Post natal development of the canine elbow joint: a light and electron microscopical study

Research in Veterinary Science 1992, 52, 67-71

Post natal development of the canine elbow joint: a light and electron microscopical study S. GUTHRIE, Department of Small Animal Medicine and Surgery, Royal Veterinary College, Hawkshead Lane, North Mymms, Hertfordshire AL9 7TA, J. M. PLUMMER, Department of

Veterinary Basic Sciences, Royal Veterinary College, Royal College Street, London NWI OTU, L. C. VAUGHAN, Department of Small Animal Medicine and Surgery, Royal Veterinary College,

Hawkshead Lane, North Mymms, Hertfordshire AL9 7TA

The elbows of 13 puppy cadavers were dissected, samples were taken for light and electron microscopy, and the thickness of the articular cartilage of the distal humerus and proximal ulna was measured. Throughout post natal development differences were found in the arrangement of the growth plate and articular chondrocytes. At birth, the articular surface had remnants of a fibrous limiting membrane that was continuous with the perichondrium, a finding not previously recorded in dogs. Orientation of the collagen fibrils within the matrix of the articular cartilage was initially lacking but became established by three weeks. In the humerus cartilage canals were present up to 12 weeks old. The articular cartilage of the humeral condyle varied in thickness across the joint surface, being thicker on the medial than on the lateral side; it was also thicker at the apex of the medial coronoid process. These regions of thick cartilage correspond with the sites where cartilage defects arise in elbow osteochondrosis. No histological evidence was found that the medial coronoid process of the ulna is a separate centre of ossification.

THE canine elbow joint is a complex structure in terms of its bone and soft tissue components, comprising three articulations (humero-radial, humero-ulnar and radio-ulnar), four ligaments (medial and lateral collaterals, oblique and annular) and several tendons. To maintain correct function it is imperative that during growth the congruency between the articular surfaces of the elbow joint is maintained. During bone growth, an increase in diaphyseal length is achieved by the orderly division and maturation of chondrocytes across cartilaginous

growth plates (Banks 1986) followed by the calcification of matrix and bone formation (endochondral ossification). Growth of the epiphysis occurs by a similar process of cell division in which the articular cartilage can be thought of as the 'growth plate' of the joint surface. Articular cartilage consists of chondrocytes embedded in matrix, moving inwards from the surface, and four distinct regions can be identified (Ghadially 1981). Zone 1, the most superficial, is the zone of surface or resting cells, zone 2 consists of groups of proliferating cells, in zone 3 the chondrocytes mature and hypertrophy and in zone 4 the matrix becomes calcified. Mineralisation is initiated in matrix vesicles which bud from cell processes (Ali 1983) and are rich in alkaline phosphatase (Glaser and Conrad 1981). The junction between calcified and uncalcified cartilage often appears as a basophilic tidemark in sections stained with haematoxylin and eosin. The type 2 collagen fibrils within the matrix are arranged in arcades (Barnett and others 1961) although those in the epiphyses of babies have been found to lack this orientation (Cameron and Robinson 1958). Necrotic ch0ndrocytes occur normally in cartilage (Ghadiall3, 1982) and can be readily identified by transmission electron microscopy because they have pyknotic nuclei with condensed chromatin, a reduced number of organelles and disrupted limiting membranes. A number of developmental abnormalities may affect the canine elbow joint and one currently causing much concern is 'osteochondrosis'. The disease, appears to result from the disruption of chondrocyte maturation at specific sites within the joint. The lesions affect the medial articular 67

68

S. Guthrie, J. M. Plummer, L. C. Vaughan

surface of the distal humeral condyle with the formation of loose cartilage flaps and erosions (Olsson 1976), and the medial coronoid process of the ulna, which often separates or fragments. Specimens from puppy cadavers were studied by light and electron microscopy to establish the normal pattern of development of the elbow and to discover whether the coronoid process is a separate centre of ossification. The thickness of the articular cartilage was measured across the joint surfaces of the humerus and ulna to determine whether the predilection sites for osteochondrosis-type lesions correspond with regions of thicker cartilage, because such regions may be more prone to ischaemic necrosis. Materials and methods

Thirteen puppy cadavers were available. They were eight labrador retrievers (one, two, three, four, eight, 12, 16 and 20 weeks old), a newborn beagle, two greyhounds (five and seven weeks old), a six-week-old rough collie and a 10-monthold Yorkshire terrier. After dissection of the elbow joint s to establish that there were no gross abnormalities of the articular surfaces, tissues were taken for histological preparation.

Light microscopy Blocks of cartilage and bone were cut with a hacksaw to include the medial coronoid process of the ulna and the humeral condyle. After fixation for 48 hours in buffered formol saline and decalcification in formic acid, transverse sections were cut from the middle third of the coronoid process and the humeral condyle. The sections were stained with haematoxylin and eosin or toluidine blue, using standard techniques.

tions were mounted on square grids and stained with uranyl acetate/lead citrate before examination in a JZOL 1200EX electron microscope.

Cartilage thickness Serial transverse sections of the entire coronoid process were cut 200 gm thick with an annular saw (Leitz 1600; Leitz UK) then fixed and stored in 70 per cent alcohol. The humeral condyle was similarly sectioned in the sagittal plane. A modified alcian blue/alizarin red S stain was used (Kelly and Bryden 1983) to differentiate cartilage and bone, the calcified tissues staining magenta and other tissues staining brilliant sky blue. The low power microscope image of each section was projected onto a digitiser tablet (Bit Pad Two 1103; Summagraphics) and an image analysis system (Digit image analysis; Hayes and Fitzke 1987) was used to measure the width of the distal humeral (medial and lateral sides of the condyle) and coronoid articular cartilage.

Results

Newborn beagle The humeral condyle, the coronoid and anconeal processes and the olecranon were entirely cartilaginous. The articular surface of both the coronoid process and the humerus were covered by a thin layer of fibrous tissue that was loosely attached in some areas and was continuous with the perichondrium (Fig 1). A cartilage canal was present towards the lateral edge of the coronoid process,

Electron microscopy For each specimen, a fresh sample of humeral articular cartilage 1 mm thick was taken from the medial side of the condyle and prefixed in 3 per cent glutaraldehyde in phosphate buffer for a minimum of three hours (usually overnight). After washing in buffer and post-fixation in 1 per cent osmium tetroxide the samples were dehydrated, cleared in propylene oxide and embedded in resin. One gm sections were cut for orientation and stained with toluidine blue. Ultrathin (90 nm) sec-

FIG 1 : Transverse section of distal humerus of newborn beagle. The articular surface is covered by a fibrous layer (F)which has separated in places on sectioning, but is continuous with the perichondrium.( Haematoxylin and eosin × 9

Post natal development of the canine elbow

69

TABLE 1 : Thickness of cartilage in dogs of different ages Coronoid

Humerus

Age of dog (weeks)

Lateral (mm)

Medial (mm)

Lateral (mm)

Medial (mm)

Newborn 1 2 3 4 5 6 7 8 12 16 20

1.2 1.1 1-2 0.94 2.6 1.7 1.2 3.8 1"1 1"6 0"94 0'5

0.57 0.40 0-44 0.43 0.60 1.1 0.93 1-18 0"78 1"8 0"55 0"57

2.4 4.1 1.3 2.1 2.3 1.24 1.7 1'8 0'8 0"8 0"4

2.7 4.6 4-8 3.2 3.6 1.8 2.1 2"0 0'83 0"80 0"60

FIG 3: Distal humerus of a one-week-old labrador retriever. Cartilage canal within zone 2. Haematoxylin and eosin x 23

where the cartilage was thickest (Table 1) and several canals were present in the humeral condyle. One-week old labrador retriever

The humeral condyle was entirely cartilaginous but bone extended slightly into the coronoid process (Fig 2). The humeral cartilage had a very organised structure and cartilage canals were present (Fig 3). Superficial zone 1 chondrocytes had a normal appearance and those of zones 2 and 3 appeared metabolically active, although there was minimal matrix organisation, the collagen fibrils having a random orientation. The uncalcified cartilage of the humerus was thicker medially (Table l). The coronoid cartilage varied in thickness across the articular surface, and was thickest at the apex. Two-week-oM labrador retriever

Two zones of hypertrophied chondrocytes were

FIG 4: Distal humerus of a two-week-old labrador retriever. Large cartilage canal (c) lined with epithelium, separating two regions of hypertrophiedchondrocytes (h) that are presumptive secondarycentres of ossification. Haematoxylin and eosin x 23

present in the humerus (Fig 4). Interterritorial and territorial matrix could be differentiated. Degenerate chondrocytes were occasionally seen in zone 3. Three-week-oM labrador retriever

cartilage

FIG 2: Line drawing taken from a radiograph of the elbow joint of a one-week-old labrador retriever. The humeral condyle (c) was entirely cartilaginous but a bony prominence (b) extended into the coronoid process

A secondary centre of ossification was present in the lateral side of the humeral condyle but no mineralisation had occurred on the medial side. Collagen fibrils were orientated parallel to the articular surface in zone 1 and then turned and were predominantly longitudinally orientated. The coronoid cartilage was thickest at the apex and the humeral cartilage became thinner from the centre to the side of the articular surface. Four-week-oM labrador retriever

Secondary centres had developed in the lateral

70

S. Guthrie, J. M. Plummer, L. C. Vaughan

and medial sides of the humeral condyle and the medial epicondyle. The cartilage:bone interface at the base of the coronoid consisted of short chondrocyte columns that were not so closely apposed as those in the humerus. Five-week-old greyhound

A secondary centre of ossification had developed in the olecranon. Cartilage canals were still present. Six-week-old rough collie

In the centre of the humeral condyle, where there was no secondary centre, the cartilage was 6.7 mm thick. Seven-week-old greyhound

The radial head had developed a secondary centre of ossification. Hypertrophied chondrocytes in the coronoid cartilage formed clusters rather than long columns (Fig 5).

Sixteen-week-old labrador retriever

The anconeal growth plate had closed radiographically. The cartilage was generally less cellular than it was in the younger dogs. Twenty-week-old labrador retriever

The anconeal and intracondylar growth plates were histologically dosed but a narrow band of cartilage still separated the medial epicondyle from the condyle. Ten-month-oM Yorkshire terrier

The articular surfaces were smooth and the cartilage had an irregular tidemark. Chondrocytes of all four zones could readily be identified. The cartilage had an ordered structure but the long columns of hypertrophied cells, typical of the growth plate, had not formed. The chondrocytes were metabolically active, with well developed pericellular capsules. Collagen fibres with type 2 banding were abundant in the matrix and were orientated predominantly at right angles to the joint surface. Discussion

FIG 5: Coronoid process of a seven-week-old greyhound, Clusters of chondrocytes at the base of the process and at the cartilage: bone interface (b Bone, c Cartilage). Haematoxylin and eosin × 45

Eight-week-old labrador retriever

A secondary centre was demonstrable in the anconeal process. No cartilage canals were found in the coronoid. Toluidine blue stained the cartilage metachromatically. Electron microscopy revealed chondrocytes which appeared metabolically active and morphologically typical of their respective zones. Twelve-week-old labrador retriever

No cartilage canals were seen.

After birth the secondary centres of ossification appeared in the following order: lateral side of the humeral condyle, medial side, medial epicondyle, olecranon, radial head and anconeal process. The last centre was present for only two months, its growth plate being the first to close. Presumptive secondary centres initially appeared as zones of hypertrophied chondrocytes. An important result of the study was that no histological evidence was found at any age to suggest that the medial coronoid process of the ulna is a separate centre of ossification. A fibrous limiting membrane which covered the articular cartilage in the newborn beagle has not been previously recorded in dogs. Classically, the process of endochondral ossification is described at the growth plates of the long bones. Although a similar process occurs below the articular cartilage there are obvious histological differences between the coronoid process, the humeral articular surface and the humeral growth plate: In the deeper zones of the humeral and coronoid articular cartilage, the chondrocytes are less densely packed than at the growth plate and

Post natal development of the canine elbow

form clusters of three or four cells rather than long columns. The region of proliferating chondrocytes (zone 2) is thicker than that in the growth plate, but zone 3 is reduced. The trabeculae of the metaphysis are broader than those of the epiphysis and are separated by larger marrow cavities. It may be hypothesised that the closer arrangement of trabeculae in the subchondral bone plate affords better support to the articular cartilage than would be provided by a coarser arrangement. Cartilage canals were more prevalent and larger in the humerus than in the coronoid process. The canals may persist longer in the humeral cartilage because it is thicker than the coronoid cartilage. The lack of orientation of the collagen in the newborn beagle supports the theory that collagen orientation is a dynamic functional adaptation to strain (Riggs and others 1990). During development the thickest cartilage occurred on the distal humerus, before the secondary centres of ossification developed. At this stage the cartilage was presumably deriving its nutrients from vessels within canals because diffusion from the synovial fluid would probably have been inadequate. Cartilage has been shown to increase in thickness in areas where there is a greater load (Muller-Gerbl and others 1987) and thus it may be hypothesised that the medial side of the humerus and the lateral edge of the coronoid are subjected to most stress. Further evidence for this hypothesis could be obtained by determining the stress pattern across the articular surfaces of the elbow during weight bearing, and by examining whether the predilection sites for osteochondrosis correspond to areas where stress is concentrated. Labrador retrievers have a predisposition to elbow osteochondrosis and it may therefore be argued that they are not the ideal breed to use when determining 'normality' for the canine elbow. Obtaining a wide variety of suitable material, however, proved to be difficult. The labrador retrievers used in this study were from osteochondrosis-free stock and no gross elbow lesions were found during the dissection of the joints. When compared with the dogs of other breeds at the same age, their elbow development proceeded along similar lines.

71

The sites at which the lesions of osteochondrosis usually occur corresponded with regions of thicker cartilage during development. However, no evidence was found that these regions of thicker cartilage were subject to ischaemic necrosis. A flaw in growth plate fusion cannot be the reason for the separation of the coronoid process, because this process is not a separate centre of ossification.

Acknowledgements This work was funded by the Guide Dogs for the Blind Association. Illustrative and technical assistance was given by Hilary Orpet, John Bredl, Geraldine Eyre, Caroline Edwards, David Gunn and Sue Evans.

References ALI, S.Y. (1983) Cartilage. Vol 1. Structure, function and biochemistry. Ed B. K. Hall. New York, Academic Press. pp 345-364 BANKS, W. J. (1986) Applied Veterinary Histology. 2nd edn. Baltimore, Williams & Wilkins. pp 108-118 BARNETT, C. H., DAVIES, D. V. & MACCONNAILL, M. A. (1961) Synovial Joints. Their Structure and Mechanics. London, Longman, Green. p 304 CAMERON, D. A. & ROBINSON, R. A. (1958) Electron microscopy of epiphyseal and articular cartilage matrix in the femur of the newborn infant. Journal of Bone and Joint Surgery 40-A, 163-170 GHADIALLY, F. N. (1981) Electron Microscopy in Human Medicine, Vol 4. Soft Tissue, Bones and Joints. Ed J.V, Johannessen. New York, McGraw-Hill. pp 3-4 GHADIALLY, F. N. (1982) Ultrastructural Pathology of the Cell and Matrix. 2nd edn. London, Butterworths. pp 16-18 GLASER, J. H. & CONRAD, H. E. (1981) The formation of matrix vesicles by cultured chick embryo chondrocytes. Journal of Biological Chemistry 256, 12607-12611 HAYES, B, & FITZKE, F. (1987) Digit Image Analysis for the Archimedes and Bit Pad 2. London, Institute of Ophthalmology. pp 1-3 KELLY, W. L. & BRYDEN, M. M. (1983) A modified differential stain for cartilage and bone in whole mount preparation of mammalian fetuses and small vertebrates. Stain Technology 58, 131-134 MULLER-GERBL, M., SCHULTE, E. & PUTZ, R. (1987) The thickness of the calcified layer of articular cartilage: a function of the load supported? Journal of Anatomy 154, 103-111 OLSSON, S-E. (1976) Osteochondrosis - a growing problem to dog breeders. Gaines Progress. pp 1-11 RIGGS, C. M., EVANS, G. P., SHAH, D., BONFIELD, W., LANYON, L. E., VAUGHAN, L. C. & BOYDE, A. (1990) Functional associations in cortical bone between collagen orientation, mechanical properties and in vivo strain history. Proceedings of the 7th meeting of the European Society of Biomechanics, Aarhus, Denmark. p 85

Received June 25, 1991 Accepted August 8, 1991