Vet. Pathol.

19: 337-354 (1982)

Canine Hypertrophic Osteodystrophy, A Study of the Spontaneous

Disease in Littermates

J. C. WOODARD

Division of Comparative Pathology, College of Veterinary Medicine, University of Florida,

Gainesville, Fla.

Abstract. All members of a weimaraner litter had clinical and radiographic signs of
hypertrophic osteodystrophy shortly after weaning. Three dogs were necropsied. Radiographic
metaphyseal densities, which are used to make a clinical diagnosis of hypertrophic osteodys-
trophy, were found to result from elongation of the calcified cartilage lattice of the primary
spongiosa. Intertrabecular acute inflammation was associated with necrosis, failure to deposit
osseous tissue on the calcified-cartilage lattice, and trabecular microfractures. This process led
to metaphyseal infraction and separation of the epiphysis. Defective bone formation (osteo-
dystrophy) was considered a secondary process resulting from inflammation of osteochondral
complexes, marrow, and periosteum. Enamel hypoplasia also was found to be associated with
inflammation of the dental crypt, and abnormal enamel matrix was observed in the developing
teeth. The histopathology of the bones and teeth was different from alterations which occur in
infantile scurvy or congenital syphilis, although these diseases of man have radiographic
similarities to canine hypertrophic osteodystrophy. Because the radiologic lesion is nonspecific,
a clinical diagnosis of hypertrophic osteodystrophy is not necessarily diagnostic of a specific
disease due to a single etiologic agent. Liver levels of ascorbic acid were within the normal
range. Although an infectious agent could not be identified, the conditions may have an
infectious origin with systemic manifestations.

Canine hypertropic osteodystrophy (metaphyseal osteopathy) is a disease of young

growing dogs and mainly affects large breeds. The spontaneous disease is character-
ized by the clinical signs of pain, lameness, and pyrexia, and by swelling and
hyperthermia of the metaphysis of the long bones. Radiographic changes are limited
usually to the ends of long bones, and lesions are symmetric bilaterally. Frequently,
there is lipping of the metaphyses of long bones, and a line of increased density is
observed next to the growth plate with an adjacent metaphyseal radiolucent zone
[26]. The costochondral junctions are enlarged with increased radiodensity.
The disease first was considered to be due to vitamin C deficiency because it
resembled human infantile scurvy [ 161. Many veterinary practitioners still consider
this disease to be canine scurvy, and large doses of ascorbic acid are claimed to cure
the disease [ll, 241. Recovery from the illness is noted also when affected dogs
receive other therapy. Some investigators think the cause is related to overnutrition
337
338 Woodard

and rapid growth, results from mineral overload, or has a multifactorial etiology [ 19,
20, 221.
Canine hypertrophic osteodystrophy occurs rather frequently, but there are few
reports of the histologic alterations occurring in dogs with the spontaneous disease
[8, 11, 171. The specificity of the radiologic changes for this disease often are taken
for granted, although the similarity between microscopic alterations has been
demonstrated in only a few instances. It is not certain that the reported histologic
changes in canine hypertrophic osteodystrophy are typical of scurvy. Moreover,
radiodense zones similar to those observed in human scurvy also have been found in
many diverse diseases [3].
The occurrence of hypertrophic osteodystrophy in a litter of weimaraner dogs
allowed us to study early histologic alterations in puppies of different ages and to
investigate certain aspects concerning the pathogenesis of this condition. The lesions
observed in dogs are compared to scorbutic lesions induced experimentally in guinea
pigs.

Case History

A litter of four weimaraner puppies, three females and one male, were weaned and
fed a commercial dry dog food supplemented with canned milk. The dogs received
postnatal veterinary care, including therapy for hookworms. The first dog to develop
clinical signs of hypertrophic osteodystrophy became listless and pyrexic at seven
weeks of age. This puppy was treated with chloramphenicol, analgesics, and vitamin
C and recovered over the next seven days and remained healthy.
Two of the other three dogs in this litter became ill at eight weeks of age. Dog 1
was severely ill and was treated with chloramphenicol and a sedative but was not
given vitamin C. The temperature was elevated, and there was leukocytosis with
increased immature neutrophils in the blood. An increased quantity of synovial fluid
was seen within the left elbow joint, and joint fluid contained polymorphonuclear
leukocytes. Bacteria could not be cultured from the joint fluid on blood agar. This
dog was in severe pain and would not stand, therefore it was killed.
Dog 2, which also became sick at eight weeks of age, was given chloramphenicol,
analgesics, and vitamin C . The dog seemed to recover, but had a relapse 20 days
after initial illness. This second illness lasted two weeks until the dog died at home.
Dog 3 became lame at 15 weeks of age. The dog was treated with chloramphenicol,
recovered, and was placed on daily oral vitamin C until it became ill again. The dog
had severe pain, remained recumbent for two weeks, but gradually improved. At 18
weeks of age it became ill again and was treated with vitamin C and analgesics. The
dog was able to walk for five weeks and became very active. The dog again became
severely ill; it then was given intravenous vitamin C and chloramphenicol. The dog’s
condition progressively became worse, and it was killed at 3 1 weeks of age.
 Canine Hypertrophic Osteodystrophy 339

Materials and Methods

The three dogs that failed to recover from the disease (dog 1, 8 weeks; dog 2, 13 weeks; dog
3, 31 weeks of age) were necropsied. Samples of blood, synovial fluid, lung, lymph node,
heart, and bone were cultured for bacteria by aerobic and anaerobic methods. Samples of liver
tissue were frozen and stored at -60°C prior to determining the content of vitamin C.
For comparative purposes, five female weanling guinea pigs of the Hartley strain (Charles
River Breeding Labs, Wilmington, Mass.) were fed a vitamin C-deficient diet (ICN Nutritional
Biochemicals, Cleveland, Ohio), and three guinea pigs were fed guinea pig chow (Ralston
Purina Co., St. Louis, Mo.). Two guinea pigs died with scorbutic lesions during the third week
and the remaining guinea pigs were killed. Bones from the killed guinea pigs and from dog 1
were futed with modified Karnovsky’s fmative, and tissues were prepared for electron micros-
copy by standard methods [9]. Other tissues from the three dogs were futed in 10% neutral
buffered formalin, and sections from bone and soft tissue were stained with hematoxylin and
eosin (HE), Taylor, Goodpasture, Brown-Hopp, Fite acid-fast, Warthin-Starry, Gomori-
methenamine silver, or von Kossa’s silver techniques [ 141. Various bones from the three dogs
were radiographed at necropsy so comparisons could be made between the radiographic and
microscopic appearance. Bone tissue from dogs 1 and 3 was embedded in methyl methacrylate,
and microradiographs of 100 pm undecalcified ground sections were obtained [25].
Liver ascorbic acid in dogs was compared with values from a guinea pig, an animal which
lacks the terminal synthetic enzyme 2-gulonolactone oxidase, and a rat, an animal that
synthesizes ascorbic acid in the liver. A weanling female, Sprague-Dawley rat and Hartley
guinea pig were fed rodent lab chow or guinea pig chow for three weeks. Freshly frozen liver
tissue from the three dogs, the rat, and the guinea pig was homogenized for two minutes in
nine volumes of ice cold 0.25 M sucrose with a plastic homogenizing pestle of the Potter-
Elvehjem type. Total ascorbic acid per gram weight of liver tissue was determined by the 2,4-
dinitrophenylhydrazine method [lo].

Results
Dog 1
Gross findings: This eight-week-old pup was in good nutritional condition and
weighed 4.5 kg. There was a bilateral, mucopurulent, nasal discharge, and some
mucopurulent material was present within the nasal cavity. Lungs contained scattered
reddened areas. There were several pinpoint hemorrhages in the pyloric mucosa and
scattered petechiae on the serosa of the bladder. The papillary muscle of the right
ventricle was yellow.
Joints of all limbs were enlarged and slightly swollen. Sawed longitudinal sections
of long bones had a 1- to 3-mm wide pale yellow zone with a soft, crumbly consistency
on the metaphyseal side of the growth cartilage. There was infraction (incomplete
fracture) of the adjacent metaphyseal bone (fig. 1). Radiographs of sawed bone slabs
illustrated that the pale yellow zone was radiodense, and the bone infraction produced
a radiolucent line across the adjacent metaphysis (fig. 2). The bone changes were
bilaterally symmetric and were most prominent at the costochondral junctions and
bone ends of the humerus, radius, femur, tibia, and distal ulna. Lesions in the first
rib were not as severe as the lesions in the other ribs which had greater respiratory
movement. Gross lesions were not observed in the vertebrae, epiphyses of long bones,
or in the proximal ulna.
340 Woodard

Fig. 1: Longitudinal section of distal humerus, dog 1. Pale zone beneath hyaline cartilage
of growth plate (arrows) and adjacent infraction.
Fig. 2: Radiograph, longitudinally sawed distal humerus, dog 1 (Cf fig. 1). Radiodense area
corresponds to the pale zone in gross specimen.
Fig. 3 Microradiograph, dog 1. Trabeculae of epiphysis (E), normal. Dense zone beneath
growth plate composed of elongated calcified cartilage lattice (arrow), impaction of fractured
trabeculae (F) within metaphysis (M).

Light microscopicfindings: There was a rather severe diffuse interstitial pneumonia

with mononuclear cell infiltrates and moderate atelectasis. The liver contained
histiocytic foci with occasional central minerahzed zone. Focal areas of mineralization
also were observed within the renal cortex. There was severe acute myocarditis of the
right ventricle with mineralization of the myocardial fibers. Lymphoid hyperplasia
 Canine Hypertrophic Osteodystrophy 34 I

was evident in spleen and lymph nodes, but microscopic lesions were not recognized
within urinary bladder, adrenal, thyroid, parathyroid, pancreas, intestinal tract, or
brain.
Microradiographs of the bone end demonstrated normal thickness and radiodensity
of the trabecular bone of the epiphysis and metaphysis (fig. 3). The radiodense zone
observed in the gross specimen was localized in the microradiograph. It consisted of
an elongated calcified cartilage lattice of the primary spongiosa and impacted
trabeculae (fig. 4). The vessels of the primary spongiosa were dilated greatly.
Intertrabecular interstitial spaces were infiltrated with inflammatory cells, principally
neutrophils. Small foci of necrosis and fibrin deposition could be seen in this area
(fig. 5). In addition to elongation of the cartilaginous lattice, the normal investment
of osseous tissue on the mineralized cartilage trabeculae was lacking. Most of the
osteoblasts in this area had pyknotic nuclei. A strihng finding was the marked degree
of suppurative inflammation below the primary spongiosa (fig. 6). In addition,
microfractures of poorly ossified cartilaginous trabeculae were frequent, and there
was partial to total infraction across the width of the metaphysis below the primary
spongiosa. The infracted area was composed of necrotic tissue, fibrin, suppurative
exudate, cartilage and bone remnants, and numerous osteoclasts. On the diaphyseal
side of the infracted region, secondary trabeculae were formed of woven bone in a
haphazard manner. The new osseous tissue was being deposited on pre-existing
fractured trabecular fragments which often were necrotic. Osteoblastic activity was
intense in this area with a variable degree of fibroplasia. Subperiosteal suppurative
inflammation was evident, and subperiosteal trabeculae of woven bone were forming.
Although gross lesions were not observed in the vertebrae, in epiphyses of long
bones, or in the proximal ulna, microscopic lesions were present. There was acute
osteochondritis in the secondary ossification center of the proximal ulna, in the
epiphyseal growth zone adjacent to the articular bone surface, and in zones of
vertebral endochondral ossification. (The term osteochondritis is used to indicate
acute inflammation of osteochondral complexes and does not signify lesions of
osteochondrosis dissecans.) The main microscopic characteristics were vascular dil-
atation, suppurative inflammation, necrosis, fibrin deposition and intense osteoclastic
activity (fig. 7). These inflammatory lesions were even more apparent in the maxillary
bone where there was an intense osteomyelitis and osteoperiostitis.

Dog 2

Grossfindings: The animal was 13 weeks of age and somewhat thin. Lungs were
red and moderately firm. Other gross findings were limited to the bones. Costochon-
dral junctions were enlarged and hyperemic. Longitudinal sections showed a hard,
pale yellow zone beneath the rib cartilage, but infraction of this region was not visible
grossly. Radiographs revealed moderate radiodensity beneath the growth plate. No
other obvious lesions were identified within the metaphyses.
Light microscopicfindings: There was a mild interstitial pneumonia, and the liver
342 Woodard

Fig. 4 Primary spongiosa, dog 1. Dilatation of vessels (V), intertrabecular hypercellularity.

Calcified cartilage lattice poorly ossified. HE.
Fig. 5 Growth plate-metaphyseal junction, dog 1. Edema, acute inflammation, and necro-
sis. HE.
Fig. 6 Suppurative inflammation (S), trabecular fractures (TF), and necrosis (N) below
elongated cartilage lattice (C). HE.
Fig. 7: Beginning secondary ossification center of epiphysis illustrates acute inflammation
(I), numerous osteoclasts (arrows). HE.
 Canine Hypertrophic Osteodystrophy 343

Fig. 8: Primary spongiosa, dog 2. Intertrabecular tissue infiltrated with neutrophils.Absence

of osteoblasts and osseous tissue which normally invest calcified cartilage lattice. HE.

contained moderate periportal, mononuclear cell infiltrates. Focal mineralization was

noted in lung and kidney, and lymph nodes were hyperplastic and edematous.
Sections of rib, distal ulna, and proximal radius were examined and showed similar
changes although of varying severity. The most significant alterations were observed
in the primary spongiosa of the metaphyses. This region was much wider than
normal, and there was necrosis of the normal mesenchymal components of the
primary spongiosa. The intertrabecular interstitium was infiltrated with massive
numbers of polymorphonuclear leukocytes. Trabeculae of the primary spongiosa
were composed of a thin, anastomosing cartilage lattice which lacked its normal
osseous tissue covering and osteoblasts (fig. 8). Microfractures of these trabeculae
sometimes were seen, but there was no infraction of this region.
Osteomyelitis was evident in the mandible, and there were hyperemia, edema, and
cellular exudation associated with the rich capillary network of the dental sac. The
odontoblastic layer lining the dental pulp of the developing canine tooth was well
differentiated, and predentine and dentine had formed normally. The dental pulp
contained pulpal calcifications. These false denticles appeared as focal homogeneous
mineralized zones or as concentric layers of calcified tissue surrounding a basophilic
homogeneous nucleus. In the region of the cervical loop and along the base of the
developing tooth, ameloblasts were identified, but these cells were absent along the
sides or at the apex of the cap-like enamel organ. A broad sheet of cells resembling
outer enamel epithelium covered the apex and sides of the external dentine surface.
344 Woodard

Table I. Hematologic values, dog 3

White blood cells 40,800 cells/pl
Differential
neutrophils 38,352 cells/pl
lymphocytes 816 cells/pl
monocytes 2,488 cells/pl
Red blood cells 5.9 x lo6 cells/pl
Hemoglobin 12.4 g/dl
Packed cell volume 38%
Mean corpuscular volume 62 fl
Mean corpuscular hemoglobin concentration 33 g/dl
Mean corpuscular hemoglobin 21 Pg
Platelets 115,000 cells/pl
Prothrombin time 8 sec
Activated partial thromboplastin time 54 sec
Plasma protein 6.4 g/dl
Fibrinogen 200 mg/dl
Calcium 11.3 mg/dl
Phosphorus 7.7 m" ddl

Cells resembling the inner enamel epithelium, stratum intermedium, or stellate

reticulum could not be identified in this location. There was abortive formation of a
narrow rim of abnormal enamel matrix located between the dentine and cells of the
enamel organ.

Dog 3
GrossJindings: A blood sample was taken prior to killing this seven-month-old
dog, and hematologic values are presented in table I. Major alterations were a
leukocytosis and an increased activated partial thromboplastin time. The dog was
malnourished and weighed only 16.8 kg. The incisor and molar teeth had a gray-
brown color, and the gums contained small, raised hemorrhagic areas. Irregular, flat,
purple patches of hemorrhage, measuring from 1 cm in diameter to 3 X 5 cm, were
present in the skin of the ventral abdomen. The center of one of these hemorrhagic
areas was ulcerated. The serosal surface of the stomach had raised, longitudinal
hemorrhagic streaks, suffusions, and ecchymoses. These sanguinous areas extended
deep into the stomach muscle. The surface of the spleen was wrinkled, and the
splenic capsule had a frosted appearance and it felt gritty. The cut surface of the
kidney showed moderate swelling; the cortex was pale and contained radially
arranged, fine yellow streaks. Lungs had bilateral areas of consolidation in the apical
and cardiac lobes. There was an extensive mural vegetative endocarditis in the left
atrium; the surface of mural thrombi was translucent white and gritty. Yellow streaks
and foci were present in the left ventricular myocardium.
Soft tissue swelling was noted around the distal radius-ulna and tibia. Necropsy
radiographs demonstrated periosteal bone proliferation around the distal radius,
ulna, humerus, and tibia and around the proximal radius and ulna. This was
 Canine Hypertrophic Osteodystrophy 345

Fig. 9 Macerated ulna, dog 3. Gross lesions of ossifying periostitis at proximal and distal
metaphyseal regions.
Fig. 10 Proximal humerus, dog 3. Metaphyseal infraction just below the growth cartilage.
Focal regions with more severe suppuration appear as white zones surrounded by dark halo
(arrows).

particularly evident in macerated bone specimens (fig. 9). The costochondral areas
were bilaterally enlarged and reddened. The cut surface of the ribs appeared similar
to the lesions described in the bones of dog 1. Radiographs of the long bones
demonstrated irregular metaphyseal densities. Longitudinally sawed bones showed
metaphyseal infraction adjacent to growth cartilage. In addition, focal light-colored
foci surrounded by a red halo were seen within the epiphyses and within metaphyses
(fig. 10). The synovial fluid of the left stifle was sanguinous.
Light microscopic findings: Soft tissue findings were characterized by suppurative
inflammation and calcinosis. There was a suppurative bronchopneumonia and
mineralization of bronchiolar walls and lung parenchyma. Diffuse and focal areas of
suppuration were found within the gastric muscular wall. In the centers of abscesses,
mineral deposits could be visualized with silver stains. There was periarterial inflam-
mation and mineralization (fig. 11). The hemorrhagic skin lesions also were charac-
terized by phlegmonous inflammation, and spherical mineral deposits were located
within the edematous reticular dermis. Extensive mineralization was noted in the
myocardium. There was severe acute endocarditis, and white mural thrombi were
permeated with mineralized material. Severe mineralization was present in the
splenic capsule, septa, and germinal centers. Nephrocalcinosis was severe also. In the
kidney, mineral could be detected in the connective tissue around arcuate arteries, in
346 Woodard

Fig. 11: Abscess and periarteritis in smooth muscle of stomach, dog 3. Basophilic area of
mineralization around vessel. HE. Inset: Mineral in center of abscess. von Kossa.
Fig. 1 2 Growth plate, dog 3. Zone of mineralized cartilage wider than normal, delineated
by basophilic line resembling “tidemark,” irregularly penetrated by vessels. Zone with hem-
orrhage, necrosis, and trabecular fractures adjacent to growth plate. HE.
 Canine Hypertrophic Osteodystrophy 347

Fig. 13: Ossifying periostitis,dog 3. Periosteal formation of woven bone trabeculae (arrows).
HE.

the interstitial spaces between tubules, and within the renal tubular epithelium and
lumina. The renal cortex was affected more severely than the medulla. Inflammation
of kidney was moderate, chronic, and was associated with some fibrosis. Arterioscle-
rosis was found in the coronary arteries, whereas mineralization of the vascular wall
was observed in medium-sized arteries of the stomach and perithyroidal connective
tissue. The thyroids and parathyroids appeared normal.
There were some minor differences in the microscopic appearance of the bones of
this dog as compared to dog 1, although suppurative inflammation was the underlying
lesion in both dogs. Some bones had no primary spongiosa. The infracted zone,
consisting of necrotic material, hemorrhage, fibrin, fractured metaphyseal trabeculae,
and suppurative exudate, was located directly adjacent to the growth plate. The
mineralized cartilage zone of the growth plate was much wider than normal. Its
width could be easily identified by the basophilic line, similar to the “tidemark”
beneath articular cartilage (fig. 12). In some bone ends, the necrotic material in the
infracted region was diminished. The grumous material seemed to have been extruded
through a sinus by mechanical action, and purulent exudate and fractured trabecular
fragments had dissected the perichondral ring. Ossifying periostitis was quite severe
(fig. 13). Microabscesses were present in a subperiosteal and extraperiosteal location.
There was an admixture of osteoblastic production of woven bone trabeculae,
osteoclastic resorption, and phlegmonous inflammation within the thickened peri-
osteal tissue. The metaphyseal region on the diaphyseal side of the infracted region
had alternating areas of trabecular sclerosis and suppuration. Microradiographs of
348 Woodard

undecalcified sections of the incisor and premolar teeth illustrated an irregular, thin
enamel layer that incompletely covered the crown.
Ultrastructural findings: Because of sampling problems attendant with electron
microscopy, it often was difficult to determine exactly which metaphyseal zone was
being examined ultrastructurally. Some mineralized surfaces were devoid of cells,
whereas other areas had osteoblasts with abundant rough endoplasmic reticulum.
These osteoblasts abutted zones of mineralizing osteoid (fig. 14). Numerous poly-
morphonuclear leukocytes, some macrophages, and necrotic cells of an unidentifiable
type were observed. The neutrophils contained numerous cytoplasmic lysosomal
bodies, but phagocytic vacuoles or microorganisms were not seen within these cells.
Several osteoclasts were observed; some had ruffled borders and were found adjacent
to resorption cavities.
Comparison with experimental scurvy: The gross and light microscopic features of
bones from scorbutic guinea pigs did not differ from those previously described
[15]. A notable difference between the lesion in guinea pigs and dog 1 was the
absence of significant numbers of polymorphonuclear leukocytes in guinea pigs.
Rather, the primary spongiosa of scorbutic guinea pigs contained many osteoprogen-
itor cells. In addition, periosteal hemorrhage occurred in guinea pigs, whereas
osteoperiostitis was observed in the dog. There was also a marked contrast in the
ultrastructural characteristics of the osteoblasts (Cf fig. 14, 15). In the scorbutic
guinea pig, osteoblasts did not contain a fully developed rough endoplasmic reticu-
lum, rather, ribosomes were scattered within the cytosol. Osteoid was not observed
on mineralized surfaces; instead, abundant quantities of unbanded fibrillar material
were adjacent to the hypoplastic osteoblasts (fig. 15).
Vitamin C assay and microbiology: The results of vitamin C assays are presented
in table 11. The range of values for the three dogs was similar to the values obtained
for rat and guinea pig. No bacteria were cultured from dog 1 or dog 3 with the
exception of the heart. Many bacteria, including bacilli, were cultured from the heart
of dog 3, but these were considered contaminants. Serum samples from dogs 1 and
3 had negative titers for Leptospira and Brucella. No microorganisms were identified
in stained tissue sections from the three dogs or by ultrastructural examinations of
bone tissues from dog 1.

Discussion
This study of hypertrophic osteodystrophy allowed microscopic observations of
spontaneous lesions during earlier disease stages than previously reported. These
observations permit a better appreciation of pathogenesis and permit an alternate
explanation as to the cause. The original description of this disease illustrated
histologic changes in advanced cases where the reparative phase predominated [ 161.
The gross changes in soft tissues in dog 3 were similar to those previously described
and reported as canine scurvy [ 111. The microscopic lesions in dog 3 also are similar
to the bone lesions previously reported in hypertrophic osteodystrophy [8, 171. The
 Canine Hypertrophic Osteodystrophy 349

Fig. 14 Osteoblast, dog 1. Cell has well-developed rough endoplasmic reticulum. Mature
collagen fibrils seen within mineralizing osteoid (0).
Fig. 1 5 Osteoblast, scorbutic guinea pig. Absence of rough endoplasmic reticulum, dis-
persed distribution of ribosomes. Unbanded fibrillar material (M) present in interstitium.
350 Woodard

Table 11. Liver ascorbic acid

Animal pg/g wet weight
Dog 1 177
Dog 2 175
Dog 3 213
Rat 297
Guinea pig 158

disease mechanism observed in this study probably is typical for dogs with clinical
signs and radiologic changes of hypertrophic osteodystrophy.
Most reports presume hypertrophic osteodystrophy to represent a disorder of bone
growth [l, 221, and some consider it to be due to a vitamin C deficiency [l, 11, 16,24,
261. Several facts could be stated in support of a nutritional cause. First, the
radiographic lesions in the bones resemble infantile scurvy in man. The bilateral
symmetry of the lesions suggests that the disease is metabolic in origin. Soft-tissue
hemorrhage is a characteristic lesion of scurvy and has been observed in dogs at
necropsy. The occurrence of extraperiosteal densities during the disease is interpreted
as hematomas undergoing mineralization [ 111, and the start of mineralization could
be a favorable healing response to ascorbic acid therapy. Serum and urine ascorbic
acid values were below normal values reported for dogs [ l l , 241. Animals may
recover from their illness after ascorbic acid therapy.
The disease occurring spontaneously in our dogs was not the result of vitamin C
deficiency. In fact, the condition probably had an infectious rather than a metabolic
cause. Two of the dogs had recurrence of their disease while receiving oral vitamin
C medication. Tissue ascorbic acid levels are considered a more reliable indication
of nutritional status than serum or urine levels. At necropsy, liver ascorbic acid levels
of all dogs were in a normal range reported for other mammals [2]. The average
tissue saturation value for ascorbic acid in animals ranges between 180 to 400 pg/g
tissue [4]. Even though dog 1 had not received vitamin C medication, the liver level
was similar to the treated dog. The ascorbic acid liver levels of these two dogs were
near saturation since the vitamin level was only slightly higher in dog 3 which had
received continous intravenous ascorbic acid (IV drip) for two days prior to death. In
guinea pigs, signs of scurvy only appear when tissue concentrations are lowered to
less than 20% of saturation [4].
The microscopic bone lesions in these dogs also differ from those observed in
scurvy. The basic defect in scurvy results from a failure of the osteoblast to
manufacture osseous matrix. However, ultrastructural studies illustrated that viable
dog osteoblasts had profiles of endoplasmic reticulum with ribosomal orientation
suggestive of collagen synthesis [ 18,211. Mature banded collagen could be discerned
in osteoid adjacent to mineralized trabeculae. Although infraction of the metaphysis
occurs in scurvy, the microscopic changes in dogs did not resemble scurvy where
zones of normal bone formation are replaced by a narrow framework composed of
mesenchymal cells or abnormal osteoblasts [ 161. Rather than the metaphysis being
 Canine Hypertrophic Osteodystrophy 35 1

composed of immature mesenchymal cells, suppurative inflammation was the pre-

dominant reaction. Suppurative inflammation is not characteristic of scorbutic bone
lesions in human infantile scurvy or experimental scurvy of the guinea pig.
Because the width of the primary spongiosa increases in both conditions, there is
a similarity between the radiographic changes in scurvy and hypertrophic osteodys-
trophy. In both diseases the primary trabeculae have a mineralized cartilage lattice
which is poorly ossified. Due to structural weakness, microfractures occur. This leads
to secondary infraction along a line of stress near the primary and secondary
trabecular interface. This is the sequence in both scurvy and hypertrophic osteodys-
trophy, but failure to ossify the primary trabeculae in the two diseases has separate
causes. In scurvy there is a metabolic defect and an inability to synthesize bone
matrix. In the dogs, acute inflammation of the osteochondral complexes (osteochon-
dritis) was the primary lesion.
Necrosis of the mesenchymal components of the primary spongiosa caused failure
to ossify the cartilaginous trabeculae of dogs. Cartilage of the growth plate obtains
much of its nutriment from branches of epiphyseal vessels. Since blood supply to the
growth cartilage was affected very little, longitudinal bone growth continued. Near
the growth plate, trabeculae in the longitudinal section actually are seen in cross
section to represent interconnecting walls of columns. The mesenchyma and vessels
run through tunnels or canals in the column centers. Since the mineralized column
wall is rigid, it is easy to envision how swelling and exudation associated with acute
inflammation could interfere with vascular perfusion and cell function. Cell kinetic
studies indicate that the metaphyseal osteoblasts develop from regional osteoprogen-
itor cells [ 131. The majority of osteoblasts are found just beneath the growth cartilage-
metaphyseal junction, and these cells migrate with the moving zone. In contrast,
osteoclasts enter the region from the blood and always remain with the area of bone
surface where they first associate [13]. Necrosis of osteoblasts would prevent the
acellular, calcified cartilage cores from becoming invested with a layer of bone.
Osteochondritis, with its attendant vascular reaction, also could prevent osteoclasts
from entering the region. Since osteoclasts remain on the surface where they first
associate, modeling of the trabeculae would not occur as elongation of the thin
trabeculae continued. Trabecular microfractures and bone infraction result from
stresses on the bone. Impaction of the infracted zone prevents normal vascular
invasion and chondrolysis of the calcified hypertrophied-cartilage zone. Increased
width of the calcified cartilage zone of the growth plate was observed in dog 3.
It could be argued that the inflammatory response was a secondary reaction to
bone infraction. This seems unlikely since dog 2 had severe osteochondritis and
metaphyseal osteomyelitis without bone infraction. In addition, inflammatory bone
lesions were seen in other bone locations where gross and radiologic lesions could
not be observed. It also has been suggested that extraperiosteal radiographic densities
represent organizing hematomas [ 1 11. This was not the case as is illustrated in dogs
1 and 2 where early lesions of ossifying periostitis could be seen. In the more
352 Woodard

advanced disease, infraction of the metaphysis and impaction led to the horizontal
displacement of fractured trabeculae and exudate. In these areas the osteoperiostitis
was more severe where the exudate had been expelled through the periochondral
ring into the periosseous connective tissues.
The radiologic characteristics of canine hypertrophic osteodystrophy should not
be considered specific for a single disease, necessarily. In addition to scurvy, radio-
graphic densities of similar appearance have been seen in many other conditions,
including malnutrition, bacteremia, anemia, birth defects, and others [3, 5 , 121.
Because of the profound inflammatory reaction in dogs, the infectious causes of these
lesions are interesting. The radiographic changes in congenital syphilis are almost
identical to those described here, and isolated cases with syphilitic-like radiographic
changes have been reported in human infants with bacteremia due to staphylococcus,
pneumonococcus, gonococcus, and tubercle bacillus [3, 51.
Congenital syphilis is characterized by osteochondritis, and evolution of the
syphilitic bone lesions offers insights into how lesions might develop in dogs.
Congenital syphilis may not be apparent during the first weeks of life, but definite
manifestations develop later. Spirochetes are found, particularly at sites of active
endochondral ossification. The organism or its products cause inflammation and
degeneration of osteoblasts [ 121. There is widening of the provisional calcification
zone, and epiphyseal separation may result. While syphilitic osteochondritis is likely
to involve all regions of endochondral ossification, the radiographic changes are
more or less symmetrically disposed. The lesions in our dogs differed from those of
congenital syphilis in man because the intertrabecular marrow was composed of
suppurative exudate rather than granulomatous inflammatory tissue.
No infectious agents could be identified in bones or soft tissues of dogs. Since the
dogs had received antibiotics prior to examination, this might have inhibited organism
isolation. It is also possible that an infectious agent might disappear rapidly from the
tissue; this is known to occur in congenital syphilis. In our dogs, there may have been
vertical transmission of an infectious agent from the dam or horizontal infection after
birth; the condition was recognized only after promulgation of sufficient clinical
signs.
The soft tissue lesions in dogs were reflective of the suppurative response observed
in all of the bones. The zones of hemorrhage observed in dog 3 were not characteristic
of those seen in scurvy [7]. They were seen in association with phlegmonous
inflammation of the skin, vegetative endocarditis, and necrosis and abscesses of the
stomach wall. Excessive tissue and joint hemorrhage in dog 3 probably was related,
in part, to the observed prolongation of the activated partial thromboplastin time.
The clotting defect seemed to develop as a terminal event. In experimental scurvy,
abnormalities have been noted in both the one-stage prothrombin time and the
thromboplastin generation time [6].
Soft tissue mineralization in hypertrophic osteodystrophy is difficult to explain.
Calcinosis is the general term used to describe calcification of extraskeletal tissues
 Canine Hypertrophic Osteodystrophy 353

when there is no apparent abnormality in calcium or bone metabolism. The highly

saturated extracellular fluid with respect to apatite predisposes all living cells to the
dangers of calcification, and a wide variety of inhibitors may protect extraskeletal
tissues from mineralization. Calcification initiators may be produced in extracellular
sites under pathological conditions. Calcinosis universalis, Monkelberg’s sclerosis,
calcinosis cutis, myocalcinosis traumatica, and nephrocalcinosis may occur in nor-
mocalcemic individuals, and they may be associated with the production of an
initiator [23]. Inflammation or degenerative changes also are thought to produce
calcification initiators or a calcifiable matrix. The association of tissue mineralization
of dogs with acute inflammation of the heart, skin, and stomach might indicate
dystrophic calcification. However, mineralization of the spleen, kidney, and vessel
walls did not seem to result from previous tissue necrosis. It may be significant that
calcinosis was much less severe in dog 2. This dog had few trabecular microfractures
and did not have bone infraction.

Acknowledgements
The author thanks Northwest Animal Hospital, 3811 NW 13th St., Gainesville, Fla. for
clinical cases and Misters Jim Faulk and Alan Bontya for dog donations. I also thank Doctors
Ackerman, Bruss, Harvey, Schaer, and White for assistance with clinical and laboratory
evaluations. This study was supported, in part, by National Research Service Award 1F33
AM 0659 I .

References
1 ALEXANDER, J.W.: Hypertrophic osteodystrophy. Canine Pract 548-52, 1978
2 BESSEY, O.A.; KING,C.G.: The distribution of vitamin C in plant and animal tissues, and
its determination. J Biol Chem 103687-698, 1933
3 CAFFEY,J.: Syphilis of the skeleton in early infancy. The nonspecificity of many of the
roentgenographic changes. Am J Roentgen01 Radium Ther Nucl Med 42637-655, 1939
4 CHATTERJEE, G.C.: Biochemical detection of deficiency. In: The Vitamins I, ed. Sebrell
and Harris, pp. 399-407, 1st ed. Academic Press, New York, 1967
5 EVANS,W.A.: Syphilis of the bones in infancy. Some possible errors in the roentgen
diagnosis. JAMA 115197-200 1940
6 FLUTE,P.T.; HOWARD, A.N.: Blood coagulation in scorbutic guinea pigs: A defect in
activation by glass contact. Br J Haematol5421-430, 1959
7 FRIEDERICI, H.H.R.; TAYLOR, H.; ROSE,R.; PIRANI,C.L.: The fine structure of capillaries
in experimental scurvy. Lab Invest 15 1442-1458, 1966
8 GRONDALEN, J.: Metaphyseal osteopathy (hypertrophic osteodystrophy) in growing dogs.
A clinical study. J Small Anim Pract 17:721-735, 1976
9 HAYAT,M.A.: Principles and Techniques of Electron Microscopy, pp. 103, Vol I . Van
Nostrand Reinhold Co., New York, 1972
10 HENRY,R.J.; CANNON, D.C.; WINKELMAN, J.W.: Clinical Chemistry Principles and Tech-
niques, pp. 1395-1398, 2nd ed. Harper and Row, Hagerstown, Maryland, 1974
11 HOLMES, J.R.: Suspected skeletal scurvy in the dog. Vet Rec 74801-813, 1962
12 JAFFE,H.L.: Metabolic, Degenerative and Inflammatory Diseases of Bones and Joints, pp.
907. Lea and Febiger, Philadelphia, 1972
13 KIMMEL, D.B.;JEE,W.S.S.: Bone cell kinetics during longitudinal bone growth in the rat.
Calcif Tissue Int 3 2 123-133, 1980
354 Woodard

14 LUNA,L.G.: Manual of Histologic Staining Methods of the Armed Forces Institute of

Pathology, pp. 176. McGraw-Hill Book Company, New York, 1968
15 MACLEAN, D.L.; SHEPPARD, M.; MCHENRY, E.W.: Tissue changes in ascorbic acid deficient
guinea pigs. Br J Exp Pathol20451-457, 1939
16 MEIER, H.; CLARK,S.T.; SCHNELLE, G.B.; WILL, D.H.: Hypertrophic osteodystrophy
associated with a disturbance of vitamin C synthesis in dogs. J Am Vet Med Assoc
130483-494, 1957
17 OLSON,S.-E.: Radiology in veterinary pathology. Acta Radio1 [Suppl] (Stockh) 319255-
270, 1972
18 PEACH,R.: An electron optical study of experimental scurvy. J Ultrastruct Res 6579-590,
1962
19 RENDANO, V.T.; DUELAND, R.; SIFFERMAN, R.L.: Letter to editor, metaphyseal osteopathy:
(hypertrophic osteodystrophy). J Small Anim Pract 18679-683, 1977
20. RISER,W.H.; SHIRER, J.F.: Normal and abnormal growth of the distal foreleg in large and
giant dogs. J Am Vet Rad SOC650-64, 1965
21 ROSS,R.; BENDITT, E.P.: Wound healing and collagen formation. J Cell Biol 22:365-387,
1964
22 TEARE,J.A.; KROOK,L.; KALLFELZ, F.A.; HINTZ,H.F.: Ascorbic acid deficiency and
hypertrophic osteodystrophy in the dog. A rebuttal. Cornell Vet 69384-401, 1979
23 URIST,M.R.: Biological initiators of calcification. In:Biological Mineralization, ed. Zipkin,
pp. 757-805, John Wiley and Sons, New York, 1973
24 VAANANEN, M.; WIKMAN,L.: Scurvy as a cause of osteodystrophy. J Small Anim Pract
20491-500, 1979
25. VJLLANEUVA, A.R.: Methods of preparing and interpreting mineralized sections of bone.
In:Proceedings of the First Workshop on Bone Histomorphometry, ed. Jaworski, pp. 341-
353. Univ. of Ottawa Press, Ottawa, Canada, 1973
26 WATSON, A.D.J.; BLAIR,R.C.; FARROW, B.R.H.; BAIRD,J.D.; COOPER, H.L.: Hypertrophic
osteodystrophy in the dog. Aust Vet J 4 9 433-439, 1973

Request reprints from J.C. Woodard, Division of Comparative Pathology, Box J- 145,
JHMHC, University of Florida, Gainesville, FL, 32610 (USA).