INTRODUCTION

Over the past decade, the reliability and success of total elbow replacement surgery has become more established. The durability of total elbow arthroplasty (TEA) for rheumatoid arthritis is now similar to that of hip replacement.² Indications have also expanded to include the full spectrum of traumatic conditions: distal humeral nonunion,⁹ instability,¹⁰ ankylosis,⁷ established arthritis,¹¹ and acute intra-articular, comminuted fractures in selected older patients.^4,5

Until recently, wear of the polyethylene articulating surface had not been recognized as a problem or addressed in the literature. Although this may not be a practical concern in unconstrained and unlinked designs, wear of the articular polyethylene bushings in linked, semiconstrained designs is a potential issue, particularly with the evidence of the increasing longevity and improved function that these designs now have, resulting in increased use of the limb. Until recently, this specific topic has not been considered when complications with TEA are being discussed.³ This chapter focuses on our experience with the problem as the development requires or implies long-term stable fixation. Hence, there has been little in the literature regarding this problem with other designs.

To address this issue, we reviewed the Mayo Clinic experience with reoperations to exchange worn bushings in the linked, semiconstrained Coonrad-Morrey TEA. Twelve of 919 TEAs reviewed (1.3%) had undergone isolated bushing exchange for wear.⁶ An additional six patients were diagnosed as having bushing wear on the basis of an asymmetric anteroposterior orientation of the ulnar component within the humeral yoke, although these patients did not have a revision.

CLINICAL AND RADIOLOGIC FEATURES

Pain, crepitus and squeaking sounds are the most common presenting features of articular bushing wear. Loss of range of motion is not marked, and there are no symptoms of functional instability or weakness. In our experience, the patients with post-traumatic arthritis had a higher prevalence of bushing wear (seven of 294; 2.4%) than did those with rheumatoid arthritis (five of 377; 1.4%).⁶

Radiographic assessment is conducted before the total elbow replacement to identify deformity and after replacement to identify signs of bushing wear, osteolysis and status of the implant. The criteria for the assessment of wear were described by Ramsey et al.¹⁰ Anteroposterior plain radiographs of the elbow in full extension made following the index arthroplasty and at the time of bushing exchange are compared. The prosthesis was designed with 7 to 10 degrees of varus-valgus laxity. A line is drawn parallel to the yoke of the humeral component, and another line is drawn parallel to the medial or lateral surface of the articular surface of the ulnar component. An angle of intersection of more than 7 degrees between these two lines indicates alteration of the bushing due to wear or plastic deformation. An angle of more than 10 degrees is considered to indicate mild to moderate bushing wear (Fig. 64-1). Of importance is that none of our patients experienced loosening of the implant because of wear. It must be emphasized the extensive osteolysis seen in the past around the ulnar component was related to osteolysis caused by loosening of a “precoat” ulnar surfaced component. These devices were implanted between 1994 and 2000. Some have confused this appearance as being caused by wear. As in the hip the proper interpretation is the implant becomes loose and the particulars associated with this are termed accelerated bushing wear (Fig. 64-2). Typically bushing wear, even when excessive, causes local osteolysis, but not implant loosening (Fig. 64-3).

FIGURE 64-1 Wear of the polyethylene is measured according to the angle formed by the articulating portion of the ulnar component and the medial or lateral aspect of the humeral yoke. Angles in excess of 7 degrees indicate displacement greater than allowed by the design tolerances. When this angle is 10 degrees, mild to moderate wear is present.

(With permission, Mayo Foundation.)

FIGURE 64-2 Extensive osteolysis occurs from loose stems usually from a precoat surface, as is the case here. This appearance is not because of worn bushings

FIGURE 64-3 Extensive bushing wear at 16 years in a patient with gun shot deformity. Note limited osteolysis not involving the stems

SURGICAL TECHNIQUE

The treatment of symptomatic wear is simply to exchange the worn articular bushings. The previous skin incision is used to explore the elbow, and the ulnar nerve is palpated. If the patient has ulnar nerve symptoms, the nerve is explored and decompressed. If the nerve is not symptomatic, it is identified proximally at the medial aspect of the triceps and is protected throughout the procedure. If the distal part of the humerus has been resected or is absent, the triceps is left attached to the ulna, the pseudocapsule is entered medially and laterally, the articulation is disengaged, and the humerus and ulna are separated. If the condyles are intact, the triceps is again reflected from the ulna according to a previously described technique. At this juncture, the anterior aspects of the medial and lateral epicondyles are removed to an extent sufficient to allow the implant locking pin to be removed both medially and laterally (Fig. 64-4). The posterior aspects of the condyles are left intact. The medial and lateral bushings are removed from the humerus, and the bushing is removed from the ulna. The soft tissue is assessed, and a thorough débridement is carried out. If the wear is sufficient to have resulted in impingement of the metallic ulnar component on themetallic humeral component, then black synovitis is the predominant feature (Fig. 64-5).

FIGURE 64-4 The condyles are preserved by removing sufficient anterior condylar bone to expose and remove the pin.

(With permission, Mayo Foundation.)

FIGURE 64-5 Metallic debris causes a black discoloration of the synovium

After débridement, the implant is inspected for the integrity of the fixation and orientation to determine if revision of either implant is necessary. If there has been resorption or osteolysis at the distal aspect of the humerus or the proximal aspect of the ulna, the interface is thoroughly cleaned and is filled with polymethyl methacrylate. Fresh bushings are then inserted in the ulnar and humeral components, and the implant is coupled with the use of the pin-within-pin snap-fit articulation (Fig. 64-6). If the preoperative assessment and intraoperative evaluation indicated a fixed angular deformity that cannot be corrected passively, then the soft tissue is released to allow correction of the deformity. Thus, an extensive flexor release from the humerus is carried out for the treatment of varus deformity. Similarly, an aggressive extensor tendon release, including the distal fibers of the brachioradialis, is performed to treat a fixed valgus deformity. The triceps is reattached with use of a cruciate and transverse drill pattern, as previously described.¹

FIGURE 64-6 The replacement of worn bushings is considered a straightforward procedure

RESULTS

The average age at the time of index TEA of the 12 patients who had bushing exchange was 44 years of age compared with a mean age of 62 years in those who did not undergo bushing exchange. No significant difference in hand dominance, body weight, occupation, or lifestyle was found in this small group.

Radiographic assessment before index TEA in this group revealed a markedly distorted joint in nine patients, with severe rheumatoid arthritis in four, marked varus or valgus deformity of more than 10 degrees in nine, and gross dissociation (a flail elbow with loss of the distal humeral condyles) in four (Fig. 64-7). Nine of the elbows had absence of one or both distal humeral condyles.

FIGURE 64-7 A, Severe elbow deformity of 50 years’ duration secondary to fracture of the humeral condyle as a child. B, Worn bushing 5 years after the surgery. The patient worked as a lumberjack against medical advice.

All 12 patients had an excellent or good result immediately following the index arthroplasty. Postoperative anteroposterior radiographs revealed the articulation to be at the limits of the designed angular tolerance in nine of the 12 patients, indicating significant stresses at the articulation from severe pre-existing deformity. At the time of bushing exchange surgery, no significant osteolysis or loss of fixation was present (Fig. 64-8).

FIGURE 64-8 At 12 years, the humeral bushing is worn. Note osteolysis is limited (A). Seven years after bushing exchange shows minimal wear (B).

Isolated bushing exchange was technically successful in all of the patients. At average follow-up of 65 months, three patients underwent a second bushing exchange for wear again. All three had marked initial deformity. Two of the twelve patients had limited success due to persistent pain from ulnar nerve neuropathy.

More recently, Wright and Hastings also documented bushing wear in 10 patients with the TEA of the same design.¹³ They identified post-traumatic arthritis, supracondylar nonunion, male sex, young age, and high activity level as associated factors. Importantly, these investigators also implicated failure of the “O” ring, which allows the axis pin to back out as an addition factor of accelerated bushing wear. In a series of semiconstrained TEAs for chronic dislocation of the elbow, Mighell et al⁸ documented one in six required bushing exchange.

Further information is partly limited by the low incidence and by lack of routine stress radiographs to document the absolute wear rate. However, probably the most significant feature with the greatest prognostic importance is the presence of severe preoperative deformity at the time of index elbow replacement. In some cases, the angular deformity may even exceed the tolerance of the design but it is these very problems that can be addressed only by this type of coupled implant.

Therefore, our current practice would be to extensively release soft tissue contractures to eliminate progressive deformity in such cases. It would be more appropriate to resect bone in order to lessen soft tissue tension that can result in an imbalance that differentially loads the polyethylene. When a patient presents with asymptomatic radiographic evidence of wear, the process is discussed with the patient. Typically, the problem is followed up radiographically because, in our experience, the wear debris does not cause fixation-compromising osteolysis. If a patient has pain or mechanical squeaking, then revision is offered.

CONCLUSIONS

Despite the increased longevity of semiconstrained linked total elbow prosthesis and the use of these implants to treat an increasingly complex array of pathologic conditions, the prevalence of articular wear requiring reoperation is very low (1.3% in our experience of such implants inserted over a 20-year period). A higher bushing revision rate is associated with a younger patient with traumatic conditions and long established fixed deformity. The presence of substantial malrotation of components at the time of implantation, even without deformity, can also contribute to increased bushing wear.¹²

Isolated bushing exchange is a successful procedure. The intervention is not extensive, and morbidity is minimal when compared with that associated with the index arthroplasty. This problem is likely to receive more attention with the rapid increase in the number of TEAs performed worldwide, improved longevity of the implant, expanded indications, and increased patient activity. Attempts to mitigate the problem of wear, such as by increasing the thickness of polyethylene bushings and using cross-linked polyethylene, are being studied. At present, the need to perform adequate soft tissue release to balance the elbow alignment in patients with long-standing deformity should be emphasized, as should cautioning patients against exceeding the recommended activity and lifting restrictions.