Portal de Periódicos da CAPES

Working Algorithm for Treatment Decision Making for Developmental Disease of the Medial Compartment of the Elbow in Dogs

2009; Wiley; Volume: 38; Issue: 2 Linguagem: Inglês

10.1111/j.1532-950x.2008.00495.x

1532-950X

Noel Fitzpatrick, Russell Yeadon,

Tendon Structure and Treatment

THERE IS disagreement about the manifestations of elbow pathology that should be included under the umbrella of elbow dysplasia. This is highlighted by variable inclusion or exclusion of diseases like ununited medial epicondyle1 and elbow incongruity2 with the more commonly recognized and historically grouped triad of (1) disease of the medial aspect of the coronoid process (medial coronoid disease; MCD); (2) osteochondrosis (OC) or osteochondritis dissecans (OCD) of the medial humeral condyle; and (3) ununited anconeal process (UAP). Variations in grouping syndromes within elbow dysplasia contributes to confusion among veterinarians and dog owners about the precise nature of disease involved. Whereas several diseases may coexist within the same joint,3–5 it has become increasingly apparent from histomorphometric,6 biomechanical,2,7,8 and genetic or heritability9,10 data that there is considerable independence in development of these multifactorial disease processes. This is further complicated by the spectrum of clinical signs and macroscopic pathology associated with any single disease process,11 which has important implications for treatment and prognosis. Clearly, there is no single treatment for all recognized manifestations of elbow dysplasia, so seemingly, the initial step in developing a clinically useful decision-making algorithm is to deconstruct this oversimplified nomenclature. A working understanding of the etiopathogenesis of these disease processes is needed to define and optimize treatments for a corresponding spectrum of pathologic change. Failure to understand and address underlying processes when treating overt pathology will almost inevitably lead to suboptimal outcome. Despite this lack of understanding, there is increasing recognition that joint incongruency is an important factor in the development of canine elbow disease, albeit in a more complex manner than initially proposed.2,11 Concurrently, it is important to understand the utility and limitations of diagnostic methods (e.g. historical, clinical, imaging) for identification of pertinent disease characteristics. To the extent that these diagnostic approaches define appreciation of disease and thus direct treatment approaches, their precision has potentially profound implications for treatment outcomes. Finally, relevant outcome measures are critical to evaluate the efficacy of treatment approaches for individuals and across patient groups. Outcome measures provide a framework for accumulation of clinical evidence and application in the context of "evidence-based practice,"12,13 and allow continued refinement of assumptions leading to progressive improvements in therapeutic strategies. Thus, in considering this larger challenge—deconstructing currently used terminology, defining the etiopathogenesis of disease processes included in elbow dysplasia, improving diagnostic tests, and developing therapeutic strategies and outcome measures—we report the decision-making algorithms currently in use in our practice for disorders involving the medial compartment of the canine elbow. These algorithms have been developed from synthesis of current scientific foundations for this constellation of disorders, and interpreted and modified based on our clinical experience. Our purpose is to stimulate dialog, highlight areas of needed investigation, and provoke continued refinement of our collective understanding through robust clinical trials and focused studies of mechanisms and pathogenesis of elbow disease. We recognize that our efforts in this and our accompanying reports might generate controversy but anticipate that the framework we provide can serve as the foundation for future improvements. Even when free fragmentation has been subsequently identified by arthroscopy or arthrotomy, identification of disease of the medial aspect of the coronoid process (MCD) has historically proven challenging despite use of diverse imaging modalities including radiography2,14–17; computed tomography (CT)11,18; scintigraphy19; and magnetic resonance imaging (MRI).20 Typically, diagnosis is based on identification of secondary markers of degenerative joint disease in the absence of other discernable discrete pathologic changes. We consider thorough clinical evaluation to be as important as subsequent diagnostic imaging techniques. Recognition of thoracic limb lameness can be relatively insensitive because of the high incidence of bilateral disease and because many younger dogs remain surprisingly mobile or behaviorally active despite severe elbow pathology. We find that evaluation of discomfort caused by elbow manipulation is a strong indicator of disease involving the medial compartment of the canine elbow. The most reproducible discomfort can be elicited on maximal flexion, and particularly on firm supination of the antebrachium while the elbow is held in moderate flexion. Response to deep digital pressure in the region of the insertion of the bicep brachii muscle over the medial aspect of the coronoid process is also a valuable indicator of potential disease. Positive responses to these tests alone, in the absence of any other identified source of lameness or pain, creates a high index of suspicion such that diagnostic imaging is directed toward identification of disease of the medial compartment of the elbow and more specifically MCD. If results of noninvasive imaging techniques are equivocal, we consider direct observation by arthroscopy or arthrotomy to be justified. Detailed radiographic examination of both elbows is essential before any intervention. This allows more detailed classification of the nature of disease (specifically, likely chronicity, radiographic severity of subchondral disease, and presence of OC lesions) that may be valuable for further decision-making, rather than simply to confirm a diagnosis of MCD. This is of particular significance because elbows affected by MCD can be radiographically normal21 or may be affected by only subtle subtrochlear sclerosis.22 Furthermore, degree of osteophytosis is considered of minimal benefit in determination of an appropriate treatment protocol as it is poorly correlated with pathology subsequently identified by direct joint inspection23,24 but has close associations with expression of cartilage genes.25 Other imaging techniques or diagnostic tests useful for exclusion of potential primary or concomitant pathologies (e.g. shoulder lesions, musculotendinous injuries, neurogenic lameness, or neoplasia) are selected based on clinical findings. In particular, diseases involving the shoulder joint and adjacent structures should be excluded as causes of thoracic limb lameness.26 CT and more recently MRI scans are increasingly used to evaluate elbows where changes are equivocal on radiographic assessment27; however, arthroscopic evaluation constitutes the most important single interrogation directing our decision-making process for MCD with or without pathology of the medial aspect of the humeral condyle. We have observed equivocal findings on both CT and MRI where subchondral pathology was subsequently proven by histologic examination of coronoid specimens after arthroscopically identified MCD.28 A spectrum of arthroscopic changes including fragmentation, cartilage fissures, chondromalacia, and cartilage fibrillation affect the medial coronoid process (MCP) and we consider these to be manifestations of a common underlying disease process.24,29 Pathologic changes initially affect subchondral bone with formation of microcracks, characteristic of local fatigue failure.6 These changes typically occur at one of two anatomic sites, either the craniodistal tip of the medial aspect of the coronoid process, or more axially oriented in the region of the radial incisure. Although the precise nature of this fatigue phenomenon remains elusive, several hypotheses encompass the disparate range of recognized pathologic changes, all of which may be attributable to some form of humeroulnar conflict (HUC). Static Radioulnar Length Disparity. Characterized by shortening of the radius relative to the ulna. This disparity effectively exposes the coronoid process of the ulna to increased load-bearing forces from the humerus throughout the full range of elbow movement. MCP disease has been identified by CT and arthroscopy in the absence of overt static incongruity,27 so it seems unlikely that static incongruency plays a major role in some dogs,11 although where definitively identified, MCD should be considered a likely sequel. This form of incongruency might be expected to result in pathologic change focused at the tip of the medial aspect of the coronoid process but further biomechanical modeling is needed to validate this hypothesis. Dynamic Radio-Ulnar Longitudinal Incongruency. Characterized by relative distal displacement of the proximal radial articular surface compared with the proximal aspect of the ulna during weight bearing or in certain positions within the range of motion of the elbow. This may account for the lack of reliable identification of "static" incongruity using conventional imaging techniques, and could result in a similar configuration of pathology to that described for static radioulnar length disparity. From an anatomic perspective, this hypothesis requires relative laxity between the radius and ulna and is thus considered less likely. Incongruency Associated with the Shape of the Ulnar Trochlear Notch. Characterized relative to either the articular contours of the radial head or humeral condyle, or both, this abnormality could also result in focally increased loading forces. These could be limited to specific angles of joint flexion or extension, and would most likely be focused in the region of the tip of the medial aspect of the coronoid process. Some breed variation in radiographic shape of the ulnar trochlear notch has been reported30 but association of this observation with clinical disease is unknown/uncertain.8 Furthermore, techniques to standardize radiographic interpretation of conformational variation in trochlear notch shape have not been reported. It is likely that three-dimensional imaging modalities (CT, MRI) will be more sensitive to subtle conformation variations within the complex structure of the elbow joint. Further imaging and biomechanical investigation is needed to clarify the potential role of ulnar trochlear notch variation in the pathogenesis of MCD. Primary Rotational Instability of the Radius and Ulna Relative to the Distal Aspect of the Humerus. From a theoretical perspective, osseous conformation discrepancies or perhaps ligamentous insufficiency, could result in instability and incongruence through part, or all, of the range of motion of the elbow joint. Angulation of the humeroulnar joint surface relative to the long axes of the component bones could effectively convert a lateral shear force into a compressive force, causing crushing of the medial aspect of the coronoid process between the radial head and the medial humeral condyle. Another possibility could be mismatch between the contour arc of the radial incisure of the MCP and the radial head. These forms of incongruency would hypothetically most likely result in pathologic changes focused in the region of the radial incisure and could partially account for clinically observed variation in distribution of pathologic changes. Musculotendinous Mismatch. Rotational instability could occur as a secondary effect of musculotendinous mismatch in relation to the bony anatomy, in particular, disparity between muscle tensions generated during supination and pronation of the antebrachium relative to the humerus.31–35 The biceps brachii/brachialis muscle complex has a large fan-shaped insertion onto the abaxial portion of the MCP and a smaller but equally robust insertion on the proximal radius. Mismatch in synchronized development of the osseous and/or musculotendinous components of the elbow joint may contribute to supra-physiologic overload of the MCP thorough this fan-shaped insertion. The biceps brachii/brachialis muscle complex has the potential to produce substantial loading forces as a major elbow flexor. In the racing greyhound the estimated combined joint moment of force produced by this muscle complex at the elbow and acting in flexion is 1990 Ncm.36 Because its distal insertion is eccentrically located medially, a susbstantial component of this force in flexion could be converted into a supination force. As illustrated for the hypothesis on primary rotational instability, such aberrant force between the radial head and the radial incisure of the MCP could give rise to shear planes because of compression of the region of the radial incisure against the radial head and could give rise to the clinically observed pattern of arcuate fragmentation radiating out from the radial surface of the MCP.34,35 Progression of subchondral microcrack formation to visible cartilage fissuring or fragmentation by coalescence, appears variable between dogs and between different regions of the coronoid process. This may reflect variable patterns of biomechanical overload as described earlier. Importantly, subchondral microcrack formation extends beyond the area of superficial or macroscopic disease.6 The pattern of visible pathologic change may be dependent on a balance between rate of microcrack formation and healing by fibrous infilling, and on the focal intensity of supraphysiologic forces, which could theoretically radiate outward from a central focus akin to fissure lines created by movement of tectonic plates. Thus shear planes might occur either at the tip, or at the radial incisure, of the MCP depending on the prevalent loading force. Fragmentation represents one potential end stage of MCD (Fig 1), whereby the medial aspect of the coronoid process cannot be salvaged to provide a functional, pain-free, load-bearing surface. Age may be an important factor with potential repetitive overload of fissure zones or traumatic disruption of fissure zones leading to fragment formation—so-called "jump-down" lesions.37 In Danielson's study,6 medial coronoid specimens from all age groups had diffuse microcrack formation. This finding lends credence to the concept that sudden fragmentation, even in mature dogs is preceded by microcrack formation. Arthroscopic image showing large fragment at radial incisure of medial aspect of coronoid process (center and right of image). However, we recognize dogs affected by a spectrum of lameness intensity and pain evident on elbow manipulation without visible fissuring or fragmentation, but with altered morphology of cartilage and subchondral bone.6,28 Seemingly, the MCP in such dogs is affected by a slow progressive cycle of microfracture and cartilage fissuring, healing with fibrocartilage infilling of cartilage clefts, and subsequent repeated microfracture. Eventually, hyaline cartilage is replaced with fibrocartilage or similar fibrous tissue over underlying subchondral bone changes.29 Lameness and pain may initially resolve with later recurrence when fracture or eburnation of mechanically inferior fibrocartilage occurs concomitant with progressive alteration in canalicular density of subchondral bone.6 Also in our experience, some dogs6,29 have fissuring and fibrocartilage healing at loci across the medial aspect of the coronoid process with apparent sudden propagation of a microcrack to create a fragment (evidenced by a local lack of marginal healing or fibrocartilaginous ingrowth). This observation suggests that abrupt overloading of existing pathologic change created the lesion. Conversely, in some dogs, a cartilage-capped free osteochondral fragment has been identified, typically with substantially hypertrophic hyaline cartilage6 with the remainder of the medial aspect of the coronoid process and typically the medial humeral condylar articular surface affected by full thickness cartilage eburnation. In this scenario, fragmentation would seemingly be an early component of the disease process occurring at a focal site of increased loading. Once the fragment is detached, its mobility may protect it from further load-bearing and cartilage abrasion, whereas loss of subchondral vascularity contributes to increasing cartilage thickness and even necrosis of the deeper layers of cartilage and/or bone typical of free osteochondral bodies within a joint, perhaps accounting for some reports of OC like lesions.38 In our experience, it is common for the remaining intact portion of the medial aspect of the coronoid process to be affected by cartilage abrasion or eburnation because of persistent supraphysiologic loading with HUC.39 Joint pain and lameness can occur when there is intact cartilage over underlying subchondral pathology. From the preceding discussion, recognition that 2 potentially different end-stage changes, of varying chronology and intensity, can exist within the same joint, argues for treatment targeted at addressing subchondral pathology rather than removal of loose fragments alone or attempted stimulation of fibrocartilaginous infilling of superficial cartilage lesions. Notwithstanding the altered subchondral bone, focused attempts at cartilage repair seem especially challenging within a perceived environment of persistent supraphysiologic loading and abrasion attributable to HUC. Considering these proposed pathways, our preference for local treatment of end-stage MCD is subtotal coronoid ostectomy (SCO; Fig 2) whereby a pyramidal portion of the medial aspect of the coronoid process extending to include the entirety of the articular portion distal to the level of the radial incisure is removed. The surgical approach for this procedure includes blunt separation of the flexor carpi radialis/pronator teres and the superficial/deep digital flexor muscles caudal to the medial collateral ligament to allow access to, and sharp incision of, the medial aspect of the joint capsule proximal to the fan-shaped insertion of the biceps brachii/brachialis muscle complex on the medial aspect of the coronoid process. Self-retaining retractors positioned caudal to the medial collateral ligament maximize exposure of the medial joint compartment. We have employed an air-powered oscillating saw for ostectomy28 but acknowledge that arthroscopically-guided ostectomy may also be achieved using an osteotome or motorized shaver with comparable efficacy. Typical configuration of subtotal coronoid ostectomy (SCO) depicted by green lines. Red line depicts typical location of fragment at radial incisure of medial aspect of coronoid process. The caudolateral boundary of the ostectomy is at the junction of the radial incisure and a point 1–2 mm distal to the sagittal ridge of the ulnar notch. Subchondral microfractures have been found extending to the edge of this osteotomy line,6 but it includes the full extent of visible cartilage pathology and the region of subchondral pathology identified by histomorphometry.6 Our initial concerns about elbow instability (because of reduced medial articular contact surface or disruption of the ulnar portion of the medial collateral ligament) have not been substantiated. SCO in 263 dogs (437 elbows) with long term follow-up (>5 years in some dogs), resulted in consistent and durable resolution of lameness with low surgical morbidity.28 Other surgical techniques for focal management of MCD include removal of free fragments alone, and varying degrees of debridement, abrasion, or excision of the visibly diseased portion of the medial aspect of the coronoid process, either arthroscopically or via arthrotomy.4,14,17,40–44 Although histologic findings suggest that such approaches would leave a substantial portion of diseased subchondral bone in situ,6 we are unaware of any clinical studies that clearly demonstrate that more aggressive arthroplasty (e.g. SCO) yields superior outcome to less aggressive approaches. A cohort comparison study is warranted. If dynamic joint incongruency or abnormal dynamic loading are potential causes of MCD, then arguably, corrective ulnar osteotomy should be considered as a treatment approach; however, without better mechanical understanding, it is unclear what osteotomy configuration might be most beneficial. In our experience, ulnar osteotomy results in lameness of several weeks duration. Further, the severity of lameness is typically more than observed preoperatively or is associated with intraarticular procedures alone. This outcome seemingly counterbalances any potential benefits and at least in our experience, long-term outcome is equivocal in dogs with MCD without substantial humeral condylar pathology. However, when there are frictional abrasion lesions associated with the medial aspect of the humeral condyle or where definitive humeroradial incongruity is evident on CT or arthroscopic assessment, ulnar osteotomy may be justified as we describe later. We do not perceive a need for ulnar ostectomy unless >4 mm humeroradial incongruity is definitively identified. When rotational instability with excessive supination loading force is suspected, we use a biceps ulnar release procedure (BURP; Fig 3) that involves tenotomy of the distal insertion of the biceps brachii/brachialis complex onto the ridge immediately caudal to the abaxial portion of the MCP.34,35 Considering our proposed pathogeneses, dogs with focal subchondral pathology in the region of the radial incisure might be potential candidates for this procedure. Clinically, case selection is based either on presence of fissure formation in the region of the radial incisure without overt fragmentation or discernable osseous incongruity, or where there is a high index of suspicion that dynamic incongruity is the underlying cause of MCD. The latter cases are typically juvenile dogs with bilateral elbow pain/lameness and minimal arthroscopic changes bilaterally or with minimal arthroscopic changes affecting the elbow contralateral to an elbow overtly affected by fragmentation (e.g. synovitis, cartilage malacia, fibrillation or radial incisure fissuring). We have used this procedure before development of end-stage disease and although early outcomes have been encouraging with resolution of clinical signs and negligible morbidity,35 further investigation of indications and outcomes is needed before recommendations for clinical use are made. Biomechanical data will be necessary to determine if BURP reduces joint contact pressure associated with HUC. Whether BURP can alter disease progression, preventing cartilage disease or fissures of the MCP becoming fragmented or reducing persistent frictional abrasion of the medial compartment after SCO remains to be determined. Equally, it is unknown at this stage whether BURP may be used for successful palliative treatment of end-stage medial compartment erosion where periarticular fibrosis or profundity of pathology may negate the positive effects of tendon release. Typical configuration of biceps ulnar release procedure (BURP) depicted by red broken line. Note fan-shaped insertion of the biceps brachii immediately surrounding the medial aspect of the coronoid process. Nonsurgical management remains the major alternative approach when focal surgical treatment is considered inappropriate or has already been performed without resolution of clinical signs. Successful nonsurgical management plans involve simultaneous use of a moderated exercise routine, body weight control, judicious use of nonsteroidal antiinflammatory medication or prescription analgesic agents, and use of nutraceuticals or disease modifying compounds (of which glucosamine and chondroitin sulfate containing preparations, or compounds such as pentosan polysulfate may hold most promise). Other adjunctive therapies should also be considered, including reduced load-bearing exercise (such as hydrotherapy), physical therapies such as massage, trans-cutaneous electrical nerve stimulation, shock wave therapy and holistic, magnetic and alternative therapies like acupuncture. There is limited scientific evidence to support use of many of these modalities but a wealth of cross-species experience and low morbidity may prompt their use in individual cases. Our current decision algorithm for the coronoid process (Fig 4) indicates SCO whenever end stage MCD, manifesting as either fragmentation, major fissuring, or full thickness eburnation of articular cartilage, is identified arthroscopically. Decision making algorithm chart for treatment of medial coronoid disease (MCD) without significant medial humeral condylar pathology. When early or mild MCD is identified arthroscopically, typically manifested as superficial cartilage fibrillation or cartilage malacia, frequently limited to the most craniomedial portion of the medial aspect of the coronoid process, other factors are considered before selecting SCO, BURP, or nonsurgical management. These factors are balanced in an attempt to answer 3 questions: Is the subchondral pathology a sufficiently important current cause of lameness or pain to justify SCO despite lack of superficial pathology? Do the arthroscopic findings suggest a rotational component manifested by pathologic changes in the region of the radial incisure, supporting use of BURP in an attempt to reduce supination forces acting on the joint? Is the current level of pathology likely to progress to end stage MCD with future lameness or pain if left untreated? The 2 factors considered most important in affirming decision to perform SCO, in light of ambiguous arthroscopic findings, are severity of clinical signs (lameness and pain on manipulation) and young age (where skeletal immaturity is considered a strong indicator for subsequent development of end stage MCD). Radiographic findings (including subjective intensity of radiographic subtrochlear sclerosis), anticipated dog and owner compliance with nonsurgical management protocols, and response to previous attempts at nonsurgical management are also considered. As an example, using our algorithm, a 6-year-old dog with subtle clinical lameness or elbow pain, with superficial cartilage fibrillation focally at the tip of MCP would be managed nonsurgically, whereas a 6-month-old dog with moderate lameness associated with low-grade superficial disease of the medial coronoid evident arthroscopically and intense subtrochlear sclerosis evident radiographically would be treated by SCO or BURP depending on the degree of medial coronoid pathology in terms of fibrillation, fissuring, or fragmentation. A sliding scale analogy may be most appropriate when trying to combine these variables (Fig 5) and therefore, a small degree of subjectivity may occur in selected cases. Ongoing studies to categorize and prioritize bone marrow lesions of the MCP by MRI and CT will undoubtedly help to deracinate this subjectivity. Cross-referencing radiographic and arthroscopic findings with micro-CT and histomorphometric analysis of excised coronoid segments will also help clarify the relationship between incongruity and morphologic changes and may help guide decision making in the future. Schematic representation of sliding scale analogy applied to incorporate additional factors into decision-making algorithms for treatment of medial elbow disease. OC (and resultant OCD) is a well-recognized disease of the medial compartment of the elbow, and concomitance with MCD is frequent (30/33 elbows in one of our studies45). This may reflect a potential role for incongruity in the etiopathogenesis of both diseases, although numerous developmental factors have been implicated including genetic46,47; dietary48; growth rate49; and endocrine50 factors. Many reports have grouped treatment of these 2 diseases together, and have not reflected the spectrum of pathology identified in our canine population. Specifically, we have commonly recognized OCD in conjunction with varying degrees of cartilage erosion (kissing lesions) of the medial humeral condyle, apparently associated with MCD, further supporting the role of incongruity in etiopathogenesis. These kissing lesions are identifiable during arthroscopy or arthrotomy as clusters of axially-orientated, linear abrasion tracts/striations ranging from superficial cartilage fibrillation to full thickness cartilage eburnation with exposure of subchondral bone. There is also substantial variation in the surface area of the medial humeral condyle affected, ranging from focal regions a few millimeters in diameter, to eburnation across almost the entire medial articular surface. These lesions are often centered at or immediately adjacent to the OCD lesion, but remain distinct in both gross appearance and depth of subchondral defect. The medial aspect of the coronoid process inevitably demonstrates a similar degree of cartilage pathology across some portion of its surface (mirror-image), whereas additional presence of macroscopic fragmentation or fissuring, although common, is more variable. Both surgical and nonsurgical management of OCD of the medial humeral condyle (with or without MCD) result in progression of osteoarthritis,3 yet variation in prognostic outcome within the spectrum of identified disease and detailed medium- and long-term outcomes have typically not been reported. In our experience, presence of marked cartilage disease of the medial humeral condyle has been associated with relatively poor clinical outcomes and in some cases, even if MCD is simultaneously treated by SCO, may continue to progress with eventual full-thickness eburnation throughout the medial joint compartment. Approximately equal load sharing between the larger humeroradial contact area and the smaller humeroulnar contact area in normal elbows51 may account for the severity of some lesions. It is considered unlikely that fibrocartilaginous ingrowth from subchondral bone in this region