SCIENTIFIC ARTICLE
Long-Term Results of Arthroscopic Wrist Synovectomy in Rheumatoid Arthritis Hyun Il Lee, MD, Keun Ho Lee, MD, Kyoung Hwan Koh, MD, Min Jong Park, MD, PhD
Purpose To investigate the effects of arthroscopic wrist synovectomy on the clinical course of rheumatoid arthritis in a large series with long-term follow-up. Methods We performed arthroscopic synovectomy on 56 wrists in 49 patients with rheumatoid arthritis. At a mean follow-up of 7.9 years (range, 5e12 y), we evaluated pain and patient satisfaction using a visual analog scale and assessed overall function using the Mayo wrist score. Radiographic stage was graded according to Larsen stage. We determined clinical outcomes on the recurrence of wrist synovitis, which we evaluated by symptoms of pain and swelling and physical examination. Preoperative variables were statistically analyzed to find factors that could influence the results. Results The mean visual analog scale score for wrist pain decreased from 6.3 to 1.7, and the mean Mayo wrist score (evaluated in 39 wrists) improved from 48 (range, 5e70) to 76 (range, 55e100). The mean visual analog scale score for patient satisfaction was 7.9. At final followup, synovitis was controlled in 42 wrists (75%) and recurred in the others. The mean Larsen stage progressed from 2.2 to 3.3. Analysis of preoperative variables revealed no factors that significantly affected clinical outcomes in terms of sex, age, duration of wrist symptom, preoperative serologic inflammatory markers, or Larsen stage. Conclusions Arthroscopic synovectomy of the wrist can provide pain relief and functional improvement with control of synovitis in 75% of rheumatoid wrists that have not responded to medication. (J Hand Surg Am. 2014;-:-e-. Copyright Ó 2014 by the American Society for Surgery of the Hand. All rights reserved.) Type of study/level of evidence Therapeutic IV. Key words Rheumatoid arthritis, wrist, arthroscopy, synovitis.
R
HEUMATOID ARTHRITIS
(RA) IS AN autoimmune-
induced synovitis affecting multiple joints by an unknown etiology. As the disease progresses, pain and dysfunction resulting from cartilage From the Department of Orthopaedic Surgery, Gangneung Asan Hospital, University of Ulsan College of Medicine, Gangneung; the Department of Orthopaedic Surgery, Samsung Medical Center, SungKyunKwan University School of Medicine, Seoul; and the Department of Orthopaedic Surgery, Inje University Ilsan Paik Hospital, Goyang, Korea. Received for publication December 22, 2013; accepted in revised form April 9, 2014. No benefits in any form have been received or will be received related directly or indirectly to the subject of this article. Corresponding author: Min Jong Park, MD, PhD, Department of Orthopedic Surgery, Samsung Medical Center, Sungkyunkwan University School of Medicine, Irwon-dong 50, Gangnam-Gu, Seoul 135-710, South Korea; e-mail:
[email protected]. 0363-5023/14/---0001$36.00/0 http://dx.doi.org/10.1016/j.jhsa.2014.04.022
destruction can markedly limit daily activities.1 Although basic treatment of RA is medication, which includes disease-modifying antirheumatic drugs and biologic agents, there are many patients with 1 or more joints resistant to medical treatment.2,3 In such cases, surgical synovectomy is an acceptable option by which to remove the uncontrolled inflammatory burden, especially in the wrist, elbow, and ankle.4e6 Arthroscopic technique renders surgical synovectomy more feasible by providing a minimally invasive approach with low morbidity and an early return to daily activity.6 Synovitis of the wrist is a frequent feature of RA.7 Within 2 years of a diagnosis of RA, more than half of patients will have wrist pain, and more than 90% will have wrist disease within 10 years.8 Although the
Ó 2014 ASSH
r
Published by Elsevier, Inc. All rights reserved.
r
1
2
ARTHROSCOPIC SYNOVECTOMY IN RHEUMATOID WRIST
short-term results for pain relief and functional improvement after arthroscopic synovectomy have been well described in previous reports,9e11 there is still controversy regarding its long-term efficacy in preventing joint destruction or recurrence. Because a large-volume prospective study found that open synovectomy in the knee and finger joints was not superior to nonsurgical control groups, many rheumatologists may believe that the surgery will not change the natural course of the disease.12 However, the recent success of arthroscopic synovectomy in many other joints led us to hypothesize that arthroscopic synovectomy of the wrist would yield satisfactory clinical outcomes with minimal morbidity on a long-term basis.4,5,13 Therefore, the purpose of the current study was to evaluate the long-term effect of arthroscopic synovectomy of rheumatoid wrists.
completed the follow-up visits and those who did not (data not shown). The average follow-up period was 7.9 years (range, 5e12 y). Our local ethics committee approved this study. Age at the time of surgery ranged from 21 to 69 years (average, 43 y). There were 11 men and 38 women. The average duration of RA was 5 years (range, 6 mo to 20 y), and the average duration of wrist symptoms was 23 months (range, 6 mo to 5 y). Tumor necrosis factor (TNF) inhibitors were not used preoperatively, and 13 patients (16 wrists; 29%) regularly used a TNF inhibitor at the final follow-up. Surgical technique The details of the operative procedures were described in previous reports.11,15 We performed arthroscopy using a 30 angled, 2.5-mm-diameter arthroscope. Standard 3e4 and 4e5 portals were used for the radiocarpal joint. Midcarpal-lateral and midcarpalmedial portals were used for the midcarpal joint. We also used the 6U portal as a viewing portal during synovectomy of the dorsal radiocarpal capsule. To achieve successful synovectomy of the midcarpal joint, an accessory portal for the scaphoid-trapeziumtrapezoid joint was used. All visible inflamed synovium was removed from the joint capsule with a motorized shaver. The distal radioulnar joint could be reached from the radiocarpal joint through a hole in the triangular fibrocartilage, which was present in all patients as a result of longstanding inflammation. We occasionally used separate portals for the distal radioulnar joint for the shaver while viewing the joint through the perforated triangular fibrocartilage. Plaster splints were removed on the third or fourth day after the operation. Formal physiotherapy was not required. After surgery, we recommended that all patients take antirheumatic medications continuously under the care of a rheumatologist.
MATERIALS AND METHODS Patient enrollment and demographics We retrospectively reviewed patients who underwent arthroscopic synovectomy for rheumatoid wrist between 2000 and 2007. Inclusion criteria were confirmed rheumatoid arthritis diagnosis by a rheumatologist, age over 18 years, no previous surgical procedures in the wrist, and painful swelling persisting for more than 6 months despite antirheumatic medication. Exclusion criteria were patients who had additional procedures such as Sauve-Kapandji, the Darrach procedure, or extensor tenosynovectomy, and patients with advanced arthritis with deformity (eg, Larsen stage 4 or 514). A total of 67 wrists from 58 patients (9 bilateral cases) met the criteria for this study and were enrolled. From electronic medical records, we extracted a pain visual analog scale (VAS) measure, wrist range of motion (ROM), and modified Mayo wrist score, which had been routinely recorded preoperatively. The patients were observed at 1-year intervals after surgery. By telephone, we regularly contacted patients who declined to visit the outpatient clinic, and an examination visit was recommended when symptoms worsened. We observed 56 wrists in 49 patients (7 bilateral cases) for a minimum of 5 years, and the overall final follow-up rate was 85%. The final follow-up visit was completed in 33 patients (39 wrists) in the outpatient setting. The remaining 16 patients (17 wrists) were followed up by telephone interview because most considered the clinic visit to be unnecessary, given the absence of discomfort in the wrist and the travel inconveniences. There were no statistical differences in age, sex, or preoperative radiographic grade of the wrist between patients who J Hand Surg Am.
Outcome measurements At the time of the final follow-up, we obtained detailed clinical evaluations as well as the pain VAS. Postoperative functional status (return to previous employment or daily activity) was also evaluated. We measured ROM and grip power when the patients visited the outpatient clinic. Patients were asked whether they would have elected to have the surgery if they had known in advance what the postoperative discomfort and the end result would be. Patient satisfaction for the surgery was also rated using a VAS. To determine whether the synovectomy was successful on a long-term basis, we used the following criteria: status A, no episode of painful swelling after r
Vol. -, - 2014
ARTHROSCOPIC SYNOVECTOMY IN RHEUMATOID WRIST
the operation; status B, intermittent painful swelling of the wrist with only short-term (< 1 mo) persistence; status C, prolonged painful swelling indicating recurrence of synovitis after a certain symptom-free period; and status D, persistent, painful swelling after surgery. We considered status A and B to be successful control of synovitis after synovectomy. Clinical failure was defined status C or D or when patients underwent additional procedures such as open or revision arthroscopic synovectomy in the same wrist. Radiological progression was not considered in the assessment of clinical success.
considered the entry point, and the time of clinical failure was used as the end point. We used kappa statistics to test intraobserver and interobserver reliability of the radiograph assessments. When the kappa value was more than 0.75, it represented excellent reliability; when it was between 0.4 and 0.75, it represents good reliability; and when it was below 0.4, it represented marginal reliability. RESULTS Clinical outcomes At the final follow-up visit, the mean VAS score for pain decreased from 6.3 (range, 3e9.5) to 1.7 (range, 0e8) (P < .001). The mean Mayo wrist score improved from 48 (range, 5e70) to 76 (range, 55e100) (P < .001). Based on the Mayo wrist score, excellent results were observed in 7 wrists (18%), good results in 10 (26%), fair results in 21 (54%), and poor results in 1 (3%). Forty-four patients (90%) were able to return to their previous levels of daily activity. Forty-two patients (86%) answered that they would undergo surgery again if presented with the same situation, and the mean VAS score of patient satisfaction of the surgery was 7.9 (range, 2e10). Mean preoperative ROM was 46 (range, 10 to 80 ) for wrist extension and 42 (range, 5 to 80 ) for wrist flexion. Postoperatively, the mean ROM was 51 (range, 10 to 80 ) for wrist extension and 44 (5 to 80 ) for wrist flexion. There was no evidence of recurrent wrist synovitis in 18 wrists, and there was intermittent painful swelling of short duration in 24 wrists, painful swelling recurring for a prolonged period in 11 wrists, and no improvement from surgery in 3 wrists after an average of 7.9 years of follow-up. Control of the synovitis (no recurrence group plus mild recurrence group) was achieved in 42 wrists (75%), whereas clinical failure (prolonged recurrence group plus surgical failure group) was observed in 14 wrists (25%). There were 3 patients who underwent additional surgical procedures owing to failure of the index surgery. Two patients had repeated arthroscopic synovectomy after 1 and 3 years, and another patient underwent a Sauve-Kapandji procedure after 6 years. When clinical failure was used as the end point, arthroscopic synovectomy achieved a 5-year survival rate of 76% and a 10-year survival rate of 65% according to Kaplan-Meyer survival analysis (Fig. 1). We observed no specific complications related to surgery.
Radiographic evaluation We compared pre- and postoperative radiographs. Larsen stage was determined based on a standard posteroanterior radiograph of the wrist.14 To evaluate interobserver and intraobserver variance in Larsen staging, 2 independent readers performed radiologic grading with time intervals of more than 2 weeks. Statistical analysis Numerical data are reported as the mean and range, and categorical data are reported as the absolute frequency and percentage distribution. We used the Shapiro-Wilk normality test to determine the distribution form. For numerical data, Student t or MannWhitney test was used according to the normality of distribution. For the analysis of categorical data, chisquare or Fisher exact test was performed. P < .05 was considered significant. We analyzed the following variables to determine predicting factors related to the success (control of synovitis) or failure of treatment (recurrence of synovitis): age at surgery, sex, symptom duration, radiological staging, and serologic inflammatory and rheumatoid markers (rheumatoid factor, C-reactive protein, and erythrocyte sedimentation rate). Wrists were divided into high or low groups relative to the median age and median duration of wrist synovitis for analysis. Furthermore, normal and abnormal values of the serologic markers (set point ¼ 42 IU/mL for rheumatoid factor, 22 mm/h for erythrocyte sedimentation rate, and 0.3 mg/dL for C-reactive protein) were organized into discrete groups for analysis. Wrists were divided into groups based on their baseline Larsen stage, resulting in early (grade 1 or 2) and advanced (grade 3) groups. We compared differences in control rates between groups defined above for each variable, and calculated significance using Fisher exact test. Kaplan-Meier survival analysis curves were constructed to describe the estimated probabilities of clinical success (control of synovitis). The surgery date was J Hand Surg Am.
3
Radiographic assessment Preoperatively, 34 wrists (61%) were Larsen stage 1 or 2 and 22 were stage 3 (39%). At the final follow-up, 13 wrists (23%) were Larsen stage 1 or 2, 21 (38%) were 3, r
Vol. -, - 2014
4
ARTHROSCOPIC SYNOVECTOMY IN RHEUMATOID WRIST
DISCUSSION We report the long-term clinical results after arthroscopic synovectomy for RA wrists refractory to medication. We achieved symptom improvement in most patients and satisfactory control of synovitis in 75% of wrists after a minimum follow-up of 5 years. A few reports previously described the clinical results of arthroscopic synovectomy in the rheumatoid wrist. Adolffson and Frisén16 reported reduced pain and improved wrist function in 24 wrists after a mean follow-up of 3.8 years. Of the 24 wrists, 17 (71%) had improved, 5 were unchanged, and 2 were worse according to the clinical scores. Wei et al10 reported good to excellent results in 13 patients, fair in 3, and poor in 1 at 12 months after office-based arthroscopic synovectomy. In the current study, we observed that 75% of wrists did not experience synovitis recurrence at the final follow-up. However, the wrist function score were not as good as we expected. The average Mayo wrist score was 76, and 56% of wrists showed a fair or poor result according to the Mayo wrist score. Although many objective measurements were available, including grip strength, motion, and the Mayo wrist score, which have been used as a reliable outcome in wrist surgeries, we decided to use recurrence of synovitis as the primary outcome for several reasons. First, a physical examination could not be conducted in patients followed up by telephone. Second, grip strength, which is also important in calculating the Mayo wrist score, is largely affected by the status of the finger joints. Third, limited ROM was less of a concern to patients with RA than pain and swelling, especially in patients with multiple joint involvement. Although synovectomy may arrest or slow progression and sometimes allows erosive lesions to heal, recurrence is always possible. Because arthroscopic synovectomy is not effective in all patients, we explored factors influencing clinical outcomes. Some authors suggested that clinical outcomes are better when synovectomy is performed at an earlier radiologic stage.4,17,18 Adolfsson and Frisén16 reported that progression of arthritic degeneration was significantly less common in patients with early-stage synovitis at the time of surgery. However, Kim et al19 found that radiographic stage was not related to clinical outcome. In a meta-analysis of either open or arthroscopic synovectomy, Chalmers et al20 reported that advanced preoperative radiographic findings did not correlate with higher pain scores or increased need for subsequent arthroplasty compared with minimal degenerative joint changes. Thirupathi et al21 reported that open
FIGURE 1: Survival probability curve for 56 wrists treated with arthroscopic synovectomy. The 5- and 10-year survival rates were 76% and 65%, respectively.
and 22 (39%) were 4 or 5. On average, Larsen stage progressed from 2.2 to 3.3. There was no radiologic progression in 17 wrists (30%). A total of 25 wrists (45%) showed a progression of 1 Larsen stage, and 14 wrists (25%) showed a progression by 2 stages or more. On average, Larsen stage progressed from 2.2 to 3.0 in the control of synovitis group, whereas it progressed from 2.0 to 4.0 in the clinical failure group. Kappa values for intraobserver agreement of the measurements were 0.53 (99% confidence interval, 0.30e0.76) for reader 1 and 0.63 (99% confidence interval, 0.40e0.86) for reader 2. The kappa value for interobserver reliability was 0.46 (P < .001), showing good inter- and intraobserver reliability. Analysis of prognostic factors When we divided patients into 2 groups (control of synovitis vs clinical failure) and analyzed according to sex, age at surgery, wrist symptom duration, serologic markers, and preoperative severity of the radiographic stage, we found no variable that was significantly different between groups (Table 1). Shorter symptom duration had a 71% control rate, compared with a 79% control rate for longer symptom duration (P ¼ .759). The rate of synovitis control of wrists with a lower radiologic stage was not superior to that of wrists with more advanced radiologic stages.
J Hand Surg Am.
r
Vol. -, - 2014
5
ARTHROSCOPIC SYNOVECTOMY IN RHEUMATOID WRIST
TABLE 1.
Relationship Between Preoperative Variables and Synovitis Control or Clinical Failure Control of Synovitis (n [%])
Clinical Failure (n [%])
Female (n ¼ 42)
30 (71%)
12 (29%)
Male (n ¼ 14)
12 (86%)
2 (14%)
Sex
P Value .48
Age, y
.76
42 (n ¼ 28)
20 (71%)
8 (29%)
> 42 (n ¼ 28)
22 (79%)
6 (21%)
12 (n ¼ 28)
20 (71%)
8 (29%)
> 12 (n ¼ 28)
22 (79%)
6 (21%)
Symptom duration, mo
.76
Erythrocyte sedimentation rate (mm/h)
.48
22 (n ¼ 14)
12 (86%)
2 (14%)
> 22 (n ¼ 42)
30 (71%)
12 (29%)
0.3 (n ¼ 20)
16 (80%)
4 (20%)
> 0.3 (n ¼ 36)
26 (72%)
10 (28%)
C-reactive protein (mg/dL)
.75
Rheumatoid factor (IU/mL)
.75
42 (n ¼ 21)
15 (71%)
6 (29%)
> 42 (n ¼ 35)
27 (77%)
8 (23%)
Stage 1 or 2 (n ¼ 34)
23 (68%)
11 (32%)
Stage 3 (n ¼ 22)
19 (86%)
3 (14%)
Larsen stage
.21
synovectomy in wrists can provide noteworthy pain relief, even in patients with advanced disease. Because the wrist is not a weight-bearing joint, disability or pain is not as severe as in knees and ankles with a similarly advanced stage of arthritis. Therefore, the benefits of synovectomy could be extended to patients with wrists of advanced radiologic stage. We also observed that patients with wrists of relatively advanced radiologic stage (Larsen stage 3) showed results comparable to those with wrists at earlier stages. An earlier radiological stage did not indicate better clinical outcomes or less recurrence, according to our current analysis. Although results in the literature remain inconsistent, our results support more aggressive intervention, even in wrists with advanced radiographic changes. In addition to the optimal time to intervene based on radiographic stage, there is also controversy about the ideal time to intervene with surgical synovectomy after a trial of medical management. Most surgeons recommend a several-month trial period of conservative therapy before synovectomy. Early intervention is advocated because there is evidence that a high proportion of patients experienced joint destruction in the first 1 to 3 years after disease onset.22 J Hand Surg Am.
Early intervention can prevent joint destruction, but a sufficient duration of medical treatment is a more important prerequisite given the chronic and systemic nature of RA. Currently, we believe that the best indication for surgical synovectomy is a single joint that is resistant to medical therapy for more than 6 months. Our study has several limitations. First, this was a retrospective case series, which was subject to several of the limitations of a level 4 study. It is difficult to be sure that arthroscopic synovectomy changes the natural course of this disease with a case series. Although there is debate about the efficacy of synovectomy, we believe that arthroscopic synovectomy can change the inflammatory course, at least in medically intractable wrists. Multicenter prospective studies with randomization will be required to further assess the efficacy of arthroscopic synovectomy. Second, antirheumatic medication could not be controlled after synovectomy because it was decided by rheumatologists based on the clinical course of the wrist and other joints. We admit that trials of new agents such as biologic TNF blockers might alter the clinical outcomes, and we cannot say that successful synovitis control solely resulted from the effects of r
Vol. -, - 2014
6
ARTHROSCOPIC SYNOVECTOMY IN RHEUMATOID WRIST
the synovectomy. However, the control rate of the patients who used TNF blockers at the final assessment was not different from that of patients who did not take a TNF blocker, which suggests that that TNF blocker usage did not seem to modify the end results (8 wrists recurred among 40 wrists of TNF blocker nonusers [20%] vs 6 wrists recurred among 16 wrists of TNF blocker users [38%]; P ¼ .190 by Fisher exact test). Third, we followed up a certain portion of patients with telephone interviews. We evaluated the synovitis status, which is the key element to assess the success or failure of the procedure, based on patient symptoms. We believe that an objective outcome measure is less important than how patients actually perceived their symptoms to have changed after surgical intervention.
8. Rizzo M, Cooney WP III. Current concepts and treatment for the rheumatoid wrist. Hand Clin. 2011;27(1):57e72. 9. Adolfsson L, Nylander G. Arthroscopic synovectomy of the rheumatoid wrist. J Hand Surg Br. 1993;18(1):92e96. 10. Wei N, Delauter SK, Beard S, Erlichman MS, Henry D. Officebased arthroscopic synovectomy of the wrist in rheumatoid arthritis. Arthroscopy. 2001;17(8):884e887. 11. Park MJ, Ahn JH, Kang JS. Arthroscopic synovectomy of the wrist in rheumatoid arthritis. J Bone Joint Surg Br. 2003;85(7): 1011e1015. 12. McEwen C. Multicenter evaluation of synovectomy in the treatment of rheumatoid arthritis: report of results at the end of five years. J Rheumatol. 1988;15(5):765e769. 13. Ishii K, Inaba Y, Mochida Y, Saito T. Good long-term outcome of synovectomy in advanced stages of the rheumatoid elbow. Acta Orthop. 2012;83(4):374e378. 14. Larsen A, Dale K, Eek M. Radiographic evaluation of rheumatoid arthritis and related conditions by standard reference films. Acta Radiol Diagn (Stockh). 1977;18(4):481e491. 15. Kim SM, Park MJ, Kang HJ, Choi YL, Lee JJ. The role of arthroscopic synovectomy in patients with undifferentiated chronic monoarthritis of the wrist. J Bone Joint Surg Br. 2012;94(3): 353e358. 16. Adolfsson L, Frisén M. Arthroscopic synovectomy of the rheumatoid wrist: a 3.8 year follow-up. J Hand Surg Br. 1997;22(6):711e713. 17. Fuerst M, Fink B, Ruther W. Survival analysis and longterm results of elbow synovectomy in rheumatoid arthritis. J Rheumatol. 2006;33(5): 892e896. 18. Roch-Bras F, Daures JP, Legouffe MC, Sany J, Combe B. Treatment of chronic knee synovitis with arthroscopic synovectomy: longterm results. J Rheumatol. 2002;29(6):1171e1175. 19. Kim SJ, Jung KA, Kim JM, Kwun JD, Kang HJ. Arthroscopic synovectomy in wrists with advanced rheumatoid arthritis. Clin Orthop Relat Res. 2006;449:262e266. 20. Chalmers PN, Sherman SL, Raphael BS, Su EP. Rheumatoid synovectomy: does the surgical approach matter? Clin Orthop Relat Res. 2011;469(7):2062e2071. 21. Thirupathi RG, Ferlic DC, Clayton ML. Dorsal wrist synovectomy in rheumatoid arthritis—a long-term study. J Hand Surg Am. 1983;8(6): 848e856. 22. Rubbert-Roth A. Assessing the safety of biologic agents in patients with rheumatoid arthritis. Rheumatology (Oxford). 2012;51(suppl 5): v38ev47.
REFERENCES 1. Conaghan PG, O’Connor P, McGonagle D, et al. Elucidation of the relationship between synovitis and bone damage: a randomized magnetic resonance imaging study of individual joints in patients with early rheumatoid arthritis. Arthritis Rheum. 2003;48(1):64e71. 2. Lie E, van der Heijde D, Uhlig T, et al. Treatment strategies in patients with rheumatoid arthritis for whom methotrexate monotherapy has failed: data from the NOR-DMARD register. Ann Rheum Dis. 2011;70(12):2103e2110. 3. Ghattas L, Mascella F, Pomponio G. Hand surgery in rheumatoid arthritis: state of the art and suggestions for research. Rheumatology (Oxford). 2005;44(7):834e845. 4. Choi WJ, Choi GW, Lee JW. Arthroscopic synovectomy of the ankle in rheumatoid arthritis. Arthroscopy. 2013;29(1):133e140. 5. Kang HJ, Park MJ, Ahn JH, Lee SH. Arthroscopic synovectomy for the rheumatoid elbow. Arthroscopy. 2010;26(9):1195e1202. 6. Adolfsson L. Arthroscopic synovectomy of the wrist. Hand Clin. 2011;27(3):395e399. 7. McInnes IB, Schett G. The pathogenesis of rheumatoid arthritis. N Engl J Med. 2011;365(23):2205e2219.
J Hand Surg Am.
r
Vol. -, - 2014