Degenerative joint disease of the shoulder, while

Arthroscopic Debridement of the Shoulder for Osteoarthritis David M. Weinstein, M.D., John S. Bucchieri, M.D., Roger G. Pollock, M.D., Evan L. Flatow, M.D., and Louis U. Bigliani, M.D. Summary: Twenty-five patients underwent arthroscopic debridement to treat early glenohumeral osteoarthritis. The group consisted of 19 men and 6 women with an average age of 46 years (range, 27 to 72 years.) The operative procedure consisted of lavage of the glenohumeral joint, debridement of labral tears and chondral lesions, loose body removal, and partial synovectomy and subacromial bursectomy. Follow-up averaged 34 months, with a range of 12 to 63 months. Overall, results were rated as excellent in 2 patients (8%), good in 19 patients (72%), and unsatisfactory in 5 (20%). Two patients had complete relief of pain, 18 patients had only occasional mild pain, and 5 had moderate to severe pain postoperatively. Of the 12 patients with marked preoperative stiffness, 10 (83%) had improvement in range of motion postoperatively. Arthroscopic debridement is a reasonable approach for treating early glenohumeral osteoarthritis that has failed to respond to nonoperative treatment, in which the humeral head and glenoid remain concentric, and where there is still a visible joint space on an axillary radiograph. The procedure is not recommended when there is severe joint incongruity or large osteophytes. Key Words: Glenohumeral joint Osteoarthritis Shoulder arthroscopy. Degenerative joint disease of the shoulder, while not occurring as frequently as in the knee or hip joint, is relatively common. Treatment of symptomatic osteoarthritis of the shoulder is nonoperative in the majority of patients. If conservative treatment fails to relieve symptoms, operative treatment is considered. As open debridement of the joint or realignment procedures have been found to be unsuccessful, operative treatment has generally centered around replacement of the joint. 1-4 Pathology in patients with glenohumeral osteoarthritis includes a degenerative labrum, loose bodies, articular cartilage defects, and osteophytes. In addition, there is often coexisting pathology, including adhesive capsulitis, glenohumeral instability, and subacromial inflammation. 2,5-9 Arthroscopic surgery may From the Premiere Orthopaedic Group, Colorado Springs, Colorado (D.M.W.); and The Shoulder Service, New York Orthopaedic Hospital, The New York Presbyterian Hospital, Columbia- Presbyterian Medical Center, New York, New York, U.S.A. Address correspondence and reprint requests to Louis U. Bigliani, M.D., 622 West 168th St, PH1130, New York, NY 10032, U.S.A. 2000 by the Arthroscopy Association of North America 0749-8063/00/1605-2172$3.00/0 doi:10.1053/jars.2000.5042 be a reasonable treatment option in these patients after conservative methods have been unsuccessful and when joint replacement is not desired. Although arthroscopy has become a well-established technique in the treatment of many shoulder disorders, the role of arthroscopy in the management of degenerative conditions has not been well delineated. 3,4,7,10-14 Several studies have recommended this as a treatment option, but only sparse follow-up data have been reported. The purpose of this study was to determine the extent and duration of symptomatic relief offered by arthroscopic debridement of the glenohumeral joint for early glenohumeral arthritis. MATERIALS AND METHODS From 1986 to 1992, 25 patients underwent shoulder arthroscopy with arthroscopic debridement to treat glenohumeral osteoarthritis. All procedures were performed at the Shoulder Service at the New York Orthopaedic Hospital, The New York Presbyterian Hospital, Columbia-Presbyterian Medical Center Campus. Inclusion in the study required a postoperative diagnosis of degenerative arthritis of the shoulder at the time of arthroscopy and a minimum of 12 months Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 16, No 5 (July-August), 2000: pp 471 476 471

472 D. M. WEINSTEIN ET AL. of postoperative follow-up. The study group consisted of 19 men and 6 women. The age at surgery ranged from 27 to 72 years, with an average of 46 years. The dominant extremity was involved in 56% of the cases. Follow-up averaged 34 months with a range of 12 to 63 months. Eight patients (32%) had a history of previous injury, including 2 patients with proximal humerus fractures and 5 patients with a history of anterior dislocation. One patient had a history of multiple subluxations. In addition, 8 patients (32%) had undergone previous surgical treatment of the shoulder. This represented a total of 15 procedures, including 6 anterior stabilizations, 3 arthroscopies, 6 rotator cuff repairs, and 1 open acromioplasty. Preoperatively, all patients complained of chronic aching pain exacerbated by activity. Seven patients were able to participate in recreational sporting activity; 1 patient played professional sports but was limited because of pain. Thirteen patients had previously participated in sports but were unable to do so because of their shoulder symptoms. At the preoperative physical examination, 13 patients had full range of motion and 12 patients had moderate to severe restricted motion with an average of 150 of forward elevation, 35 of external rotation, and internal rotation to L5. Radiographs were classified according to joint changes for comparison purposes. Stage I included normal radiographs. The diagnosis of glenohumeral osteoarthritis in this group was made at the time of arthroscopy, where moderate to severe articular cartilage damage was seen. Stage II changes included minimal joint space narrowing with a concentric head and glenoid (Fig 1). Stage III included moderate joint space narrowing with early inferior osteophyte formation (Fig 2). Stage IV changes included severe loss of joint space with osteophyte formation and loss of concentricity between the humeral head and glenoid. No patients were included in this study with stage IV changes, as this is felt to be a contraindication to arthroscopic debridement. Based on the preoperative evaluation, 16 patients had been diagnosed with osteoarthritis, 4 patients with impingement, 4 patients with a frozen shoulder, and 1 patient with glenohumeral instability. All of these patients had undergone failed nonoperative treatment for at least 3 months with nonsteroidal anti-inflammatory medication, exercises, and/or formal physical therapy. The time from initial visit to surgery ranged from 3 to 60 months with an average of 23 months. Operative reports were reviewed for the presence and number of specific pathologic entities and modes of treatment. Degeneration or changes of the articular cartilage were categorized into 4 groups: grade I, softening or blistering of the articular surface; grade II, fissure and fibrillation; grade III, deep ulceration of the articular cartilage; and grade IV, full-thickness cartilage loss with exposed subchondral bone (Fig 3). Postoperative evaluation consisted of patient interview and evaluation. Patients were questioned regarding changes in pain, function, motion, and overall subjective satisfaction with the results of the treatment. FIGURE 1. (A) Anteroposterior and (B) axillary radiographs of a patient with stage II osteoarthritic changes, including minimal joint space narrowing with a concentric humeral head and glenoid.

SHOULDER DEBRIDEMENT FOR OSTEOARTHRITIS 473 FIGURE 2. (A) Anteroposterior and (B) axillary radiographs of a patient with stage III osteoarthritis changes, demonstrating moderate joint space narrowing and early osteophyte formation. The results were rated as excellent if the patient had no pain, full use of the extremity, and essentially normal motion and strength. Results were classified as good if the patient was satisfied with only an occasional or mild pain in the shoulder, had elevation of more than 130, and had full strength. An unsatisfactory result was one that failed to meet these criteria. Statistical analysis was performed using a Wilcoxon rank test, Kruksal-Wallis test, and a paired t test when indicated. FIGURE 3. Schematic and arthroscopic photographs showing articular cartilage grading system. (A) Grade II articular changes, fissuring and fibrillation; (B) grade III articular changes, showing deep ulceration of the articular cartilage; and (C) grade IV articular changes, demonstrating full thickness cartilage loss with exposed subchondral bone.

474 D. M. WEINSTEIN ET AL. OPERATIVE TECHNIQUE Interscalene regional anesthesia was used and the patient was positioned in a beach-chair position. Standard anterior and posterior arthroscopic portals and a 30 arthroscope were used. Arthroscopic treatment included lavage, debridement of degenerative labral tears and chondral lesions, loose body removal, and partial synovectomy and osteophytectomy, depending on the precise pathology encountered in each shoulder. The subacromial space was examined using anterolateral and posterolateral portals. In 23 patients, a bursectomy was performed for the presence of chronic bursal thickening. Eight patients were found to have coexisting intra-articular pathology, including 5 labral tears, 2 partial-thickness rotator cuff tears, and 2 SLAP lesions. Three patients had loose bodies that were removed. In addition, 1 patient underwent an arthroscopic acromioclavicular resection, 1 patient underwent an arthroscopic acromioplasty, and 1 patient underwent both an arthroscopic acromioclavicular resection and acromioplasty. Postoperatively, patients were placed in a sling for 1 or 2 days. Passive range of motion was started on the first postoperative day and advanced as tolerated. Patients were allowed to return to their regular recreational activities after 4 to 6 weeks. RESULTS All 25 patients reported at least some improvement in pain, with a significant amount of pain relief noted between the preoperative and postoperative evaluations (P.05). Two patients had complete relief of pain, 18 patients had occasional to mild pain, and 5 patients had moderate to severe pain. At the final follow-up, 19 patients (76%) had maintained pain relief. Six patients noted deterioration in the pain relief, but only 2 patients reported that the pain had returned to that of the preoperative level. All 6 patients with deterioration of pain relief reported at least 7 months of pain relief with an average of 12 months and a range of 7 to 36 months. Of the 12 patients with marked preoperative stiffness, 10 patients (83%) had improvement in their range of motion following the arthroscopic treatment. Two patients reported no improvement in their motion after surgery. The active range of motion improved significantly in both forward elevation and external rotation and averaged 167 in forward elevation, external rotation to 53, and internal rotation to L5 (P.05). In the group of 13 patients who were not TABLE 1. Radiographic Changes Versus Outcome: Final Results Radiographic Stage Excellent Good Unsatisfied I 1 4 0 II 1 5 2 III 0 9 3 IV 0 0 0 able to play sports before their surgery, postoperatively 7 were able to return to their previous recreational activities. In evaluating age and gender, there was no significant difference in pain relief between patients older than 40 years or younger than 40 years (P.05). There was also no difference between male and female patients (P. 05). Preoperative radiographic changes were also compared with clinical outcome. Preoperative radiographs were normal in 5 shoulders, and showed stage II changes in 9 patients and stage III changes in 11 patients (Table 1). No patients had stage IV changes, as arthroscopy is not performed by us in patients with severe and noncongruent arthritis. There was no significant correlation between radiographic grade and clinical outcome (P.05). The degree of damage to the articular cartilage at the time of arthroscopy was also correlated with clinical outcome. No patient had findings consistent with grade I changes. Six patients had grade II findings, 11 patients had grade III findings, and 8 patients had exposed subchondral bone or grade IV changes. Similar to the radiographic findings, there was no statistically significant correlation between cartilage damage and clinical outcome (P.05). There was a definite trend toward worse results with increasing severity of articular cartilage changes. There were no unsatisfactory results in the 6 patients with grade II findings only, whereas 5 of the 19 patients (26%) with grade III or IV changes had unsatisfactory results (Table 2). Overall, the results were rated as excellent in 2 patients (8%), good in 19 patients (72%), and unsatisfactory in 5 patients (20%). Pain appeared to be the most important factor in determining a patient s final TABLE 2. Articular Findings Versus Outcome: Final Results Cartilage Grade Excellent Good Unsatisfied I 0 0 0 II 0 6 0 III 1 8 2 IV 1 4 3

SHOULDER DEBRIDEMENT FOR OSTEOARTHRITIS 475 rating. The only 2 patients with excellent results had complete resolution of symptoms. Five patients were rated as unsatisfactory, based on moderate to severe pain noted at their last follow-up (Table 3). Twentythree of the patients (92%) were satisfied with the procedure and stated that they would undergo arthroscopic debridement again. There were no infections or complications with wound healing. No reoperations have been performed. However, 3 patients are pending further surgical treatment, including 2 patients who have had pain relief for 42 months and 8 months, respectively, but now have increasing glenohumeral symptoms. Because both patients are young and obtained good relief with the first procedure, another arthroscopic debridement has been recommended. A third patient is scheduled for total shoulder arthroplasty. It has been 56 months since this patient had her arthroscopic debridement and she has been able to postpone her total shoulder arthroplasty up to this point, but presently her pain has returned to its preoperative level. DISCUSSION Preliminary studies have shown that arthroscopy may play a role in the treatment of glenohumeral osteoarthritis. Ellman 12 described using arthroscopic examination to identify coexisting osteoarthritis in a group of patients diagnosed with impingement syndrome. In this study, findings included grade II or III chondromalacia of the humeral head or glenoid, and arthroscopic debridement led to early pain relief in short-term follow-up. Ogilvie-Harris and Wiley 14 reported on their 10-year experience with shoulder arthroscopy; a subgroup of patients with degenerative changes of the shoulder joint were treated with arthroscopic debridement with subsequent good results. Both Cofield 15 and Johnson 7 also reported on small groups of patients treated with arthroscopic lavage and debridement with subsequent relief of symptoms, but the results were not quantified. In our series, arthroscopic examination of the glenohumeral joint proved helpful in diagnosing osteoarthritis and coexisting pathology. In the majority of cases, preoperative clinical findings and radiographic changes were consistent with osteoarthritis, but 9 patients had TABLE 3. Overall Rating Excellent 2 (8%) Good 18 (72%) Unsatisfactory 5 (20%) other diagnoses before surgery, including frozen shoulder, impingement, and instability. These patients were considered misdiagnosed, or osteoarthritis was diagnosed at the time of surgery and was believed to be the cause of the patients symptoms. In addition, 32% of the patients had additional intra-articular pathology, including labral tears, loose bodies, SLAP lesions, and partial-thickness articular surface rotator cuff tears. This subgroup is similar to the group reported by Ellman, in which 18 patients with the preoperative diagnosis of impingement were found to have coexisting osteoarthritis at the time of arthroscopy. Arthroscopic debridement was also particularly effective in providing pain relief. All patients reported initial pain relief lasting a minimum of 7 months, and only 2 patients felt the symptoms were as severe as before arthroscopy. The amount of pain relief was also significant. The beneficial pain relief tended to be stable in the majority of cases with only 24% of the patients noting a deterioration over time. The mechanism of pain relief following debridement is unknown. It has been hypothesized that the removal of debris, which can cause synovitis, as well as the dilution of degenerative enzymes, may contribute to the improvement found after surgery. In addition, the removal of mechanical factors such as chondral flaps, loose bodies, and osteophytes may improve joint function. Arthroscopic debridement of the subacromial space probably contributes to pain reduction and improved range of motion. A consistent finding in 23 of the 25 patients was a thickened subacromial bursa. Whether the bursa thickens as a result of an inflamed restricted glenohumeral joint is unknown. In this series, a bony spur or hooked acromion was seen in only 2 patients and probably was not a factor in the development of the coexisting bursitis in most of these cases. Functional improvement was noted particularly in the patients with significant shoulder stiffness. The osteoarthritic debris was removed and motion was restored. Unlike Johnson s findings, motion was significantly improved in 83% of the patients with preoperative stiffness. 9 The largest gains were attained for external rotation. Also, a return to sporting activity was achieved in 7 of the 13 patients who had been restricted preoperatively. An attempt was made to discern the factors that could be assessed preoperatively, which affected the results. The age of the patient did not affect the outcome, with younger patients (under 40 years) faring as well as older patients. Gender, duration of symptoms, previous surgery, and radiographic stage were not found to correlate with success of the procedure.

476 D. M. WEINSTEIN ET AL. The relationship between the severity of osteoarthritis and functional results after arthroscopic debridement is not clear. Both Cofield 15 and Ogilvie-Harris and Wiley 14 found more favorable results with lesser degrees of degeneration. Ogilvie-Harris and Wiley found successful results in two thirds of their patients when degeneration was mild; this rate dropped to one third good results when the degeneration was severe. In our study, although we were unable to show a statistically significant relationship between specific radiographic or operative findings and eventual outcome, a definite trend was noted for a worse outcome associated with increasing severity of articular cartilage damage. Arthroscopic debridement of the shoulder has relatively few complications and a low morbidity rate. 15 In this series, there were no complications. In addition, if arthroscopic debridement fails to relieve the patient s symptoms, the procedure does not jeopardize any further surgical treatment. Overall, arthroscopic debridement yielded 80% excellent or good results, and only 3 of the 25 patients (12%) have had further surgery recommended. Although there were few patients completely pain free at follow-up, the vast majority of patients did have significant improvement postoperatively with only mild symptoms and good range of motion. Ninety-two percent of patients were satisfied with their procedure and said that they would have the operation again. Admittedly, these are relatively short-term results, with an average follow-up of only 34 months. We believe that arthroscopic debridement of glenohumeral arthritis is a reasonable approach for treating concentric arthritis in which there is still a visible joint space on an axillary radiograph. Although the long-term results of this procedure are as yet unknown, early symptomatic improvement was seen in all patients. This procedure is especially useful in younger patients who wish to remain active and yet avoid prosthetic replacement. We do not recommend arthroscopy as a primary treatment before standard conservative measures have been attempted. Also, this procedure is not recommended in patients with severe joint incongruity, loss of joint space, or large osteophytes. REFERENCES 1. Bigliani LU, Nicholson GP, Flatow EL. Arthroscopic resection of the distal clavicle. Orthop Clin North Am 1993;24:133-141. 2. Neer CS II. Shoulder reconstruction. Philadelphia: WB Saunders, 1990;160-212. 3. Neer CS II. Replacement arthroplasty for glenohumeral osteoarthritis. J Bone Joint Surg Am 1974;56:1-13. 4. Rockwood CA, Matsen RA. The shoulder. Philadelphia: WB Saunders, 1990;703-713. 5. Ellman HA, Harris E, Kay S. Early degenerative joint disease simulating impingement syndrome: Arthroscopic findings. Arthroscopy 1992;8:482-487. 6. Jackson RW. Arthroscopic treatment of degenerative arthritis. Oper Arthrosc 1991;22:319-323. 7. Johnson CC. The shoulder joint: An arthroscopic perspective of anatomy and pathology. Clin Orthop 1987;223:113-125. 8. Matthews LS,Wolock BS, Martin DC. Arthroscopic management of degenerative arthritis of the shoulder. Oper Arthrosc 1991;45:567-572. 9. Small NC. Complications in arthroscopic surgery performed by experienced arthroscopists. Arthroscopy 1988;4:215-221. 10. Ark JW, Flock TJ, Flatow EL, Bigliani LU. Arthroscopic treatment of calcific tendonitis of the shoulder. Arthroscopy 1992;8:183-188. 11. Benjamin A, Hirschowitz D, Arden GP. The treatment of arthritis of the shoulder joint by double osteotomy. Orthopedics 1979;3:211-216. 12. Ellman HA. Arthroscopic subacromial decompression: Analysis of 1-3 year results. Arthroscopy 1987;3:173-181. 13. Flatow EL, Cordasco FA, Bigliani LU. Arthroscopic resection of the outer end of the clavicle from a superior approach: A critical, quantitative, radiographic assessment of bone removal. Arthroscopy 1992;8:55-64. 14. Ogilvie-Harris DJ, Wiley AM. Arthroscopic surgery of the surgery. J Bone Joint Surg Br 1986;60:201-207. 15. Cofield RH. Arthroscopy of the shoulder. Mayo Clin Proc 1983;58:501-508.