A sixty-eight-year-old retired man had a total arthroplasty of the right shoulder for treatment of disabling osteoarthrosis at another institution. The arthroplasty consisted of insertion of a humeral component and a metal-backed two-piece glenoid component without cement. The metal shell of the glenoid component was porous-coated and was secured to the glenoid with two screws.
The operative procedure was uncomplicated and resulted in substantial relief of pain and improvement of function. Three months postoperatively, the patient had an acute onset of pain with an audible pop that occurred when the shoulder was abducted and externally rotated. Standard radiographs as well as an arthrogram of the shoulder demonstrated that the polyethylene liner had detached from the metal shell of the glenoid component. At the time of operative exploration, the metal shell was found to be held securely in the glenoid and a new polyethylene liner was placed. Again, the postoperative course was uncomplicated and the patient had noticeable relief of pain and improvement of function. He was able to perform nearly all of the activities of daily living without difficulty.
Five months after the second operation, the patient again had an episode consisting of an audible pop and acute pain in the shoulder. Anterior, posterior, and lateral radiographs again revealed that the polyethylene liner had come loose from the metal shell of the glenoid component, as the metal of the humeral component was seen to be abutting the metal of the glenoid component (Fig. 1). The patient was referred to our institution for operative reconstruction at that time.
At the time of the operation, the polyethylene component was found to be distorted and loose within the glenohumeral joint. The undersurface of the polyethylene liner, which was in contact with the metal shell, had deformed into the screw-holes and the central peg of the liner had been sheared off. The humeral component was secure in the proximal aspect of the humerus and had retroverted approximately 35 degrees.
No evidence of synovitis was found within the shoulder. A tear of the rotator cuff that included the supraspinatus and infraspinatus tendons was identified and was not thought to be amenable to repair. The metal shell of the glenoid component was secure and, after removal of the screws, an osteotome was required to disrupt the bone-implant interface. The screw-holes were packed with bone graft. Because of the repeated problems with the modular glenoid component associated with an incompetent rotator cuff, we decided not to attempt to replace the glenoid component (Fig. 2).
The metal-backed shell was sectioned and evaluated histologically. No evidence of bone ingrowth was seen. Linear fibrous tissue along with polyethylene debris was found at the bone-implant interface. No evidence of osteoclastic resorption was seen.
Four months after the operation, the patient reported that he had no pain but he had limited function with active glenohumeral abduction of 60 degrees. He was subsequently lost to follow-up.
Dissociation of both modular acetabular and modular femoral total hip arthroplasty components has been reported2,6,7,10. Disassembly of the modular humeral component of a total shoulder prosthesis also has been reported4,9. To our knowledge, we are the first to report the dissociation of a modular glenoid component of a total shoulder prosthesis.
Disassembly of femoral and humeral components has been attributed to failure of the Morse taper between the head and the body of the prosthesis4,8-10. Dissociation of a femoral component has been reported after a forceful attempt at a closed reduction of a dislocation8,10. Disassembly of acetabular components differs from that of femoral and humeral components in that disassembly of acetabular components is attributable to the locking mechanism between the metal shell and the polyethylene liner. In one report of disassembly of an acetabular component, impingement of the neck of the femoral prosthesis on the polyethylene liner was shown to lever the acetabular liner out of the metal shell6.
Although the results reported for total shoulder arthroplasty have been encouraging1,3, loosening of the glenoid component has been identified as a cause of failure3,4. Tearing of the rotator cuff also has been attributed to loosening of the glenoid component as a result of proximal migration of the humeral component, which places an eccentric load on the glenoid component5. In our patient, a metal-backed glenoid component had been secured to the bone with screws and had a porous surface for biological fixation. The polyethylene liner snapped into the metal shell and was held in place by a ring around the entire circumference of the component.
The exact cause of the disassembly of the glenoid component in our patient could not be determined, but we postulated that eccentric loading of the polyethylene liner by the head of the humeral component may have had a role. On inspection, the polyethylene liner showed evidence of wear along the edge and deformation. The large tear of the rotator cuff that was noted at the time of the second revision also may have played a fundamental role in the repeated disassembly of the glenoid component.