A literature search using the MEDLINE database was performed through June 2009. In addition, the references of the retrieved articles were reviewed for relevant publications not identified in the MEDLINE search. No randomized trials on shoulder resurfacing were identified. For the Copeland prosthesis, five case series and one matched-pair analysis were identified. Three of the six studies were published by the developers of the Copeland prosthesis with likely overlap in patients; some of whom underwent TSR and some HR. Three additional case series using three different prosthetic designs were also identified. Resurfacing has been reported in patients with osteoarthritis, rheumatoid arthritis, osteonecrosis of the humeral head, instability arthropathy, post-traumatic arthropathy, and postsurgical glenohumeral arthritis. The largest prospective and/or consecutive series are described below.
The appropriate comparison for TSR would be either total shoulder arthroplasty or hemiarthroplasty, depending on whether the glenoid was resurfaced or not. Therefore, comparative outcome studies of total shoulder arthroplasty and hemiarthroplasty are also described below.
Total Shoulder Arthroplasty and Hemiarthroplasty
Bryant et al. conducted a meta-analysis of four randomized trials that compared total shoulder arthroplasty with humeral head replacement or hemiarthroplasty. Included were 112 patients with an average age of 68 years. Two-year follow-up showed an advantage of total shoulder arthroplasty over hemiarthroplasty for pain and function on the University of California at Los Angeles (UCLA) scoring system. The score for function at two years was 8.1 in the total shoulder arthroplasty group and 6.6 in the hemiarthroplasty group. There was no evidence of heterogeneity between studies for this domain. Forward elevation was improved by 13 degrees for the total shoulder vs. hemiarthroplasty groups. Pain scores also favored total shoulder arthroplasty (8.6 vs. 6.5), although the heterogeneity among the studies decreased confidence in this result. The authors noted the uncertainty in the longer-term effects of erosion of the glenoid (with hemiarthroplasty) compared with loosening of the glenoid component (with total shoulder arthroplasty), concluding that longer follow-up was needed.
Radnay and colleagues conducted a systematic review of 23 studies, primarily case series, describing outcomes from patients (n = 1952) treated with either total shoulder arthroplasty or humeral head replacement between 1966 and 2004. Patients treated with total shoulder arthroplasty were slightly older than those treated with hemiarthroplasty (average 66 vs. 63 years of age). The mean follow-up was 43 months, with a range of 30 to 116 months. Analysis showed an advantage of total shoulder arthroplasty over hemiarthroplasty for pain and function. For the 14 studies (1185 patients) that included pain as an outcome measure, postoperative pain scores were significantly improved for shoulders undergoing glenoid resurfacing (mean of 86) compared with those undergoing isolated hemiarthroplasty (mean of 78). Patients who underwent total shoulder arthroplasty outperformed those who underwent hemiarthroplasty in forward elevation (141 degrees vs. 125 degrees) and external rotation (35 degrees vs. 25 degrees). The number of revisions was significantly lower for total shoulder arthroplasty over hemiarthroplasty (6.5% vs. 10.2%) and 8.1% of the hemiarthroplasties were converted to total shoulder arthroplasty within the follow-up period. Revisions for all-polyethylene glenoid components (1.7%) were lower than for the metal-backed glenoid components (6.8%). These authors (along with a number of others) noted that the choice between total shoulder arthroplasty and hemiarthroplasty for the treatment of end-stage primary glenohumeral osteoarthritis remains controversial, due to uncertainty in long-term effects on the glenoid.
Copeland Mark-1, 2, or 3 Prosthesis
In 2001 Levy and Copeland published outcomes from a consecutive series of 103 prostheses in 94 patients treated between 1990 and 1994 with the Copeland Mark-2 prosthesis. Out of the series, one patient died less than 24 months after shoulder replacement and four patients were lost to follow-up, resulting in review of 98 shoulders. Sixty-eight shoulders also received a glenoid component for TSR while 35 received only the humeral component. Included were patients with osteoarthritis, rheumatoid arthritis, avascular necrosis, instability arthropathy, and post-traumatic arthropathy. The average age was 64 years (range of 22 to 88). About 20% of patients had irreparable or incompletely repaired cuff tear arthropathy. Independent assessment showed an improvement in the Constant score from 15 (age-adjusted of 24%) at baseline to 52 (75%) at an average 6.8 years after resurfacing (range, 5-10 years). The best results were observed in patients with primary osteoarthritis and TSR with a Constant score of 94%. Humeral resurfacing alone in this population resulted in a Constant score of 74%. Shoulders with cuff arthropathy or instability arthropathy had Constant scores of 61% and 63%, respectively. Radiological review on 88 humeral implants showed no evidence of radiolucency in 69%, a lucent line less than 1 mm in 28% and a progressive lucent line more than 2 mm in two shoulders. Eight shoulders were revised, five of which were revised to a stemmed humeral component. Mild subluxation of the humeral head was observed in 15 shoulders, moderate superior migration was observed in seven, and severe superior subluxation with obliteration of the acromiohumeral interval was observed in eight. Subluxation of the prosthesis was associated with cuff tear arthropathy. Additional reports from this group are retrospective reviews of patients with osteoarthritis or rheumatoid arthritis treated between 1986 and 1998 with Copeland Mark-1, Mark-2, or Mark-3 prostheses. Overlap in patients between these publications is likely.
Another group from England reported outcomes from a consecutive series of 52 patients (56 shoulders) who received humeral resurfacing with the Copeland Mark-3 prosthesis. Six patients died of other causes and two were lost to follow-up, resulting in an average 34-month assessment (range 24-63 months) of 44 patients (48 shoulders). Nine shoulders were followed for more than four years. The primary diagnosis was osteoarthritis in 20 patients, rheumatoid arthritis in 26, post-traumatic arthrosis in one and rotator cuff arthropathy in one. The average age was 70 years (range 34-84). Independent post-operative assessment showed an improvement from 16 to 54 in the Constant score. One patient converted to total shoulder arthroplasty, three were revised for impingement and one patient had a fracture, resulting in an estimated 98% implant survival at four years (92% survival for any revision). A German group of surgeon-investigators reported a matched pair analysis comparing 22 patients who underwent resurfacing with the Copeland Mark-3 prosthesis with 22 matched patients who had received total shoulder arthroplasty in the same year. At 12-month follow-up, total shoulder arthroplasty resulted in greater improvement in the Constant score (from 26 at baseline to 67at follow-up) in comparison with humeral resurfacing alone (from 33 at baseline to 59 at follow-up). Two of the patients who underwent humeral resurfacing converted to total shoulder arthroplasty because of painful glenoid erosion.
Durom Cup Prosthesis
A prospective study with the Durom cup prosthesis was conducted in 35 patients (42 shoulders) with pain and limited function associated with rheumatoid arthritis between 1997 and 2000. Thirteen shoulders had a normal rotator cuff or only partial tearing and thinning, another 13 shoulders had a complete rupture with a defect that was repaired. Nine shoulders had a massive rotator cuff tear with a defect of > 5 cm in diameter where the humeral head had migrated under the acromion. These were not repairable, and in these patients the Durom cup was implanted in a slightly more valgus position. The average age of the patients was 61 years (range of 27-78). For three patients who died and three who did not want to continue in the study, results were only available to the 12-month follow-up. For the remaining 29 patients, assessment at an average 73 months follow-up (all greater than 60 months) showed improvement in the Constant score from 21 to 64. Three shoulders were revised (one due to an oversized cup) and one was converted to total shoulder arthroplasty within the follow-up period. Flexion improved from 64 degrees pre-operatively to 118 degrees at a mean of 73 months after surgery. Radiographs, evaluated by two orthopedic surgeons who were blinded to the patients’ identity, showed no change in position and no sign of loosening in 33 of 35 prostheses. Proximal migration (the relationship of the humeral head to the glenoid) increased between three-month and 73-month follow-up; 22 (63%) of the shoulders had more than a 3 mm increase in proximal migration, and 13 (37%) showed 0-2 mm increase in proximal migration over follow-up. Glenoid depth increased significantly in shoulders with either intact or repaired rotator cuffs; 11 (31%) had an increase in depth of 3 mm or more. The authors concluded that humeral resurfacing with the Durom cup had less surgical morbidity and options for salvage if the implant fails, and should be considered as an option along with stemmed or reverse implants in the treatment of the rheumatoid shoulder.
Hemi-CAP Partial Resurfacing
A 2007 review article briefly described short-term outcomes (3-24 months) of 62 patients from six institutions who underwent humeral resurfacing with the HemiCAP. One prospective study in 2009 described Hemi-CAP partial humeral resurfacing in 11 patients (12 shoulders) who had advanced osteonecrosis measuring less than 40 mm (the size of the largest resurfacing device available). Half of the implants used had a diameter of 35 mm and the remaining half had a diameter of 30 mm. None of the patients had rotator cuff or labral pathology, and no patient required glenoid resurfacing. Assessments performed at three, six, 12, 18, 24, 36, and 48 months after implantation included the Western Ontario Osteoarthritis of the Shoulder (WOOS) index, the Shoulder Score Index from the American Shoulder and Elbow Surgeons (ASES) evaluation form, the Constant score, and a visual analog score (VAS) for pain. No patient was lost to follow-up. Significant improvement in function was observed at an average 30 months (range, 21-57) follow-up; the WOOS improved from 1421 to 471 (worst score 1900); the mean Shoulder Score Index improved from 24 to 75 (max of 100); and the mean Constant score improved from 23 to 62 (max of 100). Active forward elevation improved from a mean of 94 degrees to 142 degrees. All patients reported pain relief, and VAS pain scores improved from 75 at baseline to 16 at follow-up. There was no evidence of implant loosening. The authors concluded that results are promising at 30 months, but longer follow-up is required to evaluate the survivorship of the implant and its effect on the glenoid.
Shoulder resurfacing has the potential to improve pain and function to the same extent as total shoulder replacement or hemiarthroplasty, while at the same time reducing risks from the surgical procedure, preserving bone stock and reducing the difficulty with revision procedures. At this time, however, evidence in support of these proposed benefits is lacking. For some implant designs, the published literature consists of one small case series. The four independent case series identified on the Copeland prosthesis suggest better short-term outcomes with TSR or total shoulder arthroplasty than humeral head resurfacing alone. This is similar to findings of recent systematic reviews that compared hemiarthroplasty with total shoulder arthroplasty; the choice of these two procedures remains controversial due to the differing effects on glenoid erosion and glenoid component loosening. For shoulder resurfacing, questions remain about the stability and durability of these prostheses, as well as the effect of partial or total humeral resurfacing on the glenoid. Controlled studies are needed to evaluate the risks and benefits of HR and TSR in comparison with hemi- and total shoulder replacement. Several clinical trials are in progress, with estimated completion dates of 2013. At the present time, evidence is insufficient to permit conclusions concerning the effect of this procedure on health outcomes. Therefore, partial resurfacing, humeral resurfacing and TSR are considered experimental, investigational and unproven.
A search of peer-reviewed literature through March 2011 identified no new clinical trial publications or any additional information that would change the coverage position of this medical policy.
Disclaimer for coding information on Medical Policies
Procedure and diagnosis codes on Medical Policy documents are included only as a general reference tool for each policy. They may not be all-inclusive.
The presence or absence of procedure, service, supply, device or diagnosis codes in a Medical Policy document has no relevance for determination of benefit coverage for members or reimbursement for providers. Only the written coverage position in a medical policy should be used for such determinations.
Benefit coverage determinations based on written Medical Policy coverage positions must include review of the member’s benefit contract or Summary Plan Description (SPD) for defined coverage vs. non-coverage, benefit exclusions, and benefit limitations such as dollar or duration caps.