ORIGINAL RESEARCH

Veterinary Surgery
23:475-487, 1994

Incomplete Ossification of the Humeral Condyle in Spaniels

DENIS J. MARCELLIN-LITTLE, DrMedVet, DAVID J . DEYOUNG, DVM, Diplomate ACVS, Diplomate ACVA,
KELLl K. FERRIS, BS, and CLIFFORD M. BERRY, DVM, Diplomate ACVR

An evaluation of 157 dogs with humeral fractures was performed. Cocker spaniels were more
likely to have humeral condylar fractures (HCFs) than other breeds ( P < .0001). Male cocker
spaniels were at increased risk ( P < .00 I ). Cocker spaniels had more bilateral HCFs than other
breeds of dogs ( P < .OOI). Eighteen dogs (17 purebred spaniels and 1 crossbred spaniel) with
HCFs of unknown cause or occurring with normal activity were further studied, using radiography
of their humeral condyle bilaterally (n = 18), computed tomography (n = 3), biopsy (n = 2),
bone scintigraphy (n = 2), and genetic evaluation (n = 8). Fourteen of these 18 dogs had a
nonfractured contralateral condyle. Twelve (86%)of the 14 nonfractured humeral condyles had
a radiolucent line within the center of the condyle, 13 (93%)had radiographic signs of degenerative
joint disease and an abnormal medial coronoid process, and six (43%)had periosteal proliferation
involving the lateral epicondyle. Examination of biopsy samples from the fracture sites of two
cocker spaniels showed fibrous tissue present at the fracture surfaces. The results of this study
suggest an association between incomplete ossification of the humeral condyle in cocker spaniels
and Brittany spaniels and a high prevalence of HCFs. Eight affected cocker spaniels with available
pedigree information were found to be genetically related, suggesting that incomplete ossification
of the humeral condyle may be a genetic disease with a recessive mode of inheritance.
OCopjvYglif I994 by Thc. American COIILJ~~P of' rf'rtcw'narySzirgeons

H UMERAL CONDYLAR fractures (HCFs) are

common in dogs.'-4 Major trauma, such as
motor vehicle accidents, is the leading cause of HCF
or intense
posing factors.
have been suggested as predis-

Skeletally mature dogs that had humeral fractures

in dogs. HCFs can occur in association with normal and were brought to our hospital between 1983 and
activities such as running, climbing stairs, or jumping 1993 were evaluated to determine the risk factors
from a height of one meter or less. Compared with associated with these injuries. Based on the breed
other breeds, spaniels have a higher incidence of distribution of these animal patients, we hypothe-
HCF occurring with such a c t i ~ i t i e s . ~The
,~'~
history, sized that a heritable condition resulting in incom-
signalment, and fracture type suggest that a patho- plete ossification of the distal humeral condyle pre-
logical process may predispose spaniels to HCF. The disposes several spaniel breeds to HCF. To test this
pathogenesis of these fractures is unclear, however, hypothesis, further evaluation using radiology, com-
metabolic diseases,6 conformation of the e l b ~ w , ~ .puted~ tomography, nuclear scintigraphy, and his-

From the Department of Companion Animal and Special Species Medicine. the Department of Anatomy/Physiology/Radiology,
College of Veterinary Medicine. and the Department of Animal Science. North Carolina State University. Raleigh, NC.
Funded by the College of Veterinary Medicine, and the Orthopedic Research Group, North Carolina State University, Raleigh,
NC.
Address reprint requests to David 3 . DeYoung, DVM, Diplomate ACVS, Diplomate ACVA, Department of Companion Animal
and Special Species Medicine. College of Veterinary Medicine. North Carolina State University. Raleigh, NC 27606.
OCopyright 1994 by The American College of Veterinary Surgeons
0 16 1-349919412306-0006$3.00/0

47 5
476 INCOMPLETE OSSIFICATION O F THE HUMERAL CONDYLE
tology of fractured and intact humeri of 17 skeletally
mature purebred spaniels and a crossbred spaniel
with HCF was performed. To test the heritability
and mode of inheritance, 26 generation pedigrees of
skeletally mature cocker spaniels with an HCF were
compiled from the American Kennel Club registry
and familial analysis was performed.
MATERIALS AND METHODS
The medical records from 157 dogs who were pre-
sented to the North Carolina State University Veteri-
nary Teaching Hospital (NCSU-VTH) between 1983
and 1993 for treatment of a humeral fracture were re-
viewed. Signalment, medical history, physical exami-
nation findings, diagnosis, and treatment data were re-
corded for each dog. Standard orthogonal radiographic
projections that had been made of each dog's humeral
fracture were evaluated.
Twenty-eight of the 157 dogs were skeletally mature
spaniels with an HCF of unknown origin or occurring
during normal activity. These 28 spaniels (27 purebred
spaniels and one crossbred spaniel) were evaluated as three
different study groups.
Group I included all 28 spaniels. Thirty-five fractures
(21 unilateral and seven bilateral) were present in these
dogs. Preoperative radiographs of 32 of the 35 fractures
in this group were available for evaluation.
Group 2 consisted of 24 purebred cocker spaniels.
Prevalence of unilateral and bilateral fractures, sex, age,
and failure rate were compared between group 2 and
the other skeletally mature dogs with HCF treated at
the NCSU-VTH. The weight of 20 male and four female
nonaffected cocker spaniels, randomly chosen from
within the hospital population, was recorded and com-
pared with the weight of the dogs in this group. Twenty-
nine fractures ( 19 unilateral and five bilateral) were
present in these dogs.
Group 3 included 18 spaniels (17 purebred spaniels
and 1 crossbred spaniel) that were selected for in-depth
follow-up evaluation.
Radiography
In all dogs in group 3 radiographs of the fractured con-
dyles and the contralateral condyles were reviewed for the
presence of a fragmented medial coronoid process, un-
united anconeal process, osteochondrosis of the humeral
condyle, the presence of ulnar subtrochlear sclerosis, and
radial, ulnar, or humeral osteophytes." Also, the radio-
graphs were evaluated for the presence of periosteal pro-
liferation in the lateral epicondylar crest.
The following procedures were performed on selected
dogs from group 3 when circumstances allowed.
Biopsy and Histologicul Examination
Biopsy samples were obtained from two dogs during
surgical repair of the HCF using a 4.1-mm diameter
hollow crown drill bit (Howmedica, Inc., Rutherford,
NJ) to harvest a 3.2-mm diameter core biopsy specimen
of the lateral fracture fragment. The biopsy site. en-
larged to 4.5 mm, was used as a glide hole for lag screw
fixation. lnterfragmentary compression was achieved
with a fully threaded 4.5-mm cortical bone screw
(Synthes, Inc., Paoli, PA). The bone biopsy specimens
were demineralized in ethylenediamine tetraacetate
(EDTA), embedded in paraffin, sectioned, and stained
with hematoxylin-eosin.
Comput t.d Tomography
Three dogs underwent computed tomography (Model
0450, Pfizer, New York, NY) of both elbow joints. The
dogs were positioned in ventral recumbency with the
forelimbs extended. Contiguous 2-mm transverse images
of the elbow joint were recorded. Reconstructed images
in dorsal and sagittal planes were evaluated.
Bone Scint igruphjj
Bone scintigraphy was performed on two dogs by in-
travenous injection of 370 to 740 megabequerels (10 to
20 mCi) of 99"technetium-methylene diphosphonate
(osteolite, methylene diphosphonate, E. I. du Pont de
Nemours & Co., Billerica, MA). Static 128 X 128 X 16
pixel images were obtained 2 hours after injection using
a large field of view gamma camera (ADAC ARC 3000
and DPS 3300, ADAC Laboratories, Milpitas, CA) with
a parallel hole general purpose collimator. Images were
subjectively evaluated for areas of increased uptake of
the radiotracer.
Pedigree Study
Using the American Kennel Club registry. 26 genera-
tion pedigrees were compiled for eight of the cocker span-
iels in group 2. The coefficient of inbreeding (F,) of af-
fected dogs was determined using the equation:
Fx = z[(l/2)n1tn2t
( 1 + FA)]
where FA was the inbreeding coefficient of a common
ancestor, and n , and n2 were the numbers of generations
between that ancestor and the sire and the dam. respec-
tively." The coefficient of inbreeding of 100 cocker span-
 M A R C E L L I N - L I T T L E ET A L 477

Table I . Clinical Data in ZX Skeletall) Mature Dogs (27 Purebred Spaniels and a Crossbred Spaniel)
With 35 HCFs Resulting From a Minor Trauma

Dog Agc Weight Surgical Complications/

No. (yrs) Sex (kg) Color Causc o f Fracture Outcome Concurrent Medical Problems

I* 8 15.5 Buff Climbing stairs

2* 8 12.3 Buff Running Diskospondylitis.
fibrocartilaginous embolism
3* 4 15.5 Bufl' Dog fight Urinary tract infection
4* 5 13.8 Buff Chasing ball Right-sided cardiomegaly,
babesiosis
5* 2 13.1 Buff No known trauma Nonunion of the olecranon. Heartworm disease
osteom ye11tis
6* 10 16.5 Buff Dog fight Exuberant callus -
7* 8 14.5 Buff Jumping doun - Heartworm disease
8* 7 18.0 Buff Walking - -

9* 7 13.0 Black No known trauma - -

lot 9 22.7 wsto No known irmma Failure of fixation -

I I* 4 13.6 Buff Jumping down Three failures of fixation

( 2 left side, I right side)
I2* 7 12.7 Black Climhing stairs - -
I3* 7 11.8 Buff Jumping down Failure of fixation -

14* 6 12.5 Black No known lrauma - Hip dysplasia. cruciate ligament

injury
I5* 5 21.8 Black Running - -

16* 7 21.8 Buff Jumping dow,n Three failures of fixation Radius curvus. thrombocytopenia
(same side)
17% 4 12.7 Buff Caught leg in cage door LumbO-SdCra\ disease
I8* 17.2 Buff Going down stairs - -
6
19* 3 11.4 Buff No known trauma Pin tract osteomyelitis -

20* 3 14.9 Black Going down stairs - H ypot hyroidism

21* 6 18.4 Buff Chasing ball -

22* 6 16.0 Black Climbing stairs - Lateral patellar luxation

23* 10 18.2 Buff N o known trauma Two failures of fixation Radius curvus, carpal laxity
(same side)
24* 5 12.3 Buff Dog fight -

25' 5 13.0 Black Jumping down Radius curvus

26t 10 17.7 W&O Jumping down -

27 t I2 17.5 W&M Digging - -

284 6 13.0 Black Chasing ball Delayed union, fixation Radius curvus, medial patellar
fai I u re luxation

5.9 * 2.09 14.9 -+ 2.95

Abbreviations: M, male: M/C. male castrated: F/S. female spayed: W&O, white and orange; W&M. white and mahogany.
* Cocker spaniel.
7 Brittany spaniel.
$ Mixed breed dog: sire. standard poodle: dam. cocker spaniel.
5 Mean and standard deviation. cocker spaniels.

iels without medical history of HCFs registered at the
American Kennel Club between I978 and 1988 was col-
lected as a reference. Ancestors common to the eight af-
fected cocker spaniels and absent in the lineage of the 100
cocker spaniels were identified.
St at istica 1 A na Iysis
Using 2-tailed Fisher's exact tests, the number of
unilateral and bilateral HCFs in cocker spaniels in group
2 were compared with the number of HCFs occurring
in the general population presented to the NCSU-VTH
(Statistical Analysis System [SAS], version 6.07.02, SAS
Institute Inc., Cary, NC). Also, the sex of affected cocker
spaniels was compared with the sex of nonaffected
cocker spaniels presented to the NCSU-VTH during
the study period. Using 1-tailed Fisher's exact test, the
failure rate after internal fixation for repair of an HCF
was compared between skeletally mature cocker span-
478 INCOMPLETE OSSIFICATION OF THE HUMERAL CONDYLE

Table 2. Additional Data in 28 Skeletally Mature Dogs (27 Purebred Spaniels and a Crossbred Spaniel)
With 35 Humeral Condylar Fractures Resulting From a Minor Trauma
~ ~~

Contralateral Condyle-
Affected Condyle Contralateral Condyle-Intact Fractured

Dog Fracture Fracture

No. Type FCP DJD PP Line FCP DJD PP Type FCP DJD PP Further Examination

I* Medial - - - - - - -
2* Y §Yes Yes No - - - -
3* Y Yes Yes No - - - -

4* Y §Yes No No - - - -
5* Y $Yes Yes Yes - - - - -
6* Y No No No - - - - F, .32%
7* Y §Yes No No - - - -
8* Y §Yes Yes No - - - -
9* Lateral §Yes No No - - - - -
lot Lateral - - - - - - - F, = .OO%
I I* Lateral Yes No No - - - - Y Yes No No F, = ,249~
12* Y §Yes Yes No CRL §Yes Yes Yes -
13* Lateral $Yes Yes Yes CRL Yes Yes Yes -
14* Medial Yes No Yes - - - - Medial Yes Yes No Bonescan
15* Y $Yes No No CRL Yes No Yes CT/Bone scan, F, = 2.22%
16* Lateral No Yes No CRL Yes Yes Yes Lateral Yes Yes Yes F, = 25.06%
17* Medial §Yes Yes No CRL $Yes Yes Yes F, = 13.36%
18* Y §Yes Yes No CRL §Yes Yes No F, = 25.17'70
19* Y $Yes No No - - - - Lateral No Yes No -
201 Y $Yes Yes No No Yes Yes No
21* Lateral $Yes Yes No No Yes Yes No -
22* Y §Yes Yes Yes PRL Yes Yes No Bone biopsy. F, = 3.91%
23* Lateral Yes No No PRL Yes No No Lateral Yes Yes No CT scan. Bone biopsy
24* Lateral Yes No No CR L Yes No No F, = 2.34%
25* Medial Yes No No PRL Yes No No -
26t Y §Yes No No - - - - Y §Yes No No F, = 6.25%
27t Lateral No Yes No PRL No Yes No Y $Yes Yes No CT scan
28$ Y - - - CRL Yes Yes Yes

Abbreviations: FCP, radiographic changes consistent with an abnormal medial coronoid process: DJD. degenerative joint disease: PP, periosteal
proliferation; Line, radiolucent line: §Yes. separate fragmented medial coronoid process: --% not available; Y, intercondylar fracture: Medial. fracture
of the medial portion of the humeral condyle: Lateral. fracture of the lateral portion of the humeral condyle: CRL. complete radiolucent line: PRL,
partial radiolucent line: F,. coefficient of inbreeding: CT scan, computed tomography scan.
* Cocker spaniel.
t Brittany spaniel.
$ Mixed breed dog: sire, standard poodle; dam. cocker spaniel.

iels and skeletally mature dogs from all other breeds.
Two-tailed Student's t tests were used to compare the
age of affected cocker spaniels with the age of the general
population with an HCF and to compare the weight of
affected cocker spaniels with the weight of 24 control
cocker spaniels chosen randomly from within the hos-
pital population. P values < .05 were considered sig-
nificant.
RESULTS
C'linical Data
Of the I57 humeral fractures reviewed, 85 (54%)
were HCFs. Fifty-five of the HCFs occurred in skel-
etally mature dogs. Thirty-five (64%)of the 5 5 HCFs
in skeletally mature dogs occurred with normal ac-
tivity, 1 1 (20%) fractures resulted from major
trauma, eight (14%) fractures were of unknown
cause, and one (2%) fracture was secondary to a fi-
broblastic osteosarcoma.
The 43 fractures (seven bilateral) occurring with
normal activity or from an unknown cause occurred
in 36 dogs. The breed distribution was 24 cocker
spaniels ( I 9 unilateral and five bilateral HCFs), three
Brittany spaniels (one unilateral and two bilateral
HCFs), two miniature pinschers, one Pekingese, one
Maltese, one Chihuahua, a mixed breed dog from a
 MARCELLIN-LITTLE E T AL
standard poodle (sire) and a cocker spaniel (dam),
and three mixed breed dogs of unknown lineage.
The cocker spaniels had a significantly higher prev-
alence of HCF than other breeds ( P < .OOO 1).
Group 1 included 24 cocker spaniels. three Brit-
tany spaniels. and one crossbred spaniel with 35
HCFs (28 unilateral and seven bilateral). Data from
this group is presented in Tables 1 and 2. Eighteen
of 35 HCFs ( 5 1%) were intercondylar fractures'.'
(also known as "Y" or "T" fractures' and
dicondylar'.'.' fractures) with separation of the me-
dial and lateral portions of the humeral condyle from
the humeral shaft (Table 2). Twelve of the 35 HCFs
(34%)were fractures of the lateral portion of the hu-
meral condyle. with separation of the lateral portion
ofthe humeral condyle from the humeral shaft. Five
of 35 HCFs (1470)were fractures of the medial por-
tion of the humeral condyle, with separation of the
medial portion of the humeral condyle from the hu-
meral shaft. The preoperative radiographs of 32
HCFs were available for review. A separate frag-
mented medial coronoid process was observed ra-
diographically in 18 of 32 HCFs (56%) and radio-
graphic signs suggestive of elbow degenerative joint
disease and an abnormally shaped medial coronoid
process were present in 28 joints with HCFs (88%;
Table 3 ) . Periosteal proliferation was present on the
lateral epicondylar crest of five fractured condyles
( 16%). Seven dogs in this group, five cocker spaniels
and two Brittany spaniels, had bilateral HCFs (25%).
Three dogs presented initially with bilateral HCFs.
The four other dogs sustained fractures on the con-
tralateral side 2 weeks (dog no. 11). 3 months (dog
no. 27). 6 months (dog no. 23), and 3 1 months (dog
no. 16) after the first HCF.
Group 2 consisted of 24 purebred cocker spaniels
with 29 HCFs (19 unilateral and five bilateral).
Twenty (83%) of the 24 cocker spaniels were males
Table 3 . Analksis of Radiographic Data in 32 HCFs Resulting From Minor Trauma in 28 Skeletally Mature Dogs
No. FCP
Fractured condyles 32 28/32. ( 8 W )
Intact cond\les 14 13/14 (93'7)
Condkles with CRL or PRL 12 11/12 (92%)
Condyles with CRL 8 8/X (l00"i)
Condyles with PRL 4 314 ( 7 W )
Abbreviations: FCP. radiographic changes consistent with an abnormal medial coronoid process; SFCP. separate fragmented medial coronoid
process; DJD. degenerative joint disease; PP. penosteal proliferation: CRL, complete radiolucent line; PRL, partial radiolucent line.
479
and four ( 17%)were females. The prevalence of HCF
was significantly higher in males than in females
when compared with the male/female ratio of the
population of cocker spaniels presented to the
NCSU-VTH during the study period ( P < .OOl).
Dogs in group 2 were significantly older (median
age, 6 years old) than skeletally mature dogs of other
breeds with HCF (median age, 3 years old; P < .O I).
The mean weight of dogs in group 2 (14.9 kg +- 2.9)
did not differ significantly from the mean weight of
the control cocker spaniels (14.4 kg ? 2.6). Five dogs
in this group had bilateral HCF (21%). The preva-
lence of bilateral HCF was significantly higher in
cocker spaniels than in all other canine breeds ( P
< .001). Twenty-seven of the 29 HCFs in this group
were surgically reduced and stabilized using internal
fixation. Follow-up radiographs were available for
26 fractures. Implant failure and loss of fracture re-
duction occurred in six of 26 fractures repaired
(23%). Radiographic follow-up after surgical repair
was available in 22 HCFs occurring in skeletally ma-
ture dogs of nonspaniel breeds. Implant failure and
loss of fracture reduction occurred in one of these
22 fractures (4.6%). The failure rate in group 2
tended to be higher than the failure rate in skeletally
mature dogs from all other breeds after repair of
HCF ( P = .078).
Group 3 included 18 spaniels: 15 cocker spaniels
(dogs no. 1 1 to no. 25; Table I), two Brittany spaniels
(dogs no. 26 and no. 27; Table I), and a crossbred
spaniel (dog no. 28; Table 1). Bilateral radiographs
of the elbow joints were available for all dogs in this
group.
Radiography
The contralateral condyles were intact in 14 of 18
dogs in this group. Vertical radiolucent lines (0.5
SFCP DJD PP
18/32 (56%) 18/32 (56%) 5/32 ( 16'7)
3/14 (21%) 10/14 (71%) 6 / I4 (43%)
3/12 (25%) 8/12 (67%) 6/12 (50%)
3/8 (38%) 618 (75%) 618 (75%)
014 (0%) 2/4 (50%) 0/4 (0%)
480 INCOMPLETE OSSIFICATION OF THE HUMERAL CONDYLE

mm to 1 .O mm wide) were identified on the cranio-

caudal radiographic projection in 12 (86%)of the 14
intact condyles, and extended proximally from the
articular surface of the trochlea. In the 12 dogs with
a radiolucent line in their intact humeral condyle
the following findings were observed radiographi-
cally: partial radiolucent lines extending proximally
from the articular surface without reaching the phy-
seal scar (Fig 1) in four dogs (33%); complete radio-
lucent lines extending to the physeal scar or the su-
pratrochlear foramen (Figs 2 to 4) in eight dogs
(67%); a separate fragmented medial coronoid pro-
cess in three dogs (25%); radiographic features of an
abnormally shaped medial coronoid process (Fig 3;
Table 3) in 1 1 dogs (92%); degenerative joint disease

Fig 2. Cranio-caudal radiograph of the humeral condyle of dog

no. 18 (6-year-old male cocker spaniel). A radiolucent line ex-
tends from the articular surface proximally (arrow). The position
of the incomplete ossification of the distal humeral condyle is in
the anatomic location where normal fusion between the medial
and lateral humeral centers of ossification occurs.

of the elbow joint (Fig 4) in eight dogs (67%); and

Fig 1. Cranio-caudal radiograph of the humeral condyle of dog
no. 22 (6-year-old spayed female cocker spaniel). A radiolucent
line extends from the articular surface proximally (arrow).
periosteal proliferation on the lateral epicondylar
crest (Fig 3 ) in six dogs (50%).Two dogs in group 3
did not have a radiolucent line in the contralateral
humeral condyle but radiographic features consistent
with a fragmented medial coronoid process were
present. Seven dogs in group 3 had bilateral humeral
fractures. The humeral condyles of three of these
dogs had been evaluated radiographically before the
occurrence of the fracture of the contralateral con-
dyle. Two of these dogs had a partial radiolucent
line in the intact condyles (dogs no. 23 and no- 27)-
and one dog had a complete radiolucent line (dog
 MARCELLIN-LITTLE E T AL 48 I

Fig 3. Cranio-caudal radiograph of the humeral condyle (A) and illustration (B) of dog no. 17 (4-year-old castrated male cocker
spaniel). A radiolucent line (a) extends from the articular surface proximally. A fragmented medial coronoid process (b) is present.
A smooth, chronic periosteal proliferation ( c ) is observed along the lateral epicondylar crest.

no. 16). Four spaniels (dogs no. 16, no. 23, no. 25,
and no. 28; 14%)from group 3 had bilateral radius
curvus.
Computed Tomography
Computed tomography was used to evaluate the
elbows of two dogs with a partial radiolucent line
(dog no. 23 and no. 27; Fig 5) in group 3. A linear
area of hypoattenuation had a width, position, and
orientation similar to the radiolucent line, the area
started caudally and extended cranially for one third
to one half of the diameter of the condyle (Fig 6B).
Computed tomography was performed on one dog
with a complete radiolucent line (dog no. 15; Fig
6C) in group 3. A planar area of decreased atten-
uation was present perpendicular to the long axis of
the trochlea. Increased density was observed adjacent
to the radiolucent line on all three computed to-
mography scans.
Bone Scintigruphy
Bone scintigraphy was performed on two dogs in
group 3. One dog had increased radiotracer uptake
in both distal humeri. Radiographically, this dog had
a nonunion of the fractured medial portion of the
humeral condyle bilaterally. Bone scintigraphy was
482 INCOMPLETE OSSIFICATION OF THE HUMERAL CONDYLE

Fig 4. Cranio-caudal radiograph of the humeral condyle (A) an1 illustration (B) of dog no. 28 (6-year-old castrated mixei bree
dog [sire, standard poodle; dam, cocker spaniel]). A radiolucent line (a) extends from the articular surface proximally. Also, signs
of advanced degenerative joint disease are present near the medial coronoid process (b) and the epicondyles (c).

performed on the second dog 5 years after repair of
a unilateral HCF. The scan demonstrated increased
radiotracer uptake in the contralateral elbow in the
region of the lateral epicondylar crest. Radiograph-
ically, periosteal proliferation was present on the lat-
eral epicondylar crest.
Histoputhalogy
Biopsy samples were collected at the time of sur-
gical repair of an HCF from two dogs in group 3.
Both fractures had occurred 4 days previously. The
bone at the fracture surfaces in the two biopsy sam-
ples grossly appeared smooth and eburnated (Fig 7).
Histologically, the tissue consisted of dense cancel-
lous bone. Fibrous tissue, increased osteoclastic ac-
tivity, and an increased number of plasma cells were
observed at the fracture sites (Fig 8). No chondro-
cytes or cartilage matrix were observed microscop-
ically.
Genetic Anulysis
The mean coefficient of inbreeding of eight af-
fected cocker spaniels from group 3 was 9.08% (+
10.73; Table 2). The mean coefficient of inbreeding
in the control population of cocker spaniels was
2.1 1% (+ 5.20). The eight affected animals with ped-
igree data all had the same eight common ancestors
present in their pedigree within 12 generations. At
 MARCELLIN-LITTLE E T A L 483

We hypothesized that a heritable condition re-

Fig 5. Line drawing demonstrates the positioning of the dogs
on the computed tomograph palate, with the elbow in an extended
position. The plane indicates the level of the computed tomo-
graphic sections in Fig 6.
least one of the common ancestors was present on
both the sire and dam sides of the genealogy of af-
fected dogs. Also, these common ancestors were not
ancestors of any of the 100 control cocker spaniels.
DISCUSSION
Spaniels, in general, appear to be predisposed to
fractures involving the distal humeral condyle, in
comparison with other breeds treated at the NCSU-
VTH. Affected dogs were often middle-aged males
of normal body weight. These findings are consistent
with previous reports of HCF in spaniel^.^.^.^ The
cause of the HCF in our group of spaniels was either
unknown or considered to be due to minor trauma.
These findings suggest the presence of a predisposing
factor in spaniels that ultimately results in an HCF.
Although multiple causes have been suggested in
previous ~ - e p o r t s , ~ .none
~ . ~ *has
* been investigated.
sulting in incomplete ossification of the distal hu-
meral condyle predisposes several spaniel breeds to
HCF. In the normal canine humeral condyle, two
separate centers of ossification appear at 14 k 8 days
after birth.'* One ossification center includes the
capitulum and the lateral aspect of the humeral con-
dyle, whereas the other ossification center develops
into the trochlea and the medial aspect of the con-
dyle. As ossification progresses, the two centers of
ossification are separated by a thin cartilaginous plate
(Fig 9),* and normally unite at 70 k 14 days after
birth.I3
The radiolucent lines identified on the radiographs
of 1 1 skeletally mature purebred spaniels and one
crossbred spaniel had an identical location to the
cartilaginous remnant present between the lateral
and medial centers of ossification of the humeral
condyle in immature dogs. All 12 dogs in which a
radiolucent line was identified had an HCF in the
contralateral elbow. In addition, the HCF of the
purebred spaniels and crossbred spaniel occurred at
the junction of the two condylar ossification centers.
In this study, intercondylar fractures ( 5 1%) were
more common than previously reported in
Fractures of the lateral portion of the humeral con-
dyle (34%) were more common than fractures of the
medial portion of the humeral condyle (14%).
The radiolucent line in the condyle was observed
only with a cranio-caudal radiograph of the humeral
condyle. The correct orientation was achieved by
placing the dog in ventral recumbency with the fore-

Fig 6. Computed tomographic images from three different dogs taken at the level of the elbow joint illustrated in Fig 5. (A) Normal
humeral condyle, (B) humeral condyle with mildly abnormal ossification (dog no. 23, 10-year-old castrated male cocker spaniel),
and a humeral condyle with severely abnormal ossification (dog no. 15, 5-year-old male cocker spaniel). Arrows point at the area of
incomplete ossification. The subchondral bone is thickened around the area of incomplete ossification in both B and C.
484 INCOMPLETE OSSIFICATION OF T H E HUMERAL CONDYLE
Fig 7. Fractured surface of the humeral condyle in dog no. 3
(10-year-old castrated male cocker spaniel), intraoperatively.
Sclerotic cancellous bone is present in the humeral condyle (ar-
rowheads). A biopsy sample has been collected from the center
of the condyle (arrow).
leg extended, and the radiographic beam directed
from 15" cranio-medially to 15" caudo-laterally.
Rotation of the condyle greater than 5 degrees from
that incidence resulted in an inability to image the
radiolucent line (unpublished data, Marcellin-Little
DJ, 1992). In a recent report of unilateral HCF, the
inability to detect a radiolucent line in the contra-
lateral condyle was possibly caused by malposition-
ing of the limb.'
Fragmented medial coronoid process and devel-
opmental elbow joint disease are uncommon in
spaniel^.'^.'^ In this study, fragmented medial cor-
onoid processes were common findings in elbows
with fractured or intact humeral condyles. Fracture
of the medial coronoid process could result from
motion between the medial and lateral portions of
the humeral condyle.
Frequently, a smooth periosteal proliferation was
observed radiographically along the lateral and cau-
dal surface of the humeral condyle. In intact condyles
with a complete radiolucent line, periosteal prolif-
eration on the lateral epicondylar crest could result
from two mechanisms. First, microfractures and in-
stability of the lateral aspect of the humeral condyle
could result in a bone healing response. Second, ab-
normal stresses placed on the weakened condyle with
incomplete ossification could lead to remodeling of
the lateral epicondyle to strengthen the bone. Peri-
osteal proliferation was not observed in condyles
with incomplete radiolucent lines, suggesting that
instability was not present.
Four spaniels with incomplete ossification of the
humeral condyle also had bilateral radius curvus,
consistent with premature closure of the distal ulnar
physis. The relationship of these findings to the frag-
mented medial coronoid process and incomplete os-
sification of the humeral condyle cannot be deter-
mined from this study. However, it is possible that
growth discrepancies between the radius and ulna
could alter the forces applied to the growing humeral
condyle and interfere with normal ossification. The
absence of recognized incomplete ossification of the
humeral condyle in severely chondrodysplastic dogs
with radius curvus makes the association of these
two problems purely speculative.
The consistent presence of radiographic abnor-
malities in humeral condyles with incomplete ossi-
Fig 8. Histology of the fractured surface. (Hematoxylin-eosin;
original magnification X90.) The biopsy sample was collected
in a 4-day-old fracture of the lateral portion of the humeral con-
dyle (dog no. 23, 10-year-old castrated male cocker spaniel).
Fibrous tissue (arrows) covers dense cancellous bone (arrow-
heads).
 MARCELLIN-LITTLE ET AL
fication supports the recommendation for routine
radiographic evaluation of the contralateral elbow
after an HCF occurring with normal activity in a
mature spaniel.
Chronic HCF resulting in extensive degenerative
joint disease may be misdiagnosed as a neoplastic pro-
cess. In two spaniels presented at NCSU-VTH with
HCF (dogs no. 10 and no. 27), a preliminary radio-
graphic diagnosis of bone neoplasia had been made.
The radiographic diagnosis of neoplasia is suggested
by the signalment (older dogs), history (absence of
485
Fig 9. Contact radiograph (A), gross morphology (B), and his-
tology (I lematoxylin-eosin; original magnification X4) (C) of a
distal humeral condyle of a 10-week-old dog. The distal humeral
condylar physeal plate (arrowheads) is observed. A cartilaginous
plate (arrows) separates the centers of ossification of the medial
and lateral aspects of the distal humeral condyle.
trauma), clinical signs (non-weight bearing lameness
and pain on flexion of the elbow joint), and the fact
that HCF are often unilateral (75%).
Examination of the biopsy specimens obtained
from the fracture sites in two dogs showed fibrous
tissue and increased numbers of plasma cells, si:g-
gesting a chronic inflammatory process. Dense can-
cellous bone adjacent to the fracture site was con-
sistent with the radiographic appearance of bone
sclerosis, suggesting prior bone remodeling in the
humeral condyle.
486 INCOMPLETE OSSIFICATION OF THE HUMERAL CONDYLE
The high failure rate of surgical repair in affected
cocker spaniels may have been caused by abnormal
healing of the condyle. The presence of fibrous tissue
or dense cancellous bone, or both, at the fracture
site may have delayed or prevented bone healing.
Micromotion at the fracture site may lead to cyclic
fatigue and failure of the implant.
Very mild abnormalities were present on radio-
graphs and computed tomography scan of four dogs
with partial radiolucent lines. However, with no ap-
parent trauma, two of the four dogs fractured their
humeral condyle within 6 months after the initial
evaluation, suggesting that these humeral condyles
were at increased risk of fracture regardless of the
seventy of the radiographic changes. Prophylactic
placement of a transcondylar bone screw in non-
fractured condyles could be considered when a ra-
diolucent line or a periosteal reaction are present.
Lameness was inconsistent in dogs with a radio-
lucent line in the humeral condyle. Only two of eight
dogs (no. 17 and no. 28) with a radiolucent line and
degenerative joint disease were lame. None of the
four dogs with a radiolucent line and no degenerative
joint disease were lame. The diagnosis of incomplete
ossification of the contralateral humeral condyle was
often made after the occurrence of an HCF or a
chronic forelimb lameness caused by degenerative
joint disease of the elbow joint.
The incomplete ossification of the humeral con-
dyle in cocker spaniels appears to have a genetic ba-
sis. The coefficient of inbreeding of the eight affected
dogs with genetic information tended to be higher
than the coefficient of inbreeding ofthe control pop-
ulation. Also, specific common ancestors to both the
sire and the dam of affected dogs were identified.
This pattern is suggestive of a recessive mode of in-
heritance. Appropriate test mating between affected
animals, controls, and FI progeny would help con-
firm this hypothesis.
Predisposition of male versus female cocker span-
iels to the development of HCF could be an effect
of more rapid bone growth leading to structural ab-
normalities of the subchondral bone,I6 or an effect
of lower concentration of circulating estrogen hor-
mones, which have a stimulatory effect on epiphyseal
maturation,” or both. These factors may exacerbate
the expression of the disease in affected males com-
pared with affected females.
The epidemiological, radiographic, and histolog-
ical findings in this study support the hypothesis that
affected spaniels had a disturbance in ossification of
the humeral condyle, resulting in a mechanically
weak region in the condyle, predisposing affected
dogs to fractures. Incomplete ossification and de-
velopment of a fibrous union within the condyle re-
sulted in variable clinical expression of the disease,
ranging from partial fusion with no clinical signs to
complete fractuw of the humeral condyle. Osteo-
arthritis, supracondylar periosteal reaction, or any
combination of these signs may be observed radio-
graphically. Because the disease may have a genetic
basis, we recommend that affected dogs and their
sires and dams not be bred. Elbow joints of the lit-
termates of affected dogs that are maintained for
breeding should be evaluated radiographically. The
coefficient of inbreeding should be kept low in af-
fected families. Because fractures occur mostly in
middle-aged dogs, affected dogs may have had off-
spring before a humeral condyle with incomplete
ossification was diagnosed, making control of this
disease difficult. Further studies are necessary to de-
fine the precise pathogenesis and genetic transmis-
sion of humeral condyles with incomplete ossifica-
tion in the dog.
ACKNOWLEDGMENT
We thank Bonnie DeYoung for her technical assistance;
Frank Fair of the American Kennel Club for his help in
compiling pedigree information; Ree Coan, Wendy Sav-
age, Jennifer Graeber, and Joe Trumpey for photography
and illustration; and Anne Hellkamp and Judith Jaya-
wickrama for their statistical assistance.
REFERENCES
1. Bardet JF, Hohn RB. Rudy RL, et al: Fractures of the hu-
merus in dogs and cats. A retrospective study of I30 cases.
Vet Surg 12:73-77. 1983
2. Denny HR: Condylar fractures of the humerus in the dog;
a review of 133 cases. J Small Anim Pract 24:185-197,
1983
3. Cockett PA, Clayton Jones DG: The incidence of humeral
condylar fractures in the dog: A survey of seventy-nine
cases. J Small Anim Pract 26:437-444, 1985
4. Vannini R, Olmstead ML, Smeak DD: An epidemiological
study of I5 1 distal humeral fractures in dogs and cats. J
Am Anim Hosp Assoc 24:531-536, 1988
5. Anderson TJ. Carmichael S, Miller A: Intercondylar hu-
meral fracture in the dog: A review of 20 cases. J Small
Anirn Pract 3 1:437-442. 1990
 MARCELLIN-LITTLE ET A L
6. Vannini R. Olmstead ML, Smeak DD: Humeral condylar
fracture caused by minor trauma in 20 adult dogs. J Am
Anim Hosp Assoc 24:355-362, 1988 13.
7. Meustege FJ: Incomplete condylar fracture in the canine
humerus as a cause of obscure elbow lameness. Proceed-
ings of the 16th Annual Vet Ortho Sot Mtg. Park City. 14.
UT, 1989, p 1 1
8. DrapC J: Pathoghie des fractures du condyle IatCral de I’hu-
m h s . Prat MCd Chir 5:523-529, 1990 15.
9. Kaderly RE, Lamothe M: Incomplete humeral condylar
fracture due to minor trauma in a mature cocker spaniel.
J Am Anim Hosp Assoc 28:36 1-364, 1992 16.
10. Henry WB: Radiographic diagnosis and surgical manage-
ment of fragmented medical coronoid process in dogs. J
Am Vet Med Assoc 184:799-805, 1984 17.
1 1 . Willis MB: Inbreeding, in Willis MB (ed): Genetics of the
Dog. New York, NY, Macmillan. 1989, pp 320-332
12. Hare WCD: The age at which the centers of ossification
487
appear roentgenographically in the limb bones of the dog.
Am J Vet Res 223254335, 196 1
Ticer JW: General principles, in Ticer J W (ed): Radiographic
Technique in Small Animal Practice. Philadelphia, PA,
Saunders, 1975, pp 97- I02
Berzon JL, Quick CB: Fragmented coronoid process: An-
atomical, clinical and radiographic considerations with
case analyses. J Am Anim Hosp Assoc 16:24 1-25 1, 1980
Wind AP: Elbow incongruity and developmental elbow dis-
eases in the dog: part 11. J Am Anim Hosp Assoc 22:725-
730, 1986
Dammrich K: Relationship between nutrition and bone
growth in large and giant dogs. J Nutr 121:SI 14-S121,
1991
Strickland AL, Sprinz H: Studies of the influence of estradiol
and growth hormone on the hypophysectomized, im-
mature rat epiphyseal cartilage growth plate. Am J Obstret
Gynecol 115:471-477. 1973