Background: Glenoid component loosening has been a leading cause of failure of total shoulder arthroplasty. In the present study, we evaluated the outcome of reimplantation of a new glenoid component following removal of the previous glenoid component and placement of an allograft in order to determine the results, risk factors for an unsatisfactory outcome, and rate of failure associated with this procedure.
Methods: We reviewed the data on seven shoulders in seven patients. At the time of glenoid component reimplantation, two shoulders received a cemented all-polyethylene glenoid component, three received a bone-ingrowth metal-backed component with columns and screws, and two received a bone-ingrowth metal-backed component with columns and screws augmented with bone cement. The average duration of follow-up was seventy-nine months. At the time of the latest follow-up, all patients were evaluated clinically and radiographically, patient satisfaction was assessed, and the result was graded according to a modified Neer rating system.
Results: Two patients had positive growth of Propionibacterium acnes on culture of intraoperative specimens obtained at the time of revision surgery and had continuing pain, and both underwent repeat revision. The remaining five patients expressed satisfaction with the procedure and stated that they felt better following surgery. The mean preoperative pain score for these five patients (on a scale from 1 to 5) was 4.6, and the mean postoperative pain score was 2.4 (p = 0.0042). Range of motion, however, did not improve. The Neer rating of the result (determined for the five patients who did not undergo repeat revision) was excellent for one patient, satisfactory for one, and unsatisfactory (because of limitation of motion) for three.
Conclusions: Reimplantation of a glenoid component into a previously grafted bed can provide pain relief for most patients, but motion cannot be reliably improved.
Level of Evidence: Therapeutic Level IV. See Instructions to Authors for a complete description of levels of evidence.
1 Department of Orthopaedic Surgery, Stanford University, 300 Pasteur Drive, Edwards R155, Stanford, CA 94305
2 Mayo Clinic, 200 First Street S.W., Rochester, MN 55905
Revision total shoulder arthroplasty is a technically challenging procedure. This is especially true for glenoid revision surgery, a leading source of failure after total shoulder arthroplasty, and there is a paucity of information addressing this issue1-9. In a 2001 report on glenoid revision surgery2, pain relief was achieved in twenty-six of thirty patients who had received a new glenoid component, as compared with twelve of eighteen patients who had undergone glenoid component removal. Thus, when possible, a new glenoid component probably should be implanted during revision surgery for more consistent results. However, reimplantation of a component may be impossible because of insufficient glenoid bone stock, in which case component and cement removal combined with autogenous8,10,11 or allogenic2,12 bone-grafting of the deficient glenoid may be the better option. In some cases, the patient will continue to have pain and the surgeon may have to consider reimplantation of a glenoid component2,8,10-12. Case reports on this treatment option have been published1,5,12. In our earlier report on three patients who were managed with allograft bone-grafting and component reinsertion because of pain six, twenty, and thirty-six months after bone-grafting, two patients had a good clinical outcome five and eight years after component reinsertion whereas the third patient had repeat loosening of the glenoid component and was managed with component removal yet again, two years after the reinsertion procedure1.
The purpose of the present study was to expand our earlier case reports on the outcome of reimplantation of a glenoid component after prior removal of the glenoid component and allogenic bone-grafting of the deficient glenoid bone stock in order to determine the results, risk factors for an unsatisfactory outcome, and rate of failure associated with this procedure. Other than the case reports mentioned above, we are not aware of any previous reports on this technique.
Materials and Methods
Between 1976 and 2002, 110 revisions of the glenoid component of a total shoulder replacement were performed at our institution. Thirty-four shoulders were treated with removal of the glenoid implant and placement of allogenic bone graft without one-stage component reimplantation because of insufficient glenoid bone stock. Seven of these thirty-four shoulders (in seven patients) underwent reimplantation of a new glenoid component for the treatment of persistent, severe shoulder pain. All seven patients had a complete preoperative evaluation and operative report, and all were followed for a minimum of two years or until the time of repeat revision surgery. The present study was approved by our institutional review board.
The average age of the patients at the time of the index glenoid reimplantation procedure was sixty-nine years (range, sixty-five to seventy-seven years). Four of the patients were men, and three were women. Four shoulders were on the dominant side. The underlying diagnosis was osteoarthritis in four shoulders and posttraumatic arthritis in three shoulders. No patient had a rotator cuff tear at the time of repeat revision surgery. The mean time-interval from the primary arthroplasty to the revision arthroplasty with glenoid component removal and bone-grafting was thirty-eight months (range, thirteen to eighty-four months). The mean time-interval from glenoid component removal and bone-grafting to subsequent reimplantation was twenty-nine months (range, six to ninety-one months). The mean time-interval from glenoid component reimplantation to the latest follow-up or repeat revision of the glenoid component was seventy-nine months (range, thirteen to 130 months). Two patients had subsequent repeat revision with removal of the reimplanted glenoid component at thirteen and eighteen months. Excluding these two patients who had repeat revision, the mean follow-up interval for the remaining five patients was 104 months (range, sixty-two to 130 months).
The glenoid component that was used for the primary arthroplasty was a cemented all-polyethylene component in four patients, a cemented metal-backed component in two, and an uncemented metal-backed component in one. The primary humeral component was cemented in six patients and uncemented in one. The primary reason for the first glenoid revision procedure was loosening of the glenoid component in six cases and glenoid loosening with posterior instability in one. Glenoid bone stock was insufficient for the placement of a new glenoid component at the time of this procedure in all cases. The grading of glenoid bone deficiency at that time and subsequently was based on the method described previously2. Glenoid bone loss was graded as central and peripheral, severe in four patients; as central, severe in two patients; and as central, moderate in one patient. All patients undergoing removal of the glenoid component and cement had impacted cancellous allogenic bone-grafting of the glenoid. The graft material was frozen. This material initially was obtained from our bone bank, but since 1999 it has been obtained from a commercial source (CanPac; AlloSource, Centennial, Colorado). After thawing, the graft material was washed in sterile saline solution to remove blood and marrow elements. The material was then impacted into the glenoid vault; 15 to 50 cc of graft material was used.
At the time of reimplantation of the glenoid component, a deltopectoral approach was used in six cases and an antero-medial approach with detachment of the anterior part of the deltoid from the anterior aspect of the acromion and the distal aspect of the clavicle was used in one. The subdeltoid space was freed of scar. The subscapularis was divided near its insertion on the humerus, and the inferior capsule was released from the humeral neck. A modular humeral head component was removed to visualize the glenoid in six cases, and the entire humeral component was removed because of loosening of the implant in one. Glenoid bone stock had reconstituted with bone and a small amount of interposed fibrinous scar tissue. At the time of glenoid reimplantation, there was mild medial migration of the glenoid surface (≤5 mm) in five shoulders, moderate medial migration (6 to 10 mm) in zero shoulders, and severe medial migration (≥11 mm, with erosion past the lateral aspect of the coracoid base on the axillary view) in two shoulders. The newly formed glenoid was prepared with use of standard instrumentation for the placement of a primary glenoid component. In addition, with use of small burrs and curettes, the small amount of interposed fibrous tissue was removed to secure firm attachment of the new component with or without cement to the reconstituted bone.
At the time of glenoid component reimplantation, two patients had quite ample bone volume as determined by drilling a central pilot hole in the glenoid and measuring the depth with a gauge; these two patients received a cemented all-polyethylene component (Figs. 1-A, 1-B, and 1-C). Three patients with less ample bone depth received a bone-ingrowth metal-backed component with columns and screws. Two patients with severely compromised bone depth received a metal-backed component with columns and screws augmented with bone cement (Smith and Nephew, Memphis, Tennessee) (Figs. 2-A and 2-B, 2-C and 2-D). Three patients additionally had exchange of the modular humeral head component to a different size—usually one size smaller—to accommodate the volume of the replaced glenoid component. No shoulder was clinically infected. Histological analysis of frozen sections was performed for all shoulders, and the result was negative in all cases. Multiple deep specimens were taken from all shoulders intraoperatively, with the specimens from two shoulders subsequently demonstrating growth of Propionibacterium acnes on culture.
All patients who underwent shoulder arthroplasty at our institution had a clinical assessment with use of a validated shoulder-analysis sheet13,14. Pain was graded on a scale from 1 to 5, as previously described14,15, with 1 point indicating no pain, 2 points indicating slight pain, 3 points indicating pain after unusually vigorous activities, 4 points indicating moderate pain, and 5 points indicating severe pain. Patient satisfaction was assigned a score of 1 point if the patient felt much better, 2 points if the patient felt better, 3 points if the patient felt the same, and 4 points if the patient felt worse. Active elevation and external rotation were recorded in degrees. Internal rotation was graded on the basis of the posterior spinal segment that the patient could reach with the thumb. A modified Neer system was used to rate the result9. The result was considered to be excellent if the patient had no or slight pain, was satisfied with the result, and had at least 140° of active elevation and at least 45° of external rotation. The result was considered to be satisfactory if the patient had no, slight, or occasional moderate pain with vigorous activities; was satisfied with the result; and had at least 90° of active elevation and at least 20° of external rotation. If these criteria were not met, the result was considered to be unsatisfactory.
Preoperative and recent postoperative radiographs of the shoulder were made in the 40° posterior oblique position with internal and external rotation of the humerus, and an axillary radiograph was also made. The most recent follow-up radiographs for the seven shoulders in the present study were made at twelve, thirteen, thirty-four, eighty-five, ninety-eight, 115, and 130 months following glenoid component reinsertion. Prior to reimplantation, there was moderate anterior subluxation in one shoulder and severe posterior subluxation in one shoulder. No periprosthetic humeral radiolucent lines were present on the preoperative radiographs.
The preoperative to postoperative changes in activity level and satisfaction were assessed with use of Wilcoxon signed-rank tests. The level of significance was set at p < 0.05. Statistical analysis was performed with use of JMP Software (SAS Institute, Cary, North Carolina).
The two patients who had positive growth of Propionibacterium acnes on culture of intraoperative specimens had continuing pain and radiographic evidence of glenoid component loosening and underwent repeat revision surgery at thirteen and eighteen months following reinsertion of the glenoid component.
Overall, the mean preoperative pain score for all seven shoulders prior to glenoid component reimplantation was 4.6 (range, 4 to 5) and the mean postoperative pain score at the time of the latest follow-up (five shoulders) or just prior to revision surgery (two shoulders) was 2.9 (range, 2 to 4) (Table I); this difference represented a significant improvement (p = 0.0065). There was a significant change in patient satisfaction from preoperatively (with all patients having a score of 4) to postoperatively (mean score, 2.6; range, 2 to 4) (p = 0.0041). Five of the seven patients stated they felt better after reimplantation of the glenoid component, whereas two patients (both of whom had repeat revision because of infection and loosening) stated that they felt worse. There was no significant difference between preoperative and postoperative range-of-motion parameters. The average active elevation was 81° (range, 25° to 180°) preoperatively and 88° (range, 45° to 155°) postoperatively (p = 0.7422). The average external rotation was 29° (range, 10° to 70°) preoperatively and 39° (range, 10° to 70°) postoperatively (p = 0.9691). The mean internal rotation was to L3 preoperatively and to the lateral aspect of the ilium postoperatively (p = 0.1154).
With exclusion of the two patients who had positive growth on culture of intraoperative specimens and underwent repeat revision surgery, the mean pain score was 4.6 (range, 4 to 5) preoperatively and 2.4 (range, 2 to 3) postoperatively; this difference represented a significant improvement (p = 0.0042). There was a significant change in patient satisfaction from preoperatively (with all patients having a score of 4) to postoperatively (with all patients having a score of 2) (p < 0.0001). There was no significant difference between preoperative and postoperative range-of-motion parameters. According to the modified Neer rating system, one shoulder had an excellent result, one had a satisfactory result, and three had an unsatisfactory result because of limited range of motion (Table I).
Postoperative radiographs were made at a mean of 5.8 years (range, one to 10.8 years) after the time of glenoid reimplantation. The two patients who had positive growth on culture at the time of surgery had a 2-mm complete radiolucent line surrounding the glenoid prosthesis. These two components were considered to be loose radiographically and were clinically loose at the time of repeat revision surgery. Two of the remaining five patients had a 1-mm incomplete radiolucent line adjacent to the glenoid bone-cement interface. Periprosthetic humeral radiolucency was not present in any shoulder. One shoulder had severe anterosuperior subluxation.
On the basis of our previous report, approximately two-thirds of patients who have glenoid component removal and bone-grafting without reimplantation at the time of revision surgery have a satisfactory clinical result2. Thus, the present study addresses the uncommon instance in which a patient has continuing pain and wishes to proceed with another surgical intervention. During glenoid reimplantation, we consistently observed that the corticocancellous morsellized allograft generally had become consolidated; however, this consolidation had not occurred to the point that the glenoid had the same quality of bone as the native glenoid.
Surgeons should be aware that a low-grade, non-clinically apparent infection may be present in the shoulder of a patient who has had failure of prosthetic replacement, with the organism usually being Propionibacterium acnes. This finding accounted for the two clinical failures in the present study. We maintain a high index of suspicion for occult infection in patients who have had failure of a shoulder arthroplasty, and we carefully assess such patients clinically for excessive pain at rest, excessive pain with any range of motion, severe stiffness, and, of course, redness, swelling, induration, or marked local tenderness. Radiographs are carefully studied for any indication of unusual bone erosion. The white blood-cell count with differential, the erythrocyte sedimentation rate, and the C-reactive protein level are assessed in all patients. If there is any remaining suspicion at all about whether or not an infection exists, the patient undergoes aspiration of the shoulder under fluoroscopic guidance followed by analysis of the fluid for determination of the cell count as well as the evaluation of growth on aerobic and anaerobic culture.
The five patients who did not have an infection had a substantive reduction in pain and expressed satisfaction with the result of surgery. We also noted, however, that motion could not be reliably improved in spite of the performance of tissue releases. This lack of improvement may have been due to scar formation with loss of tissue flexibility associated with multiple revision procedures. The limited motion compromised the rating of the result as determined with use of the modified Neer scoring method9,14.
A weakness of the present study is the relatively small number of cases; however, to our knowledge, the present study represents the largest study, to date, that has analyzed the results of reimplantation of a glenoid component after prior resection of a failed glenoid component and placement of glenoid bone-graft at the site of a total shoulder arthroplasty. These intermediate-term results are encouraging because they show that repeat glenoid implantation can lead to pain relief and patient satisfaction. However, the long-term results, including whether or not glenoid loosening will occur subsequently, have yet to be investigated. Also, as documented in the present series, the problems encountered during revision surgery, in addition to some degree of remaining glenoid bone loss, often include tissue stiffness, glenohumeral instability, and perhaps a subclinical, low-grade infectious process.
In conclusion, reimplantation of a glenoid component following component removal and bone-grafting of the glenoid can reduce pain and can provide satisfaction for the majority of patients. However, improvement in shoulder motion is unlikely to occur. ▪
Disclosure: The authors did not receive any outside funding or grants in support of their research for or preparation of this work. Neither they nor a member of their immediate families received payments or other benefits or a commitment or agreement to provide such benefits from a commercial entity. A commercial entity (Smith and Nephew) paid or directed in any one year, or agreed to pay or direct, benefits in excess of $10,000 to a research fund, foundation, division, center, clinical practice, or other charitable or nonprofit organization with which the authors, or a member of their immediate families, are affiliated or associated.
Investigation performed at the Mayo Clinic, Rochester, Minnesota
1. , Sperling JW, Cofield RH. Reimplantation of a glenoid component after component removal and allograft bone grafting: a report of 3 cases. J Shoulder Elbow Surg. 2002;11: 637-41.
2. , Sperling JW, Cofield RH, Rowland CM. Glenoid revision surgery after total shoulder arthroplasty. J Shoulder Elbow Surg. 2001;10: 217-24.
3. , Avidor C, Krishnan SG, Walch G, Kempf JF, Mole D. Cemented polyethylene versus uncemented metal-backed glenoid components in total shoulder arthroplasty: a prospective, double-blind, randomized study. J Shoulder Elbow Surg. 2002;11: 351-9.
4. , Parsons IM, Campbell B, Titelman RM, Smith KL, Matsen FA. Patient functional self-assessment in late glenoid component failure at three to eleven years after total shoulder arthroplasty. J Shoulder Elbow Surg. 2005;14: 368-74.
5. , Greis PE, Bonutti PM. Treatment of symptomatic glenoid loosening following unconstrained shoulder arthroplasty. Orthopedics. 1999;22: 229-34.
6. , Norris TR. Influence of preoperative factors on outcome of shoulder arthroplasty for glenohumeral osteoarthritis. J Bone Joint Surg Am. 2003;85: 251-8.
7. , Zurakowski D, Thornhill TS. Uncemented glenoid component in total shoulder arthroplasty. Survivorship and outcomes. J Bone Joint Surg Am. 2005;87: 1284-92.
8. , Hawkins RJ. Revision of failed total shoulder arthroplasty. Orthop Clin North Am. 1998;29: 519-33.
9. , Cofield RH, Rowland CM. Neer hemiarthroplasty and Neer total shoulder arthroplasty in patients fifty years old or less. Long-term results. J Bone Joint Surg Am. 1998;80: 464-73.
10. , Edgerton BC. Total shoulder arthroplasty: complications and revision surgery. Instr Course Lect. 1990;39: 449-62.
11. , Sirveaux F, Roche O, Mole D, Boileau P, Walch G. [Results of revision surgery for glenoid loosening: a multicenter series of 37 shoulder prosthesis]. Rev Chir Orthop Reparatrice Appar Mot. 2004;90: 111-21. French.
12. , Segur JM, Poggio D, de Retana PF. Use of freeze-dried bone allograft with platelet-derived growth factor for revision of a glenoid component. J Bone Joint Surg Br. 2006;88: 1228-31.
13. , Barnes SA, Sperling JW, Farrell CM, Cummings JD, Cofield RH. Patient and physician-assessed shoulder function after arthroplasty. J Bone Joint Surg Am. 2006;88: 508-13.
14. , Watson KC, Stanton FJ. Recent experience in total shoulder replacement. J Bone Joint Surg Am. 1982;64: 319-37.
Copyright 2007 by The Journal of Bone and Joint Surgery, Incorporated
15. . Total shoulder arthroplasty with the Neer prosthesis. J Bone Joint Surg Am. 1984;66: 899-906.