tissue that formed following OVH,
which had been present without causing signs for two
years. The acute onset of abdominal pain and obstipation
in this patient may be due to twisting of the colon within
the scar tissue or an increased degree of adhesions across
the colon. There was no histological evidence of colitis
March/April 1999, Vol. 35
or mucosal inflammation. The third case illustrates that
extramural fibrous tissue can be present, causing signifi-
cant colonic compression, presumably for several years,
without causing clinical signs.
Conclusion
Fibrous adhesions that cause partial colonic obstruction
are a rare but significant delayed complication of OVH
in the dog and cat. The exact cause of the adhesions is
unknown, but they may lead to clinical signs of tenes-
mus, obstipation, anorexia, vomiting, and abdominal
pain. Treatment of the condition involves surgical re-
moval of the offending scar tissue. Long-term resolution
of the presenting clinical signs was seen in all cases.
a
Robaxin; Fort Dodge Laboratories, Fort Dodge, IA
b
i/d; Hill’s, Inc., Topeka, KS
c
PDS II; Ethicon, Inc., Somerville, NJ
d
Fiber optic sigmoidoscope; Welch Allyn, Inc., Skaneateles Falls, NY
e
Lactulose; Schein Pharmaceutical, Flocham Park, NJ
f
Amoxi-tabs; SmithKline Beecham, Westchester, PA
g
Cimetidine; Novapharm, Inc., Schaumburg, IL
h
Vicryl; Ethicon, Inc., Somerville, NJ
References
1. Pollari FL, Bonnett BN, Barnsey SC, et al. Postoperative complications of
elective surgeries in dogs and cats determined by examining electronic and
paper medical records. J Am Vet Med Assoc 1996;208(11):1882–6.
2. Berzon JL. Complications of elective ovariohysterectomies in the dog and
cat at a teaching institution: clinical review of 853 cases. Vet Surg
1979;8(3):89–91.
3. Pollari FL, Bonnett BN. Evaluation of postoperative complications
following elective surgeries of dogs and cats at private practices using
computer records. Can Vet J 1996;37:672–8.
4. Joshua JO. The spaying of bitches. Vet Rec 1965;77(23):642–7.
5. Pearson H. Ovario-hysterectomy in the bitch. Vet Rec 1970;87:646–7.
6. Pearson H. The complications of ovariohysterectomy in the bitch. J Sm
Anim Pract 1973;14:257–66.
7. Dorn AS, Swist RA. Complications of canine ovariohysterectomy. J Am
Anim Hosp Assoc 1977;13:720–4.
8. Spackman CJA, Caywood DD, Johnston GR, et al. Granulomas of the
uterine and ovarian stumps: a case report. J Am Anim Hosp Assoc
1984;20:449–53.
9. Furneaux RW, Baysen BG, Mero KN. Complications of ovariohysterecto-
mies. Can Vet J 1973;14(4):98–9.
10. Muir P, Goldsmid SE, Bellenger CR. Megacolon in a cat following
ovariohysterectomy. Vet Rec 1991;129:512–3.
11. Smith MC, Davies NL. Obstipation following ovariohysterectomy in a cat.
Vet Rec 1996;138:163.
12. Kunin S, Terry M. A complication following ovariohysterectomy in a dog.
Vet Med 1980;75(6):1000–1.
13. McEvoy FJ. Iatrogenic renal obstruction in a dog. Vet Rec 1994;
135:457–8.