- Email: [email protected]
Site-Specific Intraoperative Efficacy of Arthroscopic Knee Joint Synovectomy in Rheumatoid Arthritis
Site-Specific Intraoperative Efficacy of Arthroscopic Knee Joint Synovectomy in Rheumatoid Arthritis Hans-Dieter Carl, M.D., Stefan Klug, M.D., Johannes Seitz, M.D., Bernd Swoboda, M.D., and Raimund W. Kinne, M.D.
Purpose: To assess the intraoperative reduction of inflammatory infiltrates achieved by arthroscopic knee joint synovectomy in patients with rheumatoid arthritis (RA) with special regard to the removal site, using preoperative and postoperative synovial tissue (ST) samples. Type of Study: A histologic and immunohistochemical study. Methods: Eleven patients with treatment-refractory RA knee synovitis underwent arthroscopic synovectomy. In each patient, ST specimens were obtained immediately before and after synovectomy from 9 defined sites covering the whole joint. The samples were graded using an acute synovitis score (ASS; presence of polymorphonuclear neutrophilic leukocytes [PMN] and fibrin) and a chronic synovitis score (CSS; e.g., lining cell hyperplasia, presence of diffuse and lymphoid aggregates). Immunohistologic analyses were performed using 7 monoclonal antibodies directed against PMN, macrophages, and T-cell subsets (total of 1,584 preparations). Knee function was assessed after an average follow-up of 28 months by Lysholm score (modified by Klein and Jensen), Insall functional and knee scores, and Lequesne score. Results: Arthroscopic synovectomy led to an overall significant (P between .005 and .05) reduction of the acute inflammatory infiltrates (ASS) by 82.1%, but to a significant reduction of chronic inflammatory infiltrates (CSS) by only 62.5%. Accordingly, the density of PMN was reduced by 81.8%, whereas that of macrophages and different T-cell subsets was only decreased by ⱕ61.6%. With respect to the anatomic regions, a significantly (P ⱕ .05) less marked reduction of inflammatory infiltrates was observed in the upper lateral and central recess, at the medial and lateral capsule, as well as at the femoral insertion of the anterior cruciate ligament. All knee joint scores showed a significant (P ⱕ .01) improvement over preoperative values at follow-up. Conclusions: Arthroscopic synovectomy effectively reduces acute and chronic inflammatory infiltrates in patients with RA who have refractory synovitis of the knee joint (immediately after synovectomy) and improves knee function (28-month follow-up). However, the reduction of inflammatory infiltrates appears to depend on the anatomic region of the joint. Level of Evidence: Level III. Key Words: Rheumatoid arthritis—Arthroscopy—Synovectomy—Knee inflammation.
I
From the Division of Orthopedic Rheumatology, Department of Orthopedic Surgery, Friedrich Alexander University of ErlangenNuremberg (H-D.C., S.K., J.S., B.S.), Erlangen; and the Experimental Rheumatology Unit, Friedrich Schiller University Jena (R.W.K.), Jena, Germany. Address correspondence and reprint requests to Hans-Dieter Carl, M.D., Division of Orthopedic Rheumatology, Department of Orthopedic Surgery, Friedrich Alexander University of ErlangenNuremberg, Rathsberger Strasse 57, D-91054 Erlangen, Germany. E-mail: [email protected] © 2005 by the Arthroscopy Association of North America 0749-8063/05/2110-4452$30.00/0 doi:10.1016/j.arthro.2005.06.014
n patients with rheumatoid arthritis (RA), sustained refractory synovitis for a time period of more than 6 months despite adequate antirheumatic therapy represents an indication for synovectomy.1 Because the highest rate of success, defined as clinical improvement or delay of radiologic progression, has been observed in patients with limited radiologic changes, synovectomy is recommended in early stages of the disease.2-6 In a retrospective, midterm, multicenter study, the combination of arthroscopic and radiation synovectomy in early cases of RA has proven superior to either single method,7 possibly the result of a more
Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 21, No 10 (October), 2005: pp 1209-1218
1209
1210
H.-D. CARL ET AL. TABLE 1.
Clinical Characteristics of the RA Patients at the Time of Arthroscopic Synovectomy
No. of Patients
Gender (M/F)
Age (yr/range)
Disease Duration (yrs/range)
RF (⫹/⫺)
No. of ARA Criteria (RA)
Concomitant Medication
11
2/9
51.9 ⫾ 3.4 (22-63)
9.3 ⫾ 2.0 (1-20)
9/2
all ⬎4
MTX (n ⫽ 9) AZT (n ⫽ 2) NSAIDs (n ⫽ 11)
NOTE. For the parameters age and disease duration means ⫾ SEM are given. Abbreviations: RF, rheumatoid factor; ⫹/⫺, positive/negative; ARA, American Rheumatism Association (now the American College of Rheumatology); MTX, methotrexate; AZT, azathioprine, NSAIDs, nonsteroidal anti-inflammatory drugs.
thorough removal of inflamed synovial tissue (ST). In fact, the degree of clinical improvement and the postoperative prognosis of synovectomy appears to depend on the histologic composition of the rheumatoid ST8,9 and the extent of ST removal.9-13 Based on these findings, the present study sought to analyze: (1) the inflammatory cell infiltrates present in refractory synovitis of rheumatic knee joints before arthroscopic synovectomy; and (2) the efficacy of arthroscopic synovectomy in removing synovial infiltrates (as assessed immediately after synovectomy), with particular regard to the anatomic region within the knee joint. The hypothesis underlying the present study was that the extent of ST removal by arthroscopic synovectomy displays site-specific differences within the knee joint.
Surgical Procedure Arthroscopic synovectomy was carried out following previously defined protocols.1,16-17 ST samples were obtained from 9 defined sites within the knee joint (Fig 1) before synovectomy was performed and again immediately after its completion using biopsy pincers. In all cases, multiportal synovectomy was performed through the 4-6 portals using motorized shaver systems. For the procedure, a tourniquet was used and the leg was placed in a hanging position. Postoperatively, limited weight bearing (approximately 15 kg) as taught by a physiotherapist was allowed on the operated leg for 4 weeks, and physiotherapy (both active and passive) along with continuous passive motion exercises were used. Diseasemodifying antirheumatic drug therapy was not interrupted or modified during surgery or recovery.
METHODS Patients
Histologic Evaluation
ST was obtained from 11 patients with RA during arthroscopic knee synovectomy. RA was classified according to the American College of Rheumatology criteria14 (Table 1). Of the 11 RA patients, 3 had increased erythrocyte sedimentation rates (49 ⫾ 27 mm/h; mean ⫾ SEM) and C-reactive protein values. Septic arthritis was excluded by synovial fluid analysis and Gram staining of synovial fluid and ST samples. Radiologic changes were not higher than Larsen stage II for any operated knee joint.15 All patients had refractory synovitis of the knee and all had received constant disease-modifying antirheumatic drug therapy; monotherapy with methotrexate (15 mg per week) in 9 patients and a combination therapy of methotrexate and azathioprine (200 or 300 mg daily) in 2 patients (Table 1) for a minimum of 6 months. None of the patients received steroids before, during, or after surgery; nonsteroidal anti-inflammatory drugs were administered on demand.
Following standard hematoxylin and eosin staining, the extent of synovitis was semiquantitatively assessed using a scoring system initially proposed by Geiler and Stiehl18 (Table 2), with 0 ⫽ absence of parameter, 1 ⫽ weak intensity, 2 ⫽ medium intensity, and 3 ⫽ strong intensity for the given parameter. The scores for the individual parameters of acute synovitis (presence of polymorphonuclear neutrophilic leukocytes [PMN] and fibrin deposits; maximum score of 6) and chronic synovitis (lining cell hyperplasia, diffuse infiltrates, lymphoid aggregates, germinal centers, plasma cells, formation of vessels; maximum score of 18) were then added. For technical reasons (mainly lack of clear histologic structures), some specimens were not suitable for reliable analysis, which reduced the total number of samples from 198 (11 patients; 9 preoperative and 9 postoperative samples) to 180 for the acute synovitis score and 186 for the chronic synovitis score.
SYNOVECTOMY IN RHEUMATOID ARTHRITIS
1211
FIGURE 1. Preoperative and postoperative ST samples ( ⫽ synovial membrane; ⫽ synovial membrane/bone punch) were obtained from 9 different locations covering the lateral, medial, superior, and anterior part of the knee joint (exemplified for the right knee joint). Individual locations were: I ⫽ upper lateral recess, highest point; II ⫽ upper lateral recess, cartilage/pannus junction; III ⫽ upper central recess, cartilage/pannus junction, bone punch; IV ⫽ upper medial recess, cartilage/pannus junction; V ⫽ upper medial recess, highest point; VI ⫽ femur insertion ACL, cartilage/pannus junction, inter-condylar bone punch; VII ⫽ lateral joint capsule, above meniscus; VIII ⫽ tibia insertion ACL, fibro-cartilage/pannus junction; IX ⫽ medial joint capsule, above meniscus.
at room temperature. All subsequent steps were performed at room temperature in a humid chamber. Specific marker antibodies directed against PMN, macrophages, and T-cell subsets22 (Table 3) were diluted in Tris-buffered saline and added to the sections for 30 minutes. Following 3 washes for 5 minutes each with phosphate-buffered saline (PBS; 0.9% NaCl, 30 mmol/L KCl, 70 mmol/L Na2PO4, and 10 mmol/L KH2PO4, pH 7.4), a peroxidase-coupled rabbit-anti-mouse antibody (1:300 in PBS) was applied for 30 minutes. The sections were washed and then incubated for 30 minutes with a peroxidase-coupled swine-anti-rabbit antibody (1:100 in PBS). The peroxidase was developed in a solution of 5 mg diaminobenzidine (Sigma, Deisenhofen, Germany) in 10 mL PBS with a pH of 7.4. Shortly before use, the solution was filtered and 30 L of 3% H2O2 were added. The same staining procedure was performed for controls, but the specific antibodies were either omitted or replaced by isotype controls in matched concentrations (Table 3, in both cases with negative results). The density of the different inflammatory cell types (PMN, macrophages, and T-cell subsets) in the ST samples was semiquantitatively assessed using a scoring system ranging from 0 to 3, with 0 ⫽ no positive cells, 1 ⫽ less than 40% positive cells, 2 ⫽ 40% to 90% positive cells, and 3 ⫽ more than 90% positive cells. Using 7 different monoclonal antibodies and a negative control (8 sections), as well as 9 preoperative and 9 postoperative samples from 11 patients with RA (198 samples), a total of 1,584 preparations was assessed (144 specimens per knee joint). Although biopsy samples were obtained from all areas in all joints, some specimens were not suitable for reliable analysis for technical reasons (mainly failure of anti-
Immunohistochemistry
TABLE 2.
For immunohistochemical analysis, ST was embedded in Tissue-Tek O.G.T. compound (Diatec, Nuremberg, Germany), immediately frozen in isopentane (Merck, Darmstadt, Germany) cooled in liquid nitrogen, and stored at ⫺70°C. Serial cryostat sections (4-m-thick) were cut in analogy to previously published procedures19-21 using a low open-top microtome (No. 5030; Bright, Huntingdon, England) and both hematoxylin and eosin and immunohistologic staining was performed. Briefly, sections were air-dried and then fixed with acetone and chloroform (Merck) for 10 minutes each
Criteria for the Quantification of Inflammatory Activity
ASS (Actual Activity)
CSS (Basic Activity)
Presence of PMN (0-3 points) Lining cell hyperplasia (0-3 points) Fibrin deposits (0-3 points) Diffuse infiltrates (0-3 points) Maximum score 6 points Lymphoid aggregates (0-3 points) Germinal centers (0-3 points) Plasma cells (0-3 points) Formation of vessels (0-3 points) Maximum score (18 points) NOTE. The criteria for the quantification of the acute and chronic inflammatory activity in RA ST are listed as previously defined.18
1212
H.-D. CARL ET AL. TABLE 3.
Antibodies Used for Immunohistochemical Study
Antibodies (clone)
Antigen Recognized
Main Cellular Expression (as previously published)
Dilution
Source
IgG1 IgG2a MR4 MR3 UCHT1 MAX16H5 OKT8 MEM 56 UCHL1 Dako P260 Dako P217
Isotype control Isotype control CD15 CD14 CD3 CD4 CD8 CD45RA CD45RO Secondary antibody Secondary antibody
(IgG1 isotype control) (IgG2a isotype control) PMN Monocytes/macrophages T cells T-helper cells, monocytes/macrophages Cytotoxic T cells Naive T cells, B cells, monocytes Memory T cells, B cells, monocytes/macrophages Rabbit anti-mouse Swine anti-rabbit
1:5 1:5 1:100 1:100 1:10 1:10 1:10 1:5 1:10 1:300 1:100
Dako, Hamburg, Germany Dako Knapp22 Knapp22 Knapp22 Knapp22 Knapp22 Knapp22 Knapp22 Dako Dako
body staining), which reduced the total number of evaluated samples to 1,424.
score (CSS) at all different biopsy sites (exemplified for site I in Fig 2).
Knee Function Scores
ASS
Knee function was assessed the day before surgery and after an average of 28 months of follow-up (range, 12 to 55 months) using the following 3 commonly applied knee joint scores. (1) The Lysholm score modified by Klein and Jensen17 with a maximum of 100 points for an optimal result. Its variables (pain, 35 points; swelling, walking distance, range of motion, 15 points each; and squatting, stair climbing, use of walking aids, and limping, 5 points each) provide a suitable tool for evaluating inflammatory joint diseases. (2) The Insall score,23 consisting of a 100-point functional score and a 100-point knee score. (3) The Lequesne score,24 in which 0 points represents the best result and 24 points the worst.
Before synovectomy, the ASS showed no significant differences among the different biopsy sites (Fig 3A; statistical power 1- between 0.64 and 0.99 depending on the site, assuming an expected reduction of the ASS, i.e., delta of 60%). After surgery, the mean ASS over all sites was reduced from 2.8 to 0.5 (Fig 3A; P ⱕ .05 compared with preoperative values for all sites). In terms of mean percent changes, the average ASS decrease over all sites was 82.1% (Fig 3B). Notably, however, the reduction was below average at 3 biopsy sites (i.e., femoral insertion of the anterior cruciate ligament [ACL; site VI], medial joint capsule [site VII], and lateral joint capsule [site IX]) (Fig 3B). After ar-
Statistical Analysis Data were expressed as means ⫾ standard error of the mean (SEM). Data from paired samples (preoperative and postoperative) of the different biopsy sites were subjected to the multigroup Friedman test and, subsequently, to the nonparametric Wilcoxon test using the SPSS 11.0 program (SPSS, Chicago, IL). Spearman rank correlation was used to assess correlations among parameters; statistically significant differences were accepted for P ⱕ .05. RESULTS Arthroscopic synovectomy generally led to a significant (P between .005 and .05) reduction of both the acute synovitis score (ASS) and the chronic synovitis
FIGURE 2. Hematoxylin and eosin–stained preoperative (A) and postoperative (B) samples from biopsy site I. In the preoperative synovial sample, high scores for acute synovitis (presence of PMN, fibrin deposits; arrowheads in A) and chronic synovitis (lining cell hyperplasia ⫽ ll, diffuse infiltrates ⫽ di, lymphoid aggregates, germinal centers, plasma cells ⫽ pc, formation of vessels ⫽ ve in A) are present. In contrast, the postoperative sample shows no acute synovitis score and only a remnant chronic synovitis score (mononuclear infiltration) (original magnification ⫻92).
SYNOVECTOMY IN RHEUMATOID ARTHRITIS
1213
than that observed for the ASS (average of 82.1%; Figs 3B and 4B). For the CSS, the upper lateral recess (site I), the upper central recess (site III), the femur insertion of the ACL (site VI), and the lateral (site VII) and medial joint capsule (site IX) showed a reduction below average (Fig 4B). After arthroscopic synovectomy, no significant differences were observed among the different biopsy sites for the CSS by Wilcoxon test. Immunohistologic Findings Immunohistochemical analysis of ST obtained preoperatively revealed that macrophages, CD45RO-pos-
FIGURE 3. Quantification of the acute synovitis score (A) in preoperative (black columns) and (B) postoperative synovial biopsies (white columns; sites I-IX) and reduction by arthroscopic synovectomy. Data are expressed as means ⫾ standard error of the mean (means in B) of samples from n ⫽ 11 patients for all biopsy sites (except for site III with n ⫽ 7 and site VI with n ⫽ 6 samples). The solid line in A represents the mean acute synovitis score in the preoperative samples over all biopsy sites, the dashed line the mean postoperative score. The dashed line in B indicates the mean reduction by synovectomy over all biopsy sites (82.1%). *P ⱕ .05, **P ⱕ .01, and ***P ⱕ .005 in comparison with the preoperative samples (Wilcoxon test).
throscopic synovectomy, no significant differences were observed among the different biopsy sites for the ASS by Wilcoxon test. CSS Before surgery, the CSS was significantly (P ⱕ .05) lower at sites II and III compared with other sites, indicating regional heterogeneity of the chronic inflammation in RA ST. The CSS was reduced from a mean preoperative value of 8.1 to a mean postoperative value of 3.0 (Fig 4A; P ⱕ .05 compared with preoperative values for all sites, except for site III). In terms of percent changes for the different sites, the reduction of the CSS after arthroscopic synovectomy (average of 62.5%) was significantly (P ⱕ .009) lower
FIGURE 4. Quantification of the chronic synovitis score in (A) preoperative (filled columns) and (B) postoperative synovial biopsies (empty columns; sites I-IX) and reduction by arthroscopic synovectomy. Data are expressed as means ⫾ standard error of the mean (means in B) of samples from n ⫽ 11 patients for all biopsy sites (except for site III with n ⫽ 9 and site VI with n ⫽ 7 samples). The solid line in A represents the mean chronic synovitis score in the preoperative samples over all biopsy sites, the dashed line the mean postoperative score. The dashed line in B indicates the mean reduction by synovectomy over all biopsy sites (62.5%). *P ⱕ .05 and ***P ⱕ .005 in comparison with the preoperative samples (Wilcoxon test); ⫹P ⱕ .05 in comparison with site IV; §P ⱕ .05 in comparison with site VI; #P ⱕ .05 in comparison with site VIII; $P ⱕ .05 in comparison with site IX.
1214
H.-D. CARL ET AL.
FIGURE 5. (A) Density of PMN (CD15), macrophages (Mf; CD14), T cells (CD3), as well as CD4, CD8, CD45RA, and CD45RO-positive cells over all preoperative and postoperative sites of the RA synovial membrane. (B-H) Quantification of the depletion of the different cell types by synovectomy. Data are expressed as means ⫾ standard error of the mean (means in B-H) of samples from n ⫽ 11 patients for all biopsy sites (except for site III with n ⫽ 5-6 and site VI with n ⫽ 6-7 samples). The dashed lines in B-H indicate the mean reduction by synovectomy over all biopsy sites for the individual cell types. §P ⱕ .05 compared with PMN; &P ⱕ .05 compared with T cells; ⫹P ⱕ .05 compared with CD4-positive cells; %P ⱕ .05 compared with CD8-positive cells; *P ⱕ .05 compared with CD45RA-positive cells.
itive cells, and T cells dominated the inflammatory infiltrates, whereas the density of CD4-positive, CD8positive, and CD45RA-positive cells was lower (Fig 5A). The lowest density was observed for PMN. Except for PMN, the relative proportions of the cell types
remained roughly constant in postoperative samples (Fig 5A). Arthroscopic synovectomy induced a significant (P ⱕ .05) subtype-specific reduction of inflammatory cells. Whereas the density of PMN was reduced by
SYNOVECTOMY IN RHEUMATOID ARTHRITIS
1215
The reduction of the total synovitis score for individual joints (the sum of ASS and CSS for all biopsy sites) at the 12-month follow-up showed a positive correlation (P ⱕ .01; n ⫽ 11) with the improvement of the Insall knee score by arthroscopic synovectomy (increase from 45.1 to 70.0 points; P ⱕ .021; n ⫽ 11), suggesting a link between surgical radicalism and clinical improvement. However, postoperative care such as limited weight bearing for 4 weeks and consecutive physical therapy may have contributed to the clinical improvement (e.g., joint pain, range of motion), which is a question to be addressed in the future. DISCUSSION FIGURE 6. Mean reduction of inflammatory infiltration by arthroscopic synovectomy over all 7 cell types (PMN, macrophages, T cells, CD4, CD8, CD45RA, and CD45RO-positive cells) at the different biopsy sites (I-IX). Data are expressed as means ⫾ standard error of the mean. The dashed line represents the mean reduction of inflammatory cells over all biopsy sites (53.4%). !P ⱕ .05 compared with site II (Wilcoxon test); ⫹P ⱕ .05 compared with site IV; &P ⱕ .05 compared with site V; #P ⱕ .05 compared with site VIII; $P ⱕ .05 compared with site IX.
81.8% and the density of CD8-positive and CD45RApositive cells by 61.6% and 55.7%, respectively, the reduction of the remaining cell types (macrophages, T cells, CD4-positive cells, and CD45RO-positive cells) was limited to less than 46% (Figs 5B-5H). Analyses of the mean inflammatory cell reduction (over all cell types) resulted in significant (P ⱕ .05) differences among individual biopsy sites. Notably, biopsies from the femoral insertion of the ACL (site VI, 40%) and the lateral joint capsule (site VII, 45%) showed a reduction significantly lower compared with several other sites (Fig 6). In contrast, the sites I (upper lateral recess), III (upper central recess), and IX (medial joint capsule) only differed significantly from one of the sites with the highest reductions (site II, upper lateral recess, 66%; site IV, upper medial recess, 70%; Fig 6). Knee Joint Scores After an average follow-up of 28 months, all 3 knee scores displayed a highly significant (P ⱕ .01) improvement. The Lysholm score (modified by Klein and Jensen) showed an improvement from a preoperative score of 41.9 to 68.8 points at follow-up. The Insall score improved from 45.1 to 65.1 points for the knee score and from 54.1 to 71.4 points for the functional score. The Lequesne score improved from 13.2 to 8.8 points (Fig 7).
Reduction of Inflammatory Infiltrates in RA ST Arthroscopic synovectomy led to a significant reduction of both ASS and CSS in all biopsy sites of the inflamed RA joint. These histologic effects were confirmed by immunohistologic analyses, showing a considerable reduction of inflammatory cells in RA ST (PMN, macrophages, and T-cell subsets). These cell types are known to play an important role in the development and perpetuation of RA18,21,25-29 and are the selective or nonselective targets of successful antirheumatic drug treatment in RA.30-33 Previous reports on the therapeutic efficacy of synovectomy (reduction of joint pain/swelling, improvement of functional parameters, and preservation of the structural integrity of the joint) point to an important contribution of the reduction of inflammatory infiltration to this process.7,10 This is further supported by the analysis of synovial biopsies before open synovectomy and 6 to 12 months thereafter, showing a clear relationship between arthroscopic, clinical, histopathologic, and immunohistopathologic findings.34-36 Differential Reduction of ASS and CSS by Arthroscopic Synovectomy Arthroscopic synovectomy reduced the ASS on average by approximately 82%, whereas the reduction of the CSS amounted to an average of 63%. In agreement with these findings, the density of PMN was reduced by approximately 82%, whereas that of macrophages and different T-cell subsets was only diminished by ⱕ62%. These results indicate that arthroscopic synovectomy differentially removes different subtypes of infiltrating leukocytes in ST. A possible explanation for these differences may be
1216
H.-D. CARL ET AL.
FIGURE 7. Mean values ⫾ SEM for the modified Lysholm score, the Insall knee and functional score, and the Lequesne score are shown, comparing preoperative point values and 28-month follow-up point values. White columns, preoperative values; black columns, postoperative values. P ⬍ .01 compared with preoperative values.
that PMN are often found in superficial regions of RA ST, either directly below or even within the lining layer.21 Macrophages, on the other hand, are present as type A cells in the lining layer of RA ST, but also as stromal macrophages in deeper regions of the RA synovial membrane.37 Finally, T cells are rarely located in the lining layer but rather in diffuse infiltrates or lymphoid aggregates in the depth of ST.21,38 These deeper cell types may escape the “shaving” effect of arthroscopic synovectomy, due to technical aspects of the procedure, and sustain the persistence of a basic CSS. Thus, macrophages or lymphocytes, i.e., the cell types characteristic of chronic inflammation21,25-26,29,39 and presumably responsible for the formation of a neosynovial membrane and the occurrence of resynovitis,34-36 remain in RA ST after arthroscopic synovectomy, albeit at a markedly reduced density. Differential Reduction of Inflammatory Infiltrates in Different Joint Regions Arthroscopic synovectomy resulted in regional differences concerning the radicalism of ST removal, thus supporting our working hypothesis. Indeed, the
upper lateral recess (site I), the upper central recess (III), the femur insertion of the ACL (VI), as well as the lateral (VII) and medial joint capsule (IX) showed a less marked reduction of inflammatory infiltrates (in the case of ACL [VI], lateral [VII], and medial joint capsule [IX] concerning both ASS and CSS). At these locations, the full extent of synovitis may be difficult to assess during arthroscopy. In addition, surgical sparing may result from the attempt to avoid damage to ligamentous and capsular tissue. A more thorough removal of ST in these anatomic regions (possibly supported by an advanced experience of the surgeon) may help to further improve the therapeutic efficacy of arthroscopic synovectomy.6,13 Improvement of Knee Function Scores Arthroscopic synovectomy significantly improved various knee function scores, both after a 1-year (only shown for the Insall knee score) and after a 28-month follow-up. Although postoperative care, consecutive physical therapy, and in 1 case, subsequent radiation synovectomy must be taken into consideration, the present results support previous reports on the therapeutic efficacy of arthroscopic synovectomy.7,10 In
SYNOVECTOMY IN RHEUMATOID ARTHRITIS addition, there was a positive correlation between the reduction of total synovitis for individual joints at the 1-year follow-up with the improvement of the Insall knee score by arthroscopic synovectomy. These findings suggest a link between surgical radicalism and clinical improvement and further emphasize the need for radical and thorough removal of inflamed ST.6,13 It has recently been reported that the clinical improvement in RA patients with therapy-resistant synovitis following combined arthroscopic and radiation synovectomy is superior to that following either 1 of the 2 individual methods.7,13 In the former multicenter study (33-month follow-up),7 the combined approach induced a higher benefit with respect to erythrocyte sedimentation rate, Insall knee score, modified Lysholm score, Lequesne score, joint pain as measured on a 4-point scale, range of motion, and presence of effusion and synovitis (P ⱕ .05 for all parameters). Although there are as yet no studies comparing the degree of ST removal after arthroscopic synovectomy, radiation synovectomy, or combination therapy, it appears likely that more radical reduction of inflammatory infiltrates by combination therapy contributes to its augmented therapeutic efficacy. However, it is possible also that the efficacy of monotherapy by arthroscopy could be improved by more careful, radical, and thorough excision of inflamed ST, applying particular care to the critical regions identified in the present report. Future research may have to include similar studies with larger patient numbers and also biopsy and clinical outcome evaluations after a longtime follow-up, to determine the therapeutic importance of the reduction of acute and chronic inflammatory activity for arthroscopic synovectomy. CONCLUSION Although arthroscopic synovectomy effectively reduces acute inflammation of ST in RA, chronic inflammation is diminished to a more limited degree. Also, the elimination of ST is less radical in areas relatively spared from surgery in order to avoid ligament damage, the elimination of ST is less radical. Therefore, in patients with RA who have persistent knee synovitis, the continued presence of chronically inflamed ST following arthroscopic synovectomy suggests either more thorough removal of inflamed ST or the application of additional measures, e.g., the combination of arthroscopic and radiation synovectomy. Acknowledgment: Dr. E. Palombo-Kinne is gratefully acknowledged for critical revision of the manuscript, U.
1217
Vorderwülbecke, B. Müller, J. Harzendorf, and B. Niescher for excellent technical assistance.
REFERENCES 1. Klug S, Willauschus W, Weseloh G. Arthroscopic synovectomy of the knee joint. Arthroscopy 1997;16:262-267. 2. Kresnik E, Mikosch P, Gallowitsch HJ, et al. Clinical outcome of radiosynoviorthesis: A meta-analysis including 2190 treated joints. Nucl Med Commun 2002;23:683-688. 3. Gibbons CE, Gosal HS, Bartlett J. Long-term results of arthroscopic synovectomy for seropositive rheumatoid arthritis: 6-16 year review. Int Orthop 2002;26:98-100. 4. Roch-Bras F, Daures JP, Legouffe MC, Sany J, Combe B. Treatment of chronic knee synovitis with arthroscopic synovectomy: Long-term results. J Rheumatol 2002;29:1171-1175. 5. Horiuchi K, Momohara S, Tomatsu T, Inoue K, Toyama J. Arthroscopic synovectomy of the elbow in rheumatoid arthritis. J Bone Joint Surg Am 2002;84:342-347. 6. Hagena FW. Development of rheumatoid arthritis surgery in German-speaking countries. Orthopaede 2001;30:756-767. 7. Klug S, Wittmann G, Weseloh G. Arthroscopic synovectomy of the knee joint in early cases of rheumatoid arthritis: Follow-up results of a multicenter study. Arthroscopy 2000;16: 262-267. 8. Toda Y. The relationship between prognosis after one-day arthroscopic partial synovectomy on rheumatoid knee joints and the histologic features of the synovial membrane. Ryumachi 2003;43:549-556. 9. Tanaka N, Sakahashi H, Sato E, Ishi S. Immunohistological indication for arthroscopic synovectomy in rheumatoid knees: Analysis of synovial samples obtained by needle arthroscopy. Clin Rheumatol 2002;21:46-51. 10. Jüsten HP, Pommer S, Leeb I, Pilhofer C, Kisslinger E, Wessinghage D. Long-term results of open knee synovectomy in later cases of rheumatoid arthritis. Z Rheumatol 1999;58:20012006. 11. Bynum CK, Tasto J. Arthroscopic treatment of synovial disorders in the shoulder, elbow, and ankle. J Knee Surg 2002; 15:57-59. 12. Muro H, Waguri-Nagaya Y, Otsuka T, Matsui N, Asai K, Kato T. Serum gliostatin levels in patients with rheumatoid factornegative and -positive rheumatoid arthritis and changes of these levels after surgical treatment. Clin Rheumatol 2001;20: 331-336. 13. Kerschbaumer F, Kandziora F, Herresthal J, Hertel A, Hor G. Combined arthroscopic and radiation synovectomy in rheumatoid arthritis. Orthopaede 1998;27:188-196. 14. Arnett FC, Edworthy SM, Bloch DA, et al. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 1988;31:315324. 15. Larsen A, Dale K, Eek M. Radiographic evaluation of rheumatoid arthritis and related conditions by standard reference films. Acta Radiol 1977;18:481-491. 16. Highgenboten CL. Arthroscopic synovectomy. Arthroscopy 1985;1:190-193. 17. Klein W, Jensen KU. Arthroscopic synovectomy of the knee joint: Indication, technique and follow-up results. Arthroscopy 1988;4:63-71. 18. Geiler G, Stiehl P. Significance of the concepts basic activity and actual activity in the morphological evaluation of synovial membranes in rheumatoid arthritis. Z Rheumatol 1974;33:7386. 19. Duke O, Panayi GS, Janossy G, Poulter LW. Immunohistological analysis of the lymphocytic infiltrates of rheumatoid
1218
20. 21.
22. 23. 24.
25.
26. 27. 28.
29. 30.
H.-D. CARL ET AL.
synovial membrane using monoclonal antibodies. Ann Rheum Dis 1982;41:192-193. Young CL, Adamson TC, Vaughan JH, Fox RI. Immunohistologic characterization of synovial membrane lymphocytes in rheumatoid arthritis. Arthritis Rheum 1984;27:124-130. Sack U, Kinne RW, Marx T, Heppt P, Bender S, Emmrich F. Interleukin-6 in synovial fluid is closely associated with chronic synovitis in rheumatoid arthritis. Rheumatol Int 1993; 13:45-51. Knapp W. Leukocyte typing IV. Oxford: Oxford University, 1990. Insall JN, Dorr DD, Scott RD, Scott WN. Rationale of the knee society clinical rating system. Clin Orthop 1989;248:13-14. Lequesne MG, Mery C, Samson M, Gerard P. Index of severity of osteoarthritis for the hip and knee. Validation-value in comparison with other assessment tests. Scand J Rheumatol 1987;65:85-89. Smeets TJ, Kraan MC, Galjaard S, Youssef PP, Smith MD, Tak PP. Analysis of the cell infiltrate and expression of matrix metalloproteinases and granzyme B in paired synovial biopsy specimens from the cartilage-pannus junction in patients with RA. Ann Rheum Dis 2001;60:561-565. Mulherin DM, Fitzgerald O, Breshnihan B. Synovial tissue macrophage populations and articular damage in rheumatoid arthritis. Arthritis Rheum 1996;39:115-124. Firestein GS. Novel therapeutic strategies involving animals, arthritis, and apoptosis. Curr Opin Rheumatol 1998;10:236241. Tak PP, Taylor PC, Breedveld FC, et al. Decrease in cellularity and expression of adhesion molecules by anti-tumor necrosis factor alpha monoclonal antibody treatment in patients with rheumatoid arthritis. Arthritis Rheum 1996;39:1077-1081. Tak PP. Editorial. Lessons learnt from the synovial tissue response to anti-rheumatic treatment. Rheumatology 2000;39: 817-820. Smith MD, Slavotinek J, Au V, et al. Successful treatment of rheumatoid arthritis is associated with a reduction in synovial
31.
32.
33. 34.
35.
36.
37. 38. 39.
membrane cytokines and cell adhesion molecule expression. Rheumatology 2001;40:965-977. Kraan MC, Reece RJ, Barg EC, et al. Modulation of inflammation and metalloproteinase expression in synovial tissue by leflunomide and methotrexate in patients with active rheumatoid arthritis. Findings in a prospective, randomized, doubleblind, parallel-design clinical trial in thirty-nine patients at two centers. Arthritis Rheum 2000;43:1820-1830. Wong P, Cuello C, Bertouch JV, et al. The effects of pulse methylprednisolone on matrix metalloproteinase and tissue inhibitor of metalloproteinase-1 expression in rheumatoid. Rheumatology 2000;39:1067-1073. Barrera P, Blom A, van Lent PL, et al. Synovial macrophage depletion with clodronate-containing liposomes in rheumatoid arthritis. Arthritis Rheum 2000;43:1951-1959. Paus AC, Pahle JA. Arthroscopic evaluation of the synovial lining before and after open synovectomy of the knee joint in patients with chronic inflammatory joint disease. Scand J Rheumatol 1990;19:193-201. Paus AC, Refsum S, Forre O. Histopathologic changes in arthroscopic synovial biopsies before and after open synovectomy in patients with chronic inflammatory joint diseases. Scand J Rheumatol 1990;19:202-208. Paus AC, Mellbye OJ, Forre O. Immunohistopathologic findings in synovial biopsies before and after synovectomy in patients with chronic inflammatory joint diseases and their relation to clinical evaluation. A prospective study of biopsies taken from areas selected by arthroscopy. Scand J Rheumatol 1990;19:269-279. Kinne RW, Bräuer B, Stuhlmüller B, Palombo-Kinne E, Burmester GR. Macrophages in rheumatoid arthritis. Arthritis Res 2000;2:189-202. Smolen J, Steiner G. Therapeutic strategies for rheumatoid arthritis. Nature Reviews 2003;2:473-488. Kinne RW, Palombo-Kinne E, Emmrich F. T-cells in the pathogenesis of rheumatoid arthritis—Villains or accomplices. Biochimica Biophysica Acta 1997;1360:109-141.
Purpose: To assess the intraoperative reduction of inflammatory infiltrates achieved by arthroscopic knee joint synovectomy in patients with rheumatoid arthritis (RA) with special regard to the removal site, using preoperative and postoperative synovial tissue (ST) samples. Type of Study: A histologic and immunohistochemical study. Methods: Eleven patients with treatment-refractory RA knee synovitis underwent arthroscopic synovectomy. In each patient, ST specimens were obtained immediately before and after synovectomy from 9 defined sites covering the whole joint. The samples were graded using an acute synovitis score (ASS; presence of polymorphonuclear neutrophilic leukocytes [PMN] and fibrin) and a chronic synovitis score (CSS; e.g., lining cell hyperplasia, presence of diffuse and lymphoid aggregates). Immunohistologic analyses were performed using 7 monoclonal antibodies directed against PMN, macrophages, and T-cell subsets (total of 1,584 preparations). Knee function was assessed after an average follow-up of 28 months by Lysholm score (modified by Klein and Jensen), Insall functional and knee scores, and Lequesne score. Results: Arthroscopic synovectomy led to an overall significant (P between .005 and .05) reduction of the acute inflammatory infiltrates (ASS) by 82.1%, but to a significant reduction of chronic inflammatory infiltrates (CSS) by only 62.5%. Accordingly, the density of PMN was reduced by 81.8%, whereas that of macrophages and different T-cell subsets was only decreased by ⱕ61.6%. With respect to the anatomic regions, a significantly (P ⱕ .05) less marked reduction of inflammatory infiltrates was observed in the upper lateral and central recess, at the medial and lateral capsule, as well as at the femoral insertion of the anterior cruciate ligament. All knee joint scores showed a significant (P ⱕ .01) improvement over preoperative values at follow-up. Conclusions: Arthroscopic synovectomy effectively reduces acute and chronic inflammatory infiltrates in patients with RA who have refractory synovitis of the knee joint (immediately after synovectomy) and improves knee function (28-month follow-up). However, the reduction of inflammatory infiltrates appears to depend on the anatomic region of the joint. Level of Evidence: Level III. Key Words: Rheumatoid arthritis—Arthroscopy—Synovectomy—Knee inflammation.
I
From the Division of Orthopedic Rheumatology, Department of Orthopedic Surgery, Friedrich Alexander University of ErlangenNuremberg (H-D.C., S.K., J.S., B.S.), Erlangen; and the Experimental Rheumatology Unit, Friedrich Schiller University Jena (R.W.K.), Jena, Germany. Address correspondence and reprint requests to Hans-Dieter Carl, M.D., Division of Orthopedic Rheumatology, Department of Orthopedic Surgery, Friedrich Alexander University of ErlangenNuremberg, Rathsberger Strasse 57, D-91054 Erlangen, Germany. E-mail: [email protected] © 2005 by the Arthroscopy Association of North America 0749-8063/05/2110-4452$30.00/0 doi:10.1016/j.arthro.2005.06.014
n patients with rheumatoid arthritis (RA), sustained refractory synovitis for a time period of more than 6 months despite adequate antirheumatic therapy represents an indication for synovectomy.1 Because the highest rate of success, defined as clinical improvement or delay of radiologic progression, has been observed in patients with limited radiologic changes, synovectomy is recommended in early stages of the disease.2-6 In a retrospective, midterm, multicenter study, the combination of arthroscopic and radiation synovectomy in early cases of RA has proven superior to either single method,7 possibly the result of a more
Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 21, No 10 (October), 2005: pp 1209-1218
1209
1210
H.-D. CARL ET AL. TABLE 1.
Clinical Characteristics of the RA Patients at the Time of Arthroscopic Synovectomy
No. of Patients
Gender (M/F)
Age (yr/range)
Disease Duration (yrs/range)
RF (⫹/⫺)
No. of ARA Criteria (RA)
Concomitant Medication
11
2/9
51.9 ⫾ 3.4 (22-63)
9.3 ⫾ 2.0 (1-20)
9/2
all ⬎4
MTX (n ⫽ 9) AZT (n ⫽ 2) NSAIDs (n ⫽ 11)
NOTE. For the parameters age and disease duration means ⫾ SEM are given. Abbreviations: RF, rheumatoid factor; ⫹/⫺, positive/negative; ARA, American Rheumatism Association (now the American College of Rheumatology); MTX, methotrexate; AZT, azathioprine, NSAIDs, nonsteroidal anti-inflammatory drugs.
thorough removal of inflamed synovial tissue (ST). In fact, the degree of clinical improvement and the postoperative prognosis of synovectomy appears to depend on the histologic composition of the rheumatoid ST8,9 and the extent of ST removal.9-13 Based on these findings, the present study sought to analyze: (1) the inflammatory cell infiltrates present in refractory synovitis of rheumatic knee joints before arthroscopic synovectomy; and (2) the efficacy of arthroscopic synovectomy in removing synovial infiltrates (as assessed immediately after synovectomy), with particular regard to the anatomic region within the knee joint. The hypothesis underlying the present study was that the extent of ST removal by arthroscopic synovectomy displays site-specific differences within the knee joint.
Surgical Procedure Arthroscopic synovectomy was carried out following previously defined protocols.1,16-17 ST samples were obtained from 9 defined sites within the knee joint (Fig 1) before synovectomy was performed and again immediately after its completion using biopsy pincers. In all cases, multiportal synovectomy was performed through the 4-6 portals using motorized shaver systems. For the procedure, a tourniquet was used and the leg was placed in a hanging position. Postoperatively, limited weight bearing (approximately 15 kg) as taught by a physiotherapist was allowed on the operated leg for 4 weeks, and physiotherapy (both active and passive) along with continuous passive motion exercises were used. Diseasemodifying antirheumatic drug therapy was not interrupted or modified during surgery or recovery.
METHODS Patients
Histologic Evaluation
ST was obtained from 11 patients with RA during arthroscopic knee synovectomy. RA was classified according to the American College of Rheumatology criteria14 (Table 1). Of the 11 RA patients, 3 had increased erythrocyte sedimentation rates (49 ⫾ 27 mm/h; mean ⫾ SEM) and C-reactive protein values. Septic arthritis was excluded by synovial fluid analysis and Gram staining of synovial fluid and ST samples. Radiologic changes were not higher than Larsen stage II for any operated knee joint.15 All patients had refractory synovitis of the knee and all had received constant disease-modifying antirheumatic drug therapy; monotherapy with methotrexate (15 mg per week) in 9 patients and a combination therapy of methotrexate and azathioprine (200 or 300 mg daily) in 2 patients (Table 1) for a minimum of 6 months. None of the patients received steroids before, during, or after surgery; nonsteroidal anti-inflammatory drugs were administered on demand.
Following standard hematoxylin and eosin staining, the extent of synovitis was semiquantitatively assessed using a scoring system initially proposed by Geiler and Stiehl18 (Table 2), with 0 ⫽ absence of parameter, 1 ⫽ weak intensity, 2 ⫽ medium intensity, and 3 ⫽ strong intensity for the given parameter. The scores for the individual parameters of acute synovitis (presence of polymorphonuclear neutrophilic leukocytes [PMN] and fibrin deposits; maximum score of 6) and chronic synovitis (lining cell hyperplasia, diffuse infiltrates, lymphoid aggregates, germinal centers, plasma cells, formation of vessels; maximum score of 18) were then added. For technical reasons (mainly lack of clear histologic structures), some specimens were not suitable for reliable analysis, which reduced the total number of samples from 198 (11 patients; 9 preoperative and 9 postoperative samples) to 180 for the acute synovitis score and 186 for the chronic synovitis score.
SYNOVECTOMY IN RHEUMATOID ARTHRITIS
1211
FIGURE 1. Preoperative and postoperative ST samples ( ⫽ synovial membrane; ⫽ synovial membrane/bone punch) were obtained from 9 different locations covering the lateral, medial, superior, and anterior part of the knee joint (exemplified for the right knee joint). Individual locations were: I ⫽ upper lateral recess, highest point; II ⫽ upper lateral recess, cartilage/pannus junction; III ⫽ upper central recess, cartilage/pannus junction, bone punch; IV ⫽ upper medial recess, cartilage/pannus junction; V ⫽ upper medial recess, highest point; VI ⫽ femur insertion ACL, cartilage/pannus junction, inter-condylar bone punch; VII ⫽ lateral joint capsule, above meniscus; VIII ⫽ tibia insertion ACL, fibro-cartilage/pannus junction; IX ⫽ medial joint capsule, above meniscus.
at room temperature. All subsequent steps were performed at room temperature in a humid chamber. Specific marker antibodies directed against PMN, macrophages, and T-cell subsets22 (Table 3) were diluted in Tris-buffered saline and added to the sections for 30 minutes. Following 3 washes for 5 minutes each with phosphate-buffered saline (PBS; 0.9% NaCl, 30 mmol/L KCl, 70 mmol/L Na2PO4, and 10 mmol/L KH2PO4, pH 7.4), a peroxidase-coupled rabbit-anti-mouse antibody (1:300 in PBS) was applied for 30 minutes. The sections were washed and then incubated for 30 minutes with a peroxidase-coupled swine-anti-rabbit antibody (1:100 in PBS). The peroxidase was developed in a solution of 5 mg diaminobenzidine (Sigma, Deisenhofen, Germany) in 10 mL PBS with a pH of 7.4. Shortly before use, the solution was filtered and 30 L of 3% H2O2 were added. The same staining procedure was performed for controls, but the specific antibodies were either omitted or replaced by isotype controls in matched concentrations (Table 3, in both cases with negative results). The density of the different inflammatory cell types (PMN, macrophages, and T-cell subsets) in the ST samples was semiquantitatively assessed using a scoring system ranging from 0 to 3, with 0 ⫽ no positive cells, 1 ⫽ less than 40% positive cells, 2 ⫽ 40% to 90% positive cells, and 3 ⫽ more than 90% positive cells. Using 7 different monoclonal antibodies and a negative control (8 sections), as well as 9 preoperative and 9 postoperative samples from 11 patients with RA (198 samples), a total of 1,584 preparations was assessed (144 specimens per knee joint). Although biopsy samples were obtained from all areas in all joints, some specimens were not suitable for reliable analysis for technical reasons (mainly failure of anti-
Immunohistochemistry
TABLE 2.
For immunohistochemical analysis, ST was embedded in Tissue-Tek O.G.T. compound (Diatec, Nuremberg, Germany), immediately frozen in isopentane (Merck, Darmstadt, Germany) cooled in liquid nitrogen, and stored at ⫺70°C. Serial cryostat sections (4-m-thick) were cut in analogy to previously published procedures19-21 using a low open-top microtome (No. 5030; Bright, Huntingdon, England) and both hematoxylin and eosin and immunohistologic staining was performed. Briefly, sections were air-dried and then fixed with acetone and chloroform (Merck) for 10 minutes each
Criteria for the Quantification of Inflammatory Activity
ASS (Actual Activity)
CSS (Basic Activity)
Presence of PMN (0-3 points) Lining cell hyperplasia (0-3 points) Fibrin deposits (0-3 points) Diffuse infiltrates (0-3 points) Maximum score 6 points Lymphoid aggregates (0-3 points) Germinal centers (0-3 points) Plasma cells (0-3 points) Formation of vessels (0-3 points) Maximum score (18 points) NOTE. The criteria for the quantification of the acute and chronic inflammatory activity in RA ST are listed as previously defined.18
1212
H.-D. CARL ET AL. TABLE 3.
Antibodies Used for Immunohistochemical Study
Antibodies (clone)
Antigen Recognized
Main Cellular Expression (as previously published)
Dilution
Source
IgG1 IgG2a MR4 MR3 UCHT1 MAX16H5 OKT8 MEM 56 UCHL1 Dako P260 Dako P217
Isotype control Isotype control CD15 CD14 CD3 CD4 CD8 CD45RA CD45RO Secondary antibody Secondary antibody
(IgG1 isotype control) (IgG2a isotype control) PMN Monocytes/macrophages T cells T-helper cells, monocytes/macrophages Cytotoxic T cells Naive T cells, B cells, monocytes Memory T cells, B cells, monocytes/macrophages Rabbit anti-mouse Swine anti-rabbit
1:5 1:5 1:100 1:100 1:10 1:10 1:10 1:5 1:10 1:300 1:100
Dako, Hamburg, Germany Dako Knapp22 Knapp22 Knapp22 Knapp22 Knapp22 Knapp22 Knapp22 Dako Dako
body staining), which reduced the total number of evaluated samples to 1,424.
score (CSS) at all different biopsy sites (exemplified for site I in Fig 2).
Knee Function Scores
ASS
Knee function was assessed the day before surgery and after an average of 28 months of follow-up (range, 12 to 55 months) using the following 3 commonly applied knee joint scores. (1) The Lysholm score modified by Klein and Jensen17 with a maximum of 100 points for an optimal result. Its variables (pain, 35 points; swelling, walking distance, range of motion, 15 points each; and squatting, stair climbing, use of walking aids, and limping, 5 points each) provide a suitable tool for evaluating inflammatory joint diseases. (2) The Insall score,23 consisting of a 100-point functional score and a 100-point knee score. (3) The Lequesne score,24 in which 0 points represents the best result and 24 points the worst.
Before synovectomy, the ASS showed no significant differences among the different biopsy sites (Fig 3A; statistical power 1- between 0.64 and 0.99 depending on the site, assuming an expected reduction of the ASS, i.e., delta of 60%). After surgery, the mean ASS over all sites was reduced from 2.8 to 0.5 (Fig 3A; P ⱕ .05 compared with preoperative values for all sites). In terms of mean percent changes, the average ASS decrease over all sites was 82.1% (Fig 3B). Notably, however, the reduction was below average at 3 biopsy sites (i.e., femoral insertion of the anterior cruciate ligament [ACL; site VI], medial joint capsule [site VII], and lateral joint capsule [site IX]) (Fig 3B). After ar-
Statistical Analysis Data were expressed as means ⫾ standard error of the mean (SEM). Data from paired samples (preoperative and postoperative) of the different biopsy sites were subjected to the multigroup Friedman test and, subsequently, to the nonparametric Wilcoxon test using the SPSS 11.0 program (SPSS, Chicago, IL). Spearman rank correlation was used to assess correlations among parameters; statistically significant differences were accepted for P ⱕ .05. RESULTS Arthroscopic synovectomy generally led to a significant (P between .005 and .05) reduction of both the acute synovitis score (ASS) and the chronic synovitis
FIGURE 2. Hematoxylin and eosin–stained preoperative (A) and postoperative (B) samples from biopsy site I. In the preoperative synovial sample, high scores for acute synovitis (presence of PMN, fibrin deposits; arrowheads in A) and chronic synovitis (lining cell hyperplasia ⫽ ll, diffuse infiltrates ⫽ di, lymphoid aggregates, germinal centers, plasma cells ⫽ pc, formation of vessels ⫽ ve in A) are present. In contrast, the postoperative sample shows no acute synovitis score and only a remnant chronic synovitis score (mononuclear infiltration) (original magnification ⫻92).
SYNOVECTOMY IN RHEUMATOID ARTHRITIS
1213
than that observed for the ASS (average of 82.1%; Figs 3B and 4B). For the CSS, the upper lateral recess (site I), the upper central recess (site III), the femur insertion of the ACL (site VI), and the lateral (site VII) and medial joint capsule (site IX) showed a reduction below average (Fig 4B). After arthroscopic synovectomy, no significant differences were observed among the different biopsy sites for the CSS by Wilcoxon test. Immunohistologic Findings Immunohistochemical analysis of ST obtained preoperatively revealed that macrophages, CD45RO-pos-
FIGURE 3. Quantification of the acute synovitis score (A) in preoperative (black columns) and (B) postoperative synovial biopsies (white columns; sites I-IX) and reduction by arthroscopic synovectomy. Data are expressed as means ⫾ standard error of the mean (means in B) of samples from n ⫽ 11 patients for all biopsy sites (except for site III with n ⫽ 7 and site VI with n ⫽ 6 samples). The solid line in A represents the mean acute synovitis score in the preoperative samples over all biopsy sites, the dashed line the mean postoperative score. The dashed line in B indicates the mean reduction by synovectomy over all biopsy sites (82.1%). *P ⱕ .05, **P ⱕ .01, and ***P ⱕ .005 in comparison with the preoperative samples (Wilcoxon test).
throscopic synovectomy, no significant differences were observed among the different biopsy sites for the ASS by Wilcoxon test. CSS Before surgery, the CSS was significantly (P ⱕ .05) lower at sites II and III compared with other sites, indicating regional heterogeneity of the chronic inflammation in RA ST. The CSS was reduced from a mean preoperative value of 8.1 to a mean postoperative value of 3.0 (Fig 4A; P ⱕ .05 compared with preoperative values for all sites, except for site III). In terms of percent changes for the different sites, the reduction of the CSS after arthroscopic synovectomy (average of 62.5%) was significantly (P ⱕ .009) lower
FIGURE 4. Quantification of the chronic synovitis score in (A) preoperative (filled columns) and (B) postoperative synovial biopsies (empty columns; sites I-IX) and reduction by arthroscopic synovectomy. Data are expressed as means ⫾ standard error of the mean (means in B) of samples from n ⫽ 11 patients for all biopsy sites (except for site III with n ⫽ 9 and site VI with n ⫽ 7 samples). The solid line in A represents the mean chronic synovitis score in the preoperative samples over all biopsy sites, the dashed line the mean postoperative score. The dashed line in B indicates the mean reduction by synovectomy over all biopsy sites (62.5%). *P ⱕ .05 and ***P ⱕ .005 in comparison with the preoperative samples (Wilcoxon test); ⫹P ⱕ .05 in comparison with site IV; §P ⱕ .05 in comparison with site VI; #P ⱕ .05 in comparison with site VIII; $P ⱕ .05 in comparison with site IX.
1214
H.-D. CARL ET AL.
FIGURE 5. (A) Density of PMN (CD15), macrophages (Mf; CD14), T cells (CD3), as well as CD4, CD8, CD45RA, and CD45RO-positive cells over all preoperative and postoperative sites of the RA synovial membrane. (B-H) Quantification of the depletion of the different cell types by synovectomy. Data are expressed as means ⫾ standard error of the mean (means in B-H) of samples from n ⫽ 11 patients for all biopsy sites (except for site III with n ⫽ 5-6 and site VI with n ⫽ 6-7 samples). The dashed lines in B-H indicate the mean reduction by synovectomy over all biopsy sites for the individual cell types. §P ⱕ .05 compared with PMN; &P ⱕ .05 compared with T cells; ⫹P ⱕ .05 compared with CD4-positive cells; %P ⱕ .05 compared with CD8-positive cells; *P ⱕ .05 compared with CD45RA-positive cells.
itive cells, and T cells dominated the inflammatory infiltrates, whereas the density of CD4-positive, CD8positive, and CD45RA-positive cells was lower (Fig 5A). The lowest density was observed for PMN. Except for PMN, the relative proportions of the cell types
remained roughly constant in postoperative samples (Fig 5A). Arthroscopic synovectomy induced a significant (P ⱕ .05) subtype-specific reduction of inflammatory cells. Whereas the density of PMN was reduced by
SYNOVECTOMY IN RHEUMATOID ARTHRITIS
1215
The reduction of the total synovitis score for individual joints (the sum of ASS and CSS for all biopsy sites) at the 12-month follow-up showed a positive correlation (P ⱕ .01; n ⫽ 11) with the improvement of the Insall knee score by arthroscopic synovectomy (increase from 45.1 to 70.0 points; P ⱕ .021; n ⫽ 11), suggesting a link between surgical radicalism and clinical improvement. However, postoperative care such as limited weight bearing for 4 weeks and consecutive physical therapy may have contributed to the clinical improvement (e.g., joint pain, range of motion), which is a question to be addressed in the future. DISCUSSION FIGURE 6. Mean reduction of inflammatory infiltration by arthroscopic synovectomy over all 7 cell types (PMN, macrophages, T cells, CD4, CD8, CD45RA, and CD45RO-positive cells) at the different biopsy sites (I-IX). Data are expressed as means ⫾ standard error of the mean. The dashed line represents the mean reduction of inflammatory cells over all biopsy sites (53.4%). !P ⱕ .05 compared with site II (Wilcoxon test); ⫹P ⱕ .05 compared with site IV; &P ⱕ .05 compared with site V; #P ⱕ .05 compared with site VIII; $P ⱕ .05 compared with site IX.
81.8% and the density of CD8-positive and CD45RApositive cells by 61.6% and 55.7%, respectively, the reduction of the remaining cell types (macrophages, T cells, CD4-positive cells, and CD45RO-positive cells) was limited to less than 46% (Figs 5B-5H). Analyses of the mean inflammatory cell reduction (over all cell types) resulted in significant (P ⱕ .05) differences among individual biopsy sites. Notably, biopsies from the femoral insertion of the ACL (site VI, 40%) and the lateral joint capsule (site VII, 45%) showed a reduction significantly lower compared with several other sites (Fig 6). In contrast, the sites I (upper lateral recess), III (upper central recess), and IX (medial joint capsule) only differed significantly from one of the sites with the highest reductions (site II, upper lateral recess, 66%; site IV, upper medial recess, 70%; Fig 6). Knee Joint Scores After an average follow-up of 28 months, all 3 knee scores displayed a highly significant (P ⱕ .01) improvement. The Lysholm score (modified by Klein and Jensen) showed an improvement from a preoperative score of 41.9 to 68.8 points at follow-up. The Insall score improved from 45.1 to 65.1 points for the knee score and from 54.1 to 71.4 points for the functional score. The Lequesne score improved from 13.2 to 8.8 points (Fig 7).
Reduction of Inflammatory Infiltrates in RA ST Arthroscopic synovectomy led to a significant reduction of both ASS and CSS in all biopsy sites of the inflamed RA joint. These histologic effects were confirmed by immunohistologic analyses, showing a considerable reduction of inflammatory cells in RA ST (PMN, macrophages, and T-cell subsets). These cell types are known to play an important role in the development and perpetuation of RA18,21,25-29 and are the selective or nonselective targets of successful antirheumatic drug treatment in RA.30-33 Previous reports on the therapeutic efficacy of synovectomy (reduction of joint pain/swelling, improvement of functional parameters, and preservation of the structural integrity of the joint) point to an important contribution of the reduction of inflammatory infiltration to this process.7,10 This is further supported by the analysis of synovial biopsies before open synovectomy and 6 to 12 months thereafter, showing a clear relationship between arthroscopic, clinical, histopathologic, and immunohistopathologic findings.34-36 Differential Reduction of ASS and CSS by Arthroscopic Synovectomy Arthroscopic synovectomy reduced the ASS on average by approximately 82%, whereas the reduction of the CSS amounted to an average of 63%. In agreement with these findings, the density of PMN was reduced by approximately 82%, whereas that of macrophages and different T-cell subsets was only diminished by ⱕ62%. These results indicate that arthroscopic synovectomy differentially removes different subtypes of infiltrating leukocytes in ST. A possible explanation for these differences may be
1216
H.-D. CARL ET AL.
FIGURE 7. Mean values ⫾ SEM for the modified Lysholm score, the Insall knee and functional score, and the Lequesne score are shown, comparing preoperative point values and 28-month follow-up point values. White columns, preoperative values; black columns, postoperative values. P ⬍ .01 compared with preoperative values.
that PMN are often found in superficial regions of RA ST, either directly below or even within the lining layer.21 Macrophages, on the other hand, are present as type A cells in the lining layer of RA ST, but also as stromal macrophages in deeper regions of the RA synovial membrane.37 Finally, T cells are rarely located in the lining layer but rather in diffuse infiltrates or lymphoid aggregates in the depth of ST.21,38 These deeper cell types may escape the “shaving” effect of arthroscopic synovectomy, due to technical aspects of the procedure, and sustain the persistence of a basic CSS. Thus, macrophages or lymphocytes, i.e., the cell types characteristic of chronic inflammation21,25-26,29,39 and presumably responsible for the formation of a neosynovial membrane and the occurrence of resynovitis,34-36 remain in RA ST after arthroscopic synovectomy, albeit at a markedly reduced density. Differential Reduction of Inflammatory Infiltrates in Different Joint Regions Arthroscopic synovectomy resulted in regional differences concerning the radicalism of ST removal, thus supporting our working hypothesis. Indeed, the
upper lateral recess (site I), the upper central recess (III), the femur insertion of the ACL (VI), as well as the lateral (VII) and medial joint capsule (IX) showed a less marked reduction of inflammatory infiltrates (in the case of ACL [VI], lateral [VII], and medial joint capsule [IX] concerning both ASS and CSS). At these locations, the full extent of synovitis may be difficult to assess during arthroscopy. In addition, surgical sparing may result from the attempt to avoid damage to ligamentous and capsular tissue. A more thorough removal of ST in these anatomic regions (possibly supported by an advanced experience of the surgeon) may help to further improve the therapeutic efficacy of arthroscopic synovectomy.6,13 Improvement of Knee Function Scores Arthroscopic synovectomy significantly improved various knee function scores, both after a 1-year (only shown for the Insall knee score) and after a 28-month follow-up. Although postoperative care, consecutive physical therapy, and in 1 case, subsequent radiation synovectomy must be taken into consideration, the present results support previous reports on the therapeutic efficacy of arthroscopic synovectomy.7,10 In
SYNOVECTOMY IN RHEUMATOID ARTHRITIS addition, there was a positive correlation between the reduction of total synovitis for individual joints at the 1-year follow-up with the improvement of the Insall knee score by arthroscopic synovectomy. These findings suggest a link between surgical radicalism and clinical improvement and further emphasize the need for radical and thorough removal of inflamed ST.6,13 It has recently been reported that the clinical improvement in RA patients with therapy-resistant synovitis following combined arthroscopic and radiation synovectomy is superior to that following either 1 of the 2 individual methods.7,13 In the former multicenter study (33-month follow-up),7 the combined approach induced a higher benefit with respect to erythrocyte sedimentation rate, Insall knee score, modified Lysholm score, Lequesne score, joint pain as measured on a 4-point scale, range of motion, and presence of effusion and synovitis (P ⱕ .05 for all parameters). Although there are as yet no studies comparing the degree of ST removal after arthroscopic synovectomy, radiation synovectomy, or combination therapy, it appears likely that more radical reduction of inflammatory infiltrates by combination therapy contributes to its augmented therapeutic efficacy. However, it is possible also that the efficacy of monotherapy by arthroscopy could be improved by more careful, radical, and thorough excision of inflamed ST, applying particular care to the critical regions identified in the present report. Future research may have to include similar studies with larger patient numbers and also biopsy and clinical outcome evaluations after a longtime follow-up, to determine the therapeutic importance of the reduction of acute and chronic inflammatory activity for arthroscopic synovectomy. CONCLUSION Although arthroscopic synovectomy effectively reduces acute inflammation of ST in RA, chronic inflammation is diminished to a more limited degree. Also, the elimination of ST is less radical in areas relatively spared from surgery in order to avoid ligament damage, the elimination of ST is less radical. Therefore, in patients with RA who have persistent knee synovitis, the continued presence of chronically inflamed ST following arthroscopic synovectomy suggests either more thorough removal of inflamed ST or the application of additional measures, e.g., the combination of arthroscopic and radiation synovectomy. Acknowledgment: Dr. E. Palombo-Kinne is gratefully acknowledged for critical revision of the manuscript, U.
1217
Vorderwülbecke, B. Müller, J. Harzendorf, and B. Niescher for excellent technical assistance.
REFERENCES 1. Klug S, Willauschus W, Weseloh G. Arthroscopic synovectomy of the knee joint. Arthroscopy 1997;16:262-267. 2. Kresnik E, Mikosch P, Gallowitsch HJ, et al. Clinical outcome of radiosynoviorthesis: A meta-analysis including 2190 treated joints. Nucl Med Commun 2002;23:683-688. 3. Gibbons CE, Gosal HS, Bartlett J. Long-term results of arthroscopic synovectomy for seropositive rheumatoid arthritis: 6-16 year review. Int Orthop 2002;26:98-100. 4. Roch-Bras F, Daures JP, Legouffe MC, Sany J, Combe B. Treatment of chronic knee synovitis with arthroscopic synovectomy: Long-term results. J Rheumatol 2002;29:1171-1175. 5. Horiuchi K, Momohara S, Tomatsu T, Inoue K, Toyama J. Arthroscopic synovectomy of the elbow in rheumatoid arthritis. J Bone Joint Surg Am 2002;84:342-347. 6. Hagena FW. Development of rheumatoid arthritis surgery in German-speaking countries. Orthopaede 2001;30:756-767. 7. Klug S, Wittmann G, Weseloh G. Arthroscopic synovectomy of the knee joint in early cases of rheumatoid arthritis: Follow-up results of a multicenter study. Arthroscopy 2000;16: 262-267. 8. Toda Y. The relationship between prognosis after one-day arthroscopic partial synovectomy on rheumatoid knee joints and the histologic features of the synovial membrane. Ryumachi 2003;43:549-556. 9. Tanaka N, Sakahashi H, Sato E, Ishi S. Immunohistological indication for arthroscopic synovectomy in rheumatoid knees: Analysis of synovial samples obtained by needle arthroscopy. Clin Rheumatol 2002;21:46-51. 10. Jüsten HP, Pommer S, Leeb I, Pilhofer C, Kisslinger E, Wessinghage D. Long-term results of open knee synovectomy in later cases of rheumatoid arthritis. Z Rheumatol 1999;58:20012006. 11. Bynum CK, Tasto J. Arthroscopic treatment of synovial disorders in the shoulder, elbow, and ankle. J Knee Surg 2002; 15:57-59. 12. Muro H, Waguri-Nagaya Y, Otsuka T, Matsui N, Asai K, Kato T. Serum gliostatin levels in patients with rheumatoid factornegative and -positive rheumatoid arthritis and changes of these levels after surgical treatment. Clin Rheumatol 2001;20: 331-336. 13. Kerschbaumer F, Kandziora F, Herresthal J, Hertel A, Hor G. Combined arthroscopic and radiation synovectomy in rheumatoid arthritis. Orthopaede 1998;27:188-196. 14. Arnett FC, Edworthy SM, Bloch DA, et al. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 1988;31:315324. 15. Larsen A, Dale K, Eek M. Radiographic evaluation of rheumatoid arthritis and related conditions by standard reference films. Acta Radiol 1977;18:481-491. 16. Highgenboten CL. Arthroscopic synovectomy. Arthroscopy 1985;1:190-193. 17. Klein W, Jensen KU. Arthroscopic synovectomy of the knee joint: Indication, technique and follow-up results. Arthroscopy 1988;4:63-71. 18. Geiler G, Stiehl P. Significance of the concepts basic activity and actual activity in the morphological evaluation of synovial membranes in rheumatoid arthritis. Z Rheumatol 1974;33:7386. 19. Duke O, Panayi GS, Janossy G, Poulter LW. Immunohistological analysis of the lymphocytic infiltrates of rheumatoid
1218
20. 21.
22. 23. 24.
25.
26. 27. 28.
29. 30.
H.-D. CARL ET AL.
synovial membrane using monoclonal antibodies. Ann Rheum Dis 1982;41:192-193. Young CL, Adamson TC, Vaughan JH, Fox RI. Immunohistologic characterization of synovial membrane lymphocytes in rheumatoid arthritis. Arthritis Rheum 1984;27:124-130. Sack U, Kinne RW, Marx T, Heppt P, Bender S, Emmrich F. Interleukin-6 in synovial fluid is closely associated with chronic synovitis in rheumatoid arthritis. Rheumatol Int 1993; 13:45-51. Knapp W. Leukocyte typing IV. Oxford: Oxford University, 1990. Insall JN, Dorr DD, Scott RD, Scott WN. Rationale of the knee society clinical rating system. Clin Orthop 1989;248:13-14. Lequesne MG, Mery C, Samson M, Gerard P. Index of severity of osteoarthritis for the hip and knee. Validation-value in comparison with other assessment tests. Scand J Rheumatol 1987;65:85-89. Smeets TJ, Kraan MC, Galjaard S, Youssef PP, Smith MD, Tak PP. Analysis of the cell infiltrate and expression of matrix metalloproteinases and granzyme B in paired synovial biopsy specimens from the cartilage-pannus junction in patients with RA. Ann Rheum Dis 2001;60:561-565. Mulherin DM, Fitzgerald O, Breshnihan B. Synovial tissue macrophage populations and articular damage in rheumatoid arthritis. Arthritis Rheum 1996;39:115-124. Firestein GS. Novel therapeutic strategies involving animals, arthritis, and apoptosis. Curr Opin Rheumatol 1998;10:236241. Tak PP, Taylor PC, Breedveld FC, et al. Decrease in cellularity and expression of adhesion molecules by anti-tumor necrosis factor alpha monoclonal antibody treatment in patients with rheumatoid arthritis. Arthritis Rheum 1996;39:1077-1081. Tak PP. Editorial. Lessons learnt from the synovial tissue response to anti-rheumatic treatment. Rheumatology 2000;39: 817-820. Smith MD, Slavotinek J, Au V, et al. Successful treatment of rheumatoid arthritis is associated with a reduction in synovial
31.
32.
33. 34.
35.
36.
37. 38. 39.
membrane cytokines and cell adhesion molecule expression. Rheumatology 2001;40:965-977. Kraan MC, Reece RJ, Barg EC, et al. Modulation of inflammation and metalloproteinase expression in synovial tissue by leflunomide and methotrexate in patients with active rheumatoid arthritis. Findings in a prospective, randomized, doubleblind, parallel-design clinical trial in thirty-nine patients at two centers. Arthritis Rheum 2000;43:1820-1830. Wong P, Cuello C, Bertouch JV, et al. The effects of pulse methylprednisolone on matrix metalloproteinase and tissue inhibitor of metalloproteinase-1 expression in rheumatoid. Rheumatology 2000;39:1067-1073. Barrera P, Blom A, van Lent PL, et al. Synovial macrophage depletion with clodronate-containing liposomes in rheumatoid arthritis. Arthritis Rheum 2000;43:1951-1959. Paus AC, Pahle JA. Arthroscopic evaluation of the synovial lining before and after open synovectomy of the knee joint in patients with chronic inflammatory joint disease. Scand J Rheumatol 1990;19:193-201. Paus AC, Refsum S, Forre O. Histopathologic changes in arthroscopic synovial biopsies before and after open synovectomy in patients with chronic inflammatory joint diseases. Scand J Rheumatol 1990;19:202-208. Paus AC, Mellbye OJ, Forre O. Immunohistopathologic findings in synovial biopsies before and after synovectomy in patients with chronic inflammatory joint diseases and their relation to clinical evaluation. A prospective study of biopsies taken from areas selected by arthroscopy. Scand J Rheumatol 1990;19:269-279. Kinne RW, Bräuer B, Stuhlmüller B, Palombo-Kinne E, Burmester GR. Macrophages in rheumatoid arthritis. Arthritis Res 2000;2:189-202. Smolen J, Steiner G. Therapeutic strategies for rheumatoid arthritis. Nature Reviews 2003;2:473-488. Kinne RW, Palombo-Kinne E, Emmrich F. T-cells in the pathogenesis of rheumatoid arthritis—Villains or accomplices. Biochimica Biophysica Acta 1997;1360:109-141.