Historical Background

Pigmented villonodular synovitis (PVNS) is most often a monoarticular disease affecting patients in their third or fourth decade. The knee joint is most commonly affected and due to the large recesses in the knee, patients often present late after the disease has insidiously spread to extraarticular sites around the knee. Up to 50% of patients report a history of asso-ciated trauma. (18)

Although Chassaignac was the first to describe the lesion in 1852, it was Jaffe who proposed the name of pigmented vil-lonodular synovitis in 1941. (11) Jaffe thought the histology showed a fibrous stroma, pigment deposition, histiocyte infiltrate, and giant cells within the synovial membrane, hence the term pigmented villonodular synovitis. (11) Granowitz et al., (7) fur-ther classified PVNS according to whether there was a discrete lesion (localized) or a diffuse lesion.

Confusion still exists as to the etiology of PVNS and whether it is a true neoplasm or a reactive inflammatory process. (3,5,11,18,22,23,24) Without a clear understanding of the patho-physiology of the disease, treatment remains controversial. It is widely accepted that surgical resection often cures the localized form of PVNS with low rates of recurrence. In contrast, treatment of the diffuse form (DPVNS) remains controversial with high rates of recurrence and substantial morbidity after surgery. (2,3,4,6,9,16,19,20,22,25,28)

In our opinion, arthroscopic and limited open techniques (4,20,28) provide inadequate exposure of the knee joint, particularly postero-laterally and in the region of the popliteus tendon. Furthermore, most reports of treatment of DPVNS have small numbers of patients, often collected over a number of years with more than one joint included. (1,2,8,8,12,16,18,22) To the best of our knowledge, no reports in the literature contain large numbers of patients with DPVNS of the knee, with adequate preoperative staging, and long-term clinical or MRI evaluation. (15,17,21)

Materials and Methods

Over 40 consecutive patients referred to the Brigham & Women's Hospital between 1990 and 1999 with histologically confirmed DPVNS of the knee were interviewed and examined, and their clinical records reviewed. Outcome data was gath-ered from 38 patients who were referred after one or more failed arthroscopic synovectomy procedures to determine pre-and post-arthroscopic-synovectomy function. (20) The average age of the patients at index arthroscopy was 31.7 years (range 11 to 65 years). We also reviewed the records of 39 patients who underwent a standardized combined posterior and anteri-or open synovectomy by one surgeon (GWB) at our hospital. These patients were retrospectively assigned to one of three groups: group I received surgery alone (6 patients), group II had surgery and intra-articular radiation synovectomy using dysprosium-165 ( 165 Dy) (27 patients), and group III had surgery and external beam radiation (6 patients) performed three months postoperatively. Group I patients were more recent patients who were treated after the senior author (GWB) no longer favored the use of Dysprosium-165. Group II patients were mostly treated during a period when Dr. Clement Sledge and Sonya Shortkroff 26 were experimenting with Dysprossium-165 to treat DPVNS of the knee. Group III patients had post-operative MRI findings suggestive of residual disease larger than 5 millimeters. Outcome was assessed according to the clinical rating system of The Knee Society, with assignment of a maximum of 100 points for knee scores and functional scores. (10)

Figure 1 A-B Sagittal ( A ) and axial ( B ) T2-weighted MRI of the left knee show diffuse spread of DPVNS tissue circumferentially around the knee extending extra-articularly along the popliteus sheath, posterior cruciate ligament, gastrocnemius origin, and the semimembranosus bursa in a 40-year-old female with an insidious recurrence of DPVNS of the left knee 19.5 years after her first arthroscopic synovectomy.

Magnetic resonance imaging was used in all cases to identify sites of DPVNS infiltration preoperatively, to confirm adequate tissue removal at three months post-operatively prior to radiation therapy, and to follow patients annually thereafter.

Figure 2 A-B Sagittal ( A ) and axial ( B ) T2-weighted MRI of the left knee show complete removal of all DPVNS tissue at 3 months follow up. Clinical symptoms were markedly improved.

Surgical Technique

Open synovectomy was performed as a single staged procedure with the posterior approach being undertaken first. An S-shaped incision with the proximal limb based postero-laterally allows adequate exposure of the peroneal nerve proximally. The distal limb placed on the postero-medial side allows extensile exposure and excision of any large semimembranosus bursa that may be present. Extra-articular sites of DPVNS included the semimembranosus bursa, beneath both heads of the gastrocnemius muscles, posterior to the posterior cruciate ligament, the popliteus sheath, superior to the posterior aspects of the medial and lateral femoral condyles as well as previous arthroscopy portals. The posterior dissection included mobilization of both the medial and lateral heads of the gastrocnemius, exploration of the popliteus sheath, and detachment of the peripheral attachments of the menisci. A standard paramedian arthrotomy was used to expose the anterior aspect of the joint. Over three liters of irrigation was used during the posterior and anterior approaches to reduce the risk of leaving residual DPVNS tissue.

Postoperative management included epidural anesthesia for 48 hours with continuous passive motion (CPM) immediately post-op and for three weeks for 1.5 hours each day. Weight bearing as tolerated with a pair of crutches is allowed immediately post-op.

Results

The most common presenting symptoms were pain and swelling. Approximately half recalled trauma to the affected knee prior to being diagnosed with DPVNS, but less than 10% recalled trauma after they were diagnosed with DPVNS. Many patients had associated diagnoses such as medial and lateral meniscus tears, Baker's cyst, torn anterior cruciate ligament, and recurrent patella dislocation. More than half of the patients had one or more procedures after the first arthroscopic syn-ovectomy without eradication of their disease.(Fig. 1 A-B) Most patients reported some improvement of their symptoms fol-lowing arthroscopic synovectomy. However, all patients rated their symptoms and function as worse than prior to arthro-scopic synovectomies at the latest follow up of about four years (range 0.25 to 19.5 years) after the first arthroscopic synovec-tomy. Most patients lost some degree of knee motion longterm. The mean post-operative score for pain, function, range of motion, and synovitis was substantially worse than the preoperative scores in all patients who had recurrence after arthroscopic synovectomy.

In contrast, the patients who had the standardized open antero-posterior synovectomy had an average of a twenty-point improvement in Knee Society Scores for pain and function at an average of 4.5 years following the index procedure. Most had substantial improvements in post-operative knee motion. Most patients considered themselves improved at the latest follow up. There was less than 20% incidence of recur-rence on MRI.(Fig. 2 A-B) A majority of these patients had no clinical signs of a recurrence. Recurrence rate was zero in group I, and slightly higher in groups II and III. Three patients with early postoperative stiffness required closed manipulation under anesthesia before regaining almost full range of motion. Four patients with advanced preoperative degenerative arthri-tis continued to have pain postoperatively necessitating a total knee replacement.

Discussion

Recent reports of "complete" arthroscopic synovectomy suggest good results following arthroscopic synovectomy for DPVNS of the knee. (20,28) However, none of these reports include both preoperative and post-operative MRI assessment to docu-ment the extent of the disease or to follow the patients after arthroscopy. When MRI confirms extra-articular disease, arthro-scopic synovectomy, even in the best hands, is unlikely to eradicate extra-articular disease in hard to reach locations such as the popliteus sheath, medial and lateral gutters, behind the posterior cruciate ligament, and the infra-patellar fat pad. In the absence of MRI assessment, no comment can be made regard-ing the true extent of the disease in any series. Nevertheless, studies by Zvijac et al., (28) and Ogilvie et al., (20) have confirmed that arthroscopic synovectomy can provide short-term relief of symptoms associated with DPVNS of the knee. This conclusion was borne out in the present study with most patients experi-encing some subjective improvements in their symptoms in the short-term following arthroscopic synovectomy.

It has been reported that failure to eradicate DPVNS of the knee is associated with poor outcomes and a need for further surgery. (2,18,19,20) We have found similar results, with over half the patients who underwent arthroscopic synovectomy requiring one or more additional procedures. This finding is especially concerning in light of recent reports that have suggested that PVNS is a neoplastic proliferation of synovial fibroblasts and histiocytes. (3,5,22,23) Thus, failure to eradicate DPVNS by limited open techniques or arthroscopic synovectomies may predispose patients to locally disseminated disease, major symptoms, and subsequent loss of function. Furthermore, it is theoretically possible that arthroscopy and surgical manipulation could result in dissemination of the tumor beyond the joint space.

Conclusion

A critical review of prior reports and our experience does not support routine arthroscopic synovectomy alone or in combination with limited open synovectomy for the treatment of DPVNS of the knee, especially when the disease is also pre-sent in extra-articular sites. It is our current recommendation to treat any patient presenting with advanced DPVNS of the knee with the combined anterior and posterior open synovec-tomy technique, post-operative CPM, and weight bearing as tolerated. Extensive pre-operative degenerative joint disease predisposes patients to a poor result. Patients should be fol-lowed preoperatively and postoperatively with MRI assessment. We could not confirm the benefits of Dysprossium-165 or external beam radiation to treat DPVNS of the knee, since recurrences occurred following both types of treatment.

The results of our study suggest that complete synovectomy is the procedure of choice to decrease the risk of recurrence of DPVNS of the knee. Arthroscopic or limited open techniques or radiotherapy may reduce the amount of DPVNS tissue and associated symptoms in the short term, but will not eradicate the disease. In contrast, an extensile open surgical approach allows excellent visualization for removal of intra-and extra-articular DPVNS tissue with an acceptably low complication rate. We continue to use arthroscopy for pre- or post-operative diagnostic biopsies or treatment of early primary or residual disease following open synovectomy. Our current practice is to assess patients with MRI pre-operatively, three months post-operatively, and annually thereafter.