Elbow Dysplasia

Originally, the term "elbow dysplasia" was introduced to describe a generalized osteoarthritis of the elbow joint in which an ununited anconeal process occurred in some but not in all of the affected joints [1]. Later, fragmentation of the medial coronoid process of the ulna and osteochondritis dissecans of the medial aspect of the humeral condyle were recognized to also be associated with this generalized osteoarthritis [2]. The modern concept of elbow dysplasia means a complex of polygenic hereditary developmental abnormalities of the elbow, including ununited anconeal process (UAP), fragmented medial coronoid process (FMCP), osteochondrosis of the medial humeral condyle (OCD), and elbow incongruity (EI) [3-5]. The term elbow dysplasia has been accepted to describe all conditions resulting in arthrosis of the elbow joint regardless of the underlying cause (1990, International Elbow Working Group meeting, San Francisco).

The same form of dysplasia may be found in both elbows or a combination of forms may be observed, such as FMCP and OCD [6-11]. Breeds may be predisposed to a particular form of elbow dysplasia. The German shepherd dog suffers more from UAP [12-14]. FMCP and incongruity are the most common findings in the Bernese mountain dog [15-16]. Rottweilers most frequently have FMCP but are rarely affected by OCD [6,8,11]. The Labrador retriever is most likely to have OCD and FMCP [11,17]. The suggestion that different patterns of disease exist within breeds or within families is extremely important in gaining a complete understanding of this complicated syndrome [18].

The pathogenesis of this disease remains poorly understood and controversial [4,19]. The finding of all forms of dysplasia within both elbows of a single dog, plus the finding of various forms of elbow dysplasia within the same elbow suggests a common etiology [18]. Incongruity exists in the different forms of elbow dysplasia [20]. Three major pathogenetic mechanisms proposed to explain the development of elbow dysplasia are asynchronous growth of the radius and ulna, osteochondrosis, and trochlear notch dysplasia.

Ununited Anconeal Process

The pathogenesis of UAP is controversial. The heritability in the German shepherd dog was reported early and conclusively [1,21,22]. The lesion in other breeds has not been so well studied; however, it is thought that it is probably inherited in these dogs as well [18].

Ununited anconeal process has been described as a manifestation of osteochondrosis.3 Osteochondrosis (OC) is disturbed endochondral ossification of articular or physeal cartilage [23,24]. The cartilage continues to grow and it becomes abnormally thick and obviously less resistant to mechanical stress [23,24]. In other cases, OC causes retarded growth or premature closure of a growth plate [24]. Retention of cartilage in the distal ulna may occasionally lead to retarded growth of the ulna; overgrowth of the radius with a relatively shorter ulna could be the cause of UAP (Fig. 102-1) [24,25]. The radius that is too long relative to the ulna forces the trochlea of the humerus in a proximal direction, which exerts more pressure than normal on the anconeal process, resulting in damage to the joint cartilage and to the ossification center of the anconeal process [3,25]. If OC is present, the entire structure is less resistant to trauma and a tear in the weakened cartilage prevents osseous bridging of the gap [3]. Measurements made of the olecranon in German shepherd dogs with UAP showed a significantly shorter olecranon in affected limbs [12]. The UAP may be completely separated or joined to the ulna by fibrous or fibrocartilaginous tissue [19]. In large-breed puppies, deformed anconeal processes with degenerated articular surfaces have lesions such as ischemia, degeneration, and multifocal dystrophic mineralization, which were apparently a result of pressure exerted by an upward thrust of the condylar surface of the humerus [4]. In a study on dogs with UAP this pressure on the anconeal process was eliminated by lengthening the ulna through osteotomy [19].

Figure 102-1. LCT scan of two dogs affected by UAP. Overgrowth of the radius with a relatively shorter ulna could be the cause of UAP (right) but is not systematically observed (left).

Although asynchronous growth of the radius and ulna may play a role in the pathogenesis of UAP, breeds of dogs having secondary centers of ossification at that location are most likely predisposed to develop that lesion. In the German shepherd dog, the separate ossification center of the anconeal process appears at 11 to 14 weeks and the anconeal process is united with the olecranon at 20 to 22 weeks [9]. In the greyhound, UAP is unknown; this is assumed to be because the anconeal process of the ulna ossifies more rapidly than that in the German shepherd dog [26]. Bony union is completed in approximately 2 to 3 weeks in the greyhound against 6 to 8 weeks in the German shepherd dog [26]. Surgical specimens of UAP from breeds with secondary centers of ossification of the anconeal process had a well formed articular surface of the proximal trochlear notch in which the articular cartilage was supported by an oriented cancellous bony base [4]. The fracture line travelled along what appeared to be a fenestrated closing growth plate. By contrast, surgical specimens of UAP from large-breed dogs that did not have separate secondary centers of ossification of this process were deformed, had less ordered cancellous bone architecture, and did not have the remnants of a closing growth plate along the base. The fracture lines of these specimens had the disordered appearance of a fatigue or stress fracture [4].

Fragmented Medial Coronoid Process – Osteochondritis Dissecans

Fragmented medial coronoid process is commonly seen with OCD or erosive lesions [2,7,11]. Nevertheless, the lesions can occur independently and one does not lead on to the other [27]. Olsson proposed that both FMCP and OCD are manifestations of OC and could result in overgrowth of the ulna [3,9]. No histologic evidence was found that the medial coronoid process of the ulna has a separate center of ossification [28]. The cartilaginous medial coronoid process ossifies from its base to the tip and ossification is completed at 20 to 22 weeks [9]. This delay in ossification of the medial coronoid process may predispose it to fragmentation when mechanical stresses are applied [25]. There is no step between the coronoid process and the radius during the stance phase of the gait cycle [29]. A study of the distribution of forces between the articular surface of the humerus and radius, and between the humerus and ulna in normal canine joints shows that the ulna is an important structure in the transfer of load. The ratio of the mean forces between the proximal articular surfaces of the radius and ulna remained close to a 50:50 distribution regardless of the applied load [30]. In human elbow joints the coronoid process appears to be an essential osseous block to prevent posterior subluxation [31].

In puppies up to 5 months of age, the bones would move past one another along the longitudinal axis over a distance of about 1 to 2 mm. In puppies older than 5 months, longitudinal movement of the radius and ulna was limited. No movement could be elicited in any of the specimens older than 6 to 7 months. This may be due in part to the natural slowing of longitudinal bone growth as skeletal maturity is approached, but it may also be due in part to the restraining effect on growth plates exerted by the maturing intraosseous membrane and the ligament that begins to mature and resist longitudinal movement at about this time [4].

Overgrowth of the ulna places an abnormal load on the medial coronoid process and the medial condyle of the humerus; FMCP, OCD, or both in combination could appear [3]. The resulting malalignment of the proximal ulna and radius caused much of the weight that was transferred from the humeral condyle to the antebrachium to be transmitted through the small, elevated medial coronoid process, leading to fragmentation or abnormal ossification of the coronoid process [3,29]. Owing to mechanical stress, fissures can result within the thickened cartilage of medial condyle that eventually leads to the formation of a cartilage flap or OCD lesion [25]. If the overgrowth of the ulna develops late, the development of the skeleton has reached a stage of near maturity that does not provide the ground for the appearance of FMCP and OCD. In that case, the increased load on the humeroulnar compartment of the elbow joint leads only to erosion, or "kissing lesion" that may extend into the underlying bone [3].

The split line reveals the prevailing alignment of collagen fibers in subchondral bone. An interrelation exists between development and differentiation of connective and supporting tissues and the load acting on them [32]. The orientation of the split line denotes the direction of maximal tensile strength. Three main types of split-line pattern could be differentiated and correspond well with the different fissure and fragmentation line pattern of the medial coronoid process [32]. These similarities may indicate an association between orientation of the split line and type of lesion [32].

Up to seven types of lesions of the medial coronoid process are described, with articular cartilage of the medial coronoid process at the same level as that of the radial head or elevated (Fig. 102-2) [33]. A decrease of the height of the step between the articular surface of the coronoid and radial head is obtained after a proximal ulna sliding osteotomy [33,34]. The damage to the medial coronoid process varies widely. It may remain cartilaginous, have a delayed ossification, or may ossify at a more normal time. The lesion may appear as a fissure within the cartilage with complete separation of the cartilage fragment or it may appear as a fissure within the bone, with an incomplete fracture or complete separation of a bony fragment [18]. A FMCP might be separated from the fracture bed and remain viable because it is loosely attached to the joint capsule from which it continues to maintain its blood supply. Thus, if displaced, the fragments may remain unchanged or, because of a persistent blood supply, the fragments of cartilage may become larger and may ossify or fragments of bone may enlarge.9 The OCD lesion lies opposite the medial coronoid process. The fragment may remain in situ. If it is not removed surgically, it gradually becomes free but rarely mineralizes; it may detach and form a joint mouse [9].

Figure 102-2. CT scan of two 2 dogs affected by FMCP. The medial coronoid process could be at the same level as that of the radial head (left) or could be elevated (right).

Medial compartment osteoarthritis with eburnation of the medial humeral condyle is currently diagnosed by arthroscopy. The canine medial compartment osteoarthritis is similar to medial compartment osteoarthritis of the human knee [35]. Research shows an association of the medial compartment osteoarthritis with a varus deformity of the elbow similar to what is described in the knee in people. The humeroulnar angle in normal dogs measured on craniocaudal radiographs varies from 0.53° to 9.09° [35]. Dogs with medial compartment osteoarthritis have a varus deformity with this angle greater than 7° [36].

The cartilaginous growth disturbance is likely to have genetic and environmental, and mainly nutritional and traumatic causes [23,25]. However, the theory that the lesions are caused by OC is not supported by consistent findings of histologically defective cartilage at early lesion sites. Histologic evidence of thickened degenerated cartilage is reported in less than 20% of 120 dogs with elbow dysplasia [37]. In a multiyear necropsy study, histologic lesions consistent with OC lesions are observed in elbows from 3- to 5-month-old puppies [4]. Both the medial humeral condyle and medial coronoid process have compressed the subchondral capillary bed, disrupted subchondral spongiosa, and focally thickened cartilage lesions. In puppies older than 5 months of age, focal lesions on the medial humeral condyle were roughened and fibrillated but had an intact articular surface. Although the lesions in these older puppies were located at the same site on the medial condyle as were the lesions in the younger puppies, they more closely resembled osteoarthritis than OC [4]. Differences in the histologic and ultrastructural appearance of the OCD and FMCP lesions tend to suggest that they are two separate disease entities. It is hypothesized that OCD results from incorrect cartilage maturation and endochondral ossification, and the etiology of FMCP is subchondral fracture with an ineffective fibrous repair in some cases [27].

Joint Incongruity

The normal canine joint showed a continuous arc formed by the central ridge of the trochlear notch and the proximal radial articulation [20]. Specific areas of articular contact were identified on the radius, the craniolateral aspect of the anconeal process, and the medial coronoid process [29]. The medial coronoid and radial contact areas were continuous across the radioulnar articulation [29]. The incongruity observed between the radius and the ulna suggests either a lagging growth rate of the radius or an ulnar trochlear notch of smaller diameter.

The most obvious abnormality was the finding that the distal edge of the trochlear notch of the ulna with its medial coronoid process lay approximately 2 mm proximally to the articular surface of the radius [20]. Measurements of radial and ulnar length have indicated a statistically significant relative increase in length of the proximal ulna in middle-size and large-breed dogs during a vital period of limb development (before 16 to 20 weeks of age). The ulna was temporarily up to 3 mm longer than the paired radius.38 Incongruity may not always be evident owing to compensatory adjustment during growth [38]. Examples of bilateral FMCP were found in which a 1- to 2-mm elevation of the medial coronoid process above the radial articular surface remained in one elbow joint [4]. This degree of incongruity, although compatible with a lagging radial growth, is in itself insufficient to explain the relative increase in the length of the proximal ulna.

Disparity in radial and ulnar lengths were noted in both normal and lame dogs and failed to correlate with obvious radiographic joint incongruity [39]. This indicates that variation in the length of these bones may be a normal variation of growth and a common finding in large-breed dogs. Only 5 of 15 lame dogs demonstrated obvious disparity in radial and ulnar length, although it is possible that joint incongruity might have temporarily occurred during growth yet was not apparent at presentation [39].

Another explanation for the incongruity between the radius and ulna is an abnormal development of the ulnar trochlear notch, resulting in a slightly elliptic joint surface with an arc of curvature insufficient to encompass the humeral trochlea. This creates a joint with major contact points in the areas of the anconeal process and medial coronoid process but without a point of contact between the trochlear notch and the humeral trochlea (Fig. 102-3) [20]. Underdevelopment of the trochlear notch may be temporary or persistent and becomes evident between the ages of 4 to 6 months [20]. If the incongruity is present at the age of 4 to 6 months, the skeleton at that time is as yet incompletely ossified. Fracture, fragmentation, or fissure of the medial coronoid process could occur owing to the increased weight-bearing forces on the too-high lying parts of the ulna [20]. For the same reason, the lesion could be an OCD lesion [20]. The humeral condyle is pressed against the anconeal process and causes micromovement of the cartilage bridge and partial to complete separation of the articular process [20,38]. If the incongruity is present after 6 months, it may be present on its own and interferes with the formation of the articular surfaces and the manner in which they articulate; consequently, with time, osteoarthritis develops as a secondary joint disease [20]. Nevertheless, the grade of arthrosis is not always higher in the older dog than in the young dog. Dogs over 5 years of age were identified suffering from FMCP even in the absence of osteophyte formation [40]. It is possible that a different etiopathogenesis such as traumatic fracture or persistent physeal cartilage may have been involved [40].

Figure 102-3. An abnormal development of the ulnar trochlear notch with an arc of curvature insufficient to encompass the humeral trochlea creates a joint with major contact points in the areas of the anconeal process and medial coronoid process.

The increase in the relative size of the proximal ulna in heavier breeds could be necessary to accommodate a larger trochlear notch for articulation with a heavier humerus [38]. This need for rapid growth may also be a contributing factor for the failure of the trochlear notch to reach the required arc of curvature to encompass the humeral trochlea [38].

In severe cases, it is obvious that the increased load on the medial coronoid process has led to remodeling of the ulnar part of the elbow joint, leading to a more open trochlear notch [3]. This observation however does not support the hypothesis that an elliptic trochlear notch makes it too small for the trochlea of the humerus. Moreover, the hypothesis of an abnormally developed, slightly elliptical trochlear notch seems unlikely to be a major contributing factor, as combined UAP and FMCP should then occur in much larger numbers of dogs and ulnar ostectomy would have no effects in cases with UAP [3,19].

One study compared the radius of curvature of the ulnar trochlear notch of the Rottweiler, a breed predisposed to FMCP, to that of greyhound, a breed that has never reported be affected by this disease [41]. The greater radius of curvature within the distal part of the trochlear notch in the Rottweiler results in a more open elliptical shape. It may result in dynamic or positional instability of the cubital joint, increasing stress leading to FMCP. It may also be formed as a result of asynchronous growth of the radius and ulna [41].

Heritability

Elbow dysplasia is inherited as multifactorial or polygenic traits [5,42,43]. Osteochondritis dissecans and FMCP were found to be inherited independently as polygenic traits [17]. Fragmented coronoid process and incongruity were found to be genetically separate diseases [16]. If elbow dysplasia is the phenotypic effect of independent genetic diseases, the phenotype may behave polygenetically even though the underlying components are monogenous [16]. The incidence of elbow dysplasia was found to be strongly breed-specific statistically and was higher in intermediate and heavy-set breeds when compared with sight hounds and setter-type breeds [38]. The exact nature of the environmental contribution to the disease remains unclear, although several putative factors have been put forward such as nutrition and exercise [5,16].

Males generally are more frequently affected [7,10,11,42,44]. Heritability in Labrador retrievers was 0.77 for males and 0.45 for females [45]. In Rottweilers, estimation of heritability for elbow dysplasia was 0.34 +/- 0.04.46 Reported incidence for this breed in Norway was 40% [8]. In Bernese mountain dogs and Rottweilers, heritability based on regression of sons on sires was 0.34, and the estimated heritability based on regression of daughters on dams was 0.28 and 0.4 respectively [43]. An inherited basis for elbow osteochondrosis has also been reported in the Labrador retriever with a greater contribution of the dam suggesting a "maternal effect" [10]. The differences in the prevalence can be explained by a direct effect of genes on the sex chromosomes or by an effect of secondary sex characteristics such as differences in growth rate, sex hormones, or behavioral patterns. For a polygenic trait such as arthrosis in the elbow joint, the effect of secondary sex characteristics is a more plausible explanation [43].

Conclusion

Elbow dysplasia describes all conditions (UAP, FMCP, OCD, EI) resulting in arthrosis of the elbow regardless of the underlying cause. A genetic predisposition is widely recognized. All forms of elbow dysplasia were initially thought to be secondary to osteochondrosis but the research of the last 20 years using histology, traumatic tactile pressure sensor, computed tomography, and subchondral split-line patterns clearly indicates a biomechanical basis as the major pathophysiologic mechanism of UAP, FMCP, and EI. Differences in the pathology and ultrastructural appearance of the OCD and FMCP lesions tend to suggest that they are two separate diseases. The proximal joint surface of the ulna contributes substantially to load transfer through the canine elbow joint. Anomalies that increase this load appear to contribute to canine elbow dysplasia, mainly fragmentation of the medial coronoid process. Because of the great interest in this common invalidating pathology and the modern tools of investigation, we share a better understanding on how elbow dysplasia might appear, but more research is needed to refine the latest knowledge and to design prevention techniques for elbow osteoarthritis.