This article is a retrospective review of 13 patients treated for a failed total elbow arthroplasty due to loosening with an allograft-prosthetic composite. The composite was placed on the humeral side in 4 patients and on the ulnar side in 9. The mean age at the time of surgery was 62 years (range, 39 to 77 years). Six patients had rheumatoid arthritis and 7 patients had post-traumatic arthritis. The interval between the last total elbow arthroplasty and the composite procedure averaged 8 years (range, 1 to 23 years). A Coonrad-Morrey semiconstrained total elbow arthroplasty was used in all but 2 patients. The patients were evaluated with radiographs and physical examination at an average of 42 months after the revision. The mean elbow performance score was excellent in 4 elbows, good for 3, fair for 1 and poor for 5. Pre-operatively, all patients had pain whereas postoperatively 9 of the 13 patients had no or only slight pain in the elbow. The mean arc of flexion was 97 degrees with an average of 28 degrees of extension to 125 degrees of flexion. Seven elbows were affected by 7 complications and 5 of the 7 required a revision procedure. A deep infection occurred in 4 elbows resulting in 3 complete removal of the composite allograft. Two non-unions occurred at the allograft-humeral junction whereas no non-unions occurred in the ulna-allograft junction.
This article confirms the high complication rate previously noted for the use of allografts in revision elbow replacement, which in this series was greater than 50%. Infection, nonunion, and loosening are significant and devastating risks of this procedure. Because of these results, the authors reserve the use of allograft-prosthesis composite elbow replacement for those elbows not amenable to primary arthroplasty with strut graft reconstruction. In addition, they routinely incorporate one gram of Vancomycin in every 40 grams of cement at the time of reinsertion of the implant. Because of the high humeral non-union rate, a step cut is incorporated at the humeral allograft junction. In addition, a long-stem prosthesis is critical to bypass the allograft-host bone junction. The use of this technique should be used with caution, with limited indications and discussion of the high potential risks with these patients.
Journal of Bone and Joint Surgery