Pigmented villonodular synovitis is a rare benign disorder of the synovial tissue. The etiology is unknown1,13,15, although recent evidence supports a benign neoplastic origin3,18,21,24. The knee is the most commonly involved large joint6,11,13,14,17. Both diffuse and focal forms of the disease occur in the knee. The two forms have an identical histological appearance, but they are quite different with regard to the clinical presentation and the disease process.
Patients who have focal pigmented villonodular synovitis usually have mechanical symptoms, such as locking and catching. A localized, or focal, area of the synovial tissue, which often is pedunculated, is involved. These patients can be well managed with partial synovectomy. Ogilvie-Harris et al. (in a study of five patients), Johansson et al. (in a study of seven), and Donde and Funding (in a study of six) reported no recurrences after partial synovectomy. Schwartz et al. and Byers et al. both reported two recurrences after partial synovectomy in twelve and thirteen patients, respectively, and Rao and Vigorita reported one recurrence in eight patients.
The diffuse form of the disease is much more difficult to treat because the entire synovial membrane in both the anterior and the posterior aspect of the joint typically is involved6,7. Treatment is complicated by the relatively high rate of recurrence. Byers et al. reported recurrence in eleven of twenty-four patients, and Schwartz et al. documented recurrence in twenty-three of sixty-three patients. It has been recommended that all of the synovial tissue be removed with use of both anterior and posterior capsulotomy to limit the chance of recurrence6,9,19,22. In addition, notable morbidity has been associated with the treatment. Total synovectomy through anterior and posterior arthrotomy can result in atrophy, fibrosis, and a decreased range of motion2,9,14. Arthroscopic total synovectomy has been performed in an attempt to lessen these problems. Ogilvie-Harris et al. reported good short-term results in eleven patients who had had a complete synovectomy for the treatment of diffuse disease. One recurrence was noted at 3.5 years, but otherwise all patients had a decrease in pain and an improvement in function19.
Patients who have diffuse pigmented villonodular synovitis may have destructive changes in the joint, including periarticular cysts, erosions, and narrowing of the joint space8,23. Total synovectomy alone may not relieve the symptoms in these patients. Treatment options include arthrodesis or total joint replacement, depending on the age, medical condition, and functional demands of the patient. We are not aware of any reports of the results of total knee arthroplasty in patients who had pigmented villonodular synovitis. The purpose of the current paper is to report our experience with total knee arthroplasty in such patients, focusing on recurrence, survival of the implant, and function.
*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.
†Department of Orthopedics, Mayo Clinic, 200 First Street S.W., Rochester, Minnesota 55905.
The records of all eighteen patients who had had a total knee arthroplasty for the treatment of biopsy-proved pigmented villonodular synovitis between 1974 and 1992 were reviewed. The review included preoperative and postoperative anteroposterior and lateral radiographs, made with the patient standing; pathology and operative reports; and clinical notes. The knee scores and functional scores were calculated according to the clinical rating system of The Knee Society, with assignment of a maximum of 100 points for each12.
Nine of the patients were men and nine were women. At the time of the arthroplasty, fifteen patients had evidence of active disease both grossly and histologically: eleven patients had diffuse disease, and four had focal disease. The three remaining patients had a history of diffuse pigmented villonodular synovitis but no evidence of disease at the time of the arthroplasty.
We established two study groups on the basis of the underlying abnormality. Group I consisted of patients who had diffuse pigmented villonodular synovitis, and Group II consisted of those who had focal pigmented villonodular synovitis. Group I was further subdivided into patients who had active (IA) and those who had inactive (IB) disease.
No patient was lost to follow-up. One patient (Case 5) in Group IA (active diffuse disease) died ten years after the total knee replacement; the prosthesis was functioning well at the time of death. The overall duration of follow-up averaged 10.8 years (range, 3.6 to 21.9 years: 11.9 years (range, 3.6 to 21.9 years) for Group IA, 10.1 years (5.3, 10.2, and 14.7 years) for Group IB, and 8.4 years (range, 5.4 to 10.5 years) for Group II.
Group IA: Active Diffuse Disease (Table I)
Group IA consisted of eleven patients (six men and five women) whose average age at the time of the arthroplasty was fifty-nine years (range, thirty-six to eighty-two years). All eleven patients had a complete synovectomy in combination with the arthroplasty. Three polycentric prostheses (one of which was unicondylar) and eight condylar prostheses were used. Six of the condylar implants were inserted with cement, one was inserted with hybrid fixation (the femoral component was inserted without cement, and the tibial component was inserted with cement), and one was inserted without cement.
Before the operation, all eleven patients had pain in the knee that limited function. Nine patients had synovitis; it was mild in one, moderate in five, and severe in three. Nine patients had an effusion, which was mild in two, moderate in four, and severe in three. Four patients had a history of a bloody effusion. Eight patients had a flexion contracture, which averaged 12 degrees (range, 2 to 30 degrees). The average flexion for the entire group was 108 degrees. The knee and functional scores averaged 48 points (range, 27 to 71 points) and 43 points (range, 0 to 80 points), respectively.
The disease was diagnosed before the arthroplasty in seven patients, it was suspected in two, and it was not diagnosed until the time of the operation in two. The interval between the diagnosis or the symptoms of pigmented villonodular synovitis and the arthroplasty averaged eight years (range, five months to twenty-three years). Six patients had a total of fourteen recurrences before the arthroplasty; of these six, five had a total of eight repeat synovectomies. All six patients had the total knee arthroplasty at the time of the last recurrence. One patient was managed with external beam irradiation before the arthroplasty.
Nine patients had a total of nineteen previous operations, including the eight open synovectomies, two removals of a Baker cyst, one proximal tibial osteotomy for the treatment of degenerative joint disease of the medial compartment, one open reduction and internal fixation for the treatment of a fracture of the proximal aspect of the tibial shaft, and seven arthroscopies. One excision of the proximal aspect of the fibula, one lateral meniscectomy, and one patelloplasty were performed in combination with three of the open synovectomies. Five of the arthroscopic procedures were total synovectomies, and two were performed for the purpose of obtaining tissue to confirm the suspected diagnosis.
The preoperative radiographs showed destructive changes in the joints. Eight patients had erosions, seven had cysts, and all eleven had narrowing of the joint space (which was predominantly medial in four patients, lateral in two, and symmetrical in five). Additionally, nine patients had soft-tissue swelling, three had osteophytes, and six had subchondral sclerosis.
Group IB: Inactive Diffuse Disease (Table I)
Group IB consisted of three patients (one man and two women) who were sixty, sixty, and seventy-three years old at the time of the total knee arthroplasty. All three patients had a history of biopsy-proved diffuse pigmented villonodular synovitis. None of these patients was managed with a synovectomy in combination with the arthroplasty. There was no obvious sign of gross disease at the time of the arthroplasty in two of the patients; in the third patient, pigmented villonodular synovitis was suspected, but histological examination of a biopsy specimen revealed negative findings. Three different condylar designs of prosthesis were used, and all were inserted with cement.
Before the operation, all three patients had pain in the knee that limited function. Physical examination revealed mild synovitis in two patients and a mild effusion in all three. One patient had a history of a bloody effusion. All three patients had a flexion contracture (of 3, 15, and 20 degrees; average, 13 degrees), and the average flexion was 105 degrees. The knee and functional scores were 12, 28, and 70 points (average, 37 points) and 0, 45, and 50 points (average, 32 points), respectively.
All three patients had previous operations, including three open synovectomies and one arthroscopy. One medial and lateral meniscectomy was performed in combination with a synovectomy. The one arthroscopic procedure was a partial synovectomy for the purpose of biopsy and diagnosis.
The interval between the diagnosis and the arthroplasty was four, fifteen, and twenty-eight years (average, 15.7 years). No patient had a recurrence of the disease between the previous synovectomy and the arthroplasty. One patient had external beam irradiation before the arthroplasty.
Preoperative radiographs revealed destructive changes in the joints in all three patients. One patient had erosions, two had cysts, and all three had narrowing of the joint space (which was predominantly medial in one, lateral in one, and both medial and lateral in one). No patient had soft-tissue swelling, two had osteophytes, and all three had subchondral sclerosis.
Group II: Focal Disease (Table I)
Group II consisted of four patients (two men and two women) whose average age at the time of the arthroplasty was sixty-three years (range, fifty-six to seventy-three years). All four patients had the total knee arthroplasty because of degenerative joint disease; the focal pigmented villonodular synovitis was diagnosed on the basis of histological examination of synovial tissue. All four patients had a partial synovectomy in combination with the arthroplasty. Four different condylar designs of prosthesis were used, and all were inserted with cement.
Before the operation, all four patients had pain in the knee that limited function. None had mechanical symptoms, such as catching or locking. Physical examination revealed mild synovitis in one patient and a mild effusion in two. Three patients had a flexion contracture of 5, 5, and 15 degrees (average, 8 degrees). The average flexion for all four patients was 115 degrees. The knee and functional scores averaged 45 points (range, 40 to 50 points) and 64 points (range, 50 to 80 points), respectively. No patient had had a previous operation.
Degenerative joint disease was evident on all preoperative radiographs. All four patients had narrowing of the joint space, which was predominantly medial in two and symmetrical in the other two. No patient had soft-tissue swelling, cysts, or erosions. All four patients had osteophytes and subchondral sclerosis.
Group IA: Active Diffuse Disease (Table I)
Of the eleven patients in this group, seven (including the one who died) had an intact, well functioning prosthesis and no evidence of recurrence at the latest follow-up evaluation (or the time of death), an average of 10.8 years (range, 3.6 to 20.1 years) postoperatively. These seven patients were functioning well, with an average knee score of 88 points (range, 80 to 99 points) and an average functional score of 74 points (range, 20 to 100 points).
Two patients needed manipulation under anesthesia. One of them (Case 5) had three such manipulations without improvement and subsequently had a posterior capsular release; ten years after the arthroplasty, she had a range of motion of 0 to 80 degrees. All patients eventually achieved full extension except for one who had a persistent flexion contracture of 20 degrees at 16.7 years. The average flexion for the seven patients who had an intact prosthesis was 102 degrees.
Six of the total condylar prostheses and one of the polycentric implants were well fixed at the latest follow-up evaluation. Radiolucent lines were present about three tibial components; all were at the bone-cement interface, measured less than one millimeter in width, were non-circumferential, and were non-progressive. One femoral component had a non-progressive radiolucent line that was less than one millimeter in width at the component-cement interface in zone 4, according to The Knee Society total knee arthroplasty roentgenographic evaluation and scoring system5.
Of the eleven patients, three had a revision because of aseptic loosening. The unicondylar polycentric prosthesis was revised twelve years after the arthroplasty, one polycentric prosthesis was revised seven years after the arthroplasty, and one total condylar prosthesis that had been inserted with hybrid fixation was revised five years after the arthroplasty. The patient (Case 1) who had the unicondylar polycentric implant had gross loosening of the tibial component. The symptoms were not directly related to the pigmented villonodular synovitis; rather, she had pain at the lateral joint line and instability in the region of the loose component. At the time of the revision arthroplasty and repeat synovectomy, the pigmented villonodular synovitis was noted to have recurred, but it was not directly involved in the loosening of the component and it was restricted to the suprapatellar region. Ten years after the revision, the implant was well fixed; the knee score was 91 points, the functional score was 80 points, and there was no evidence of recurrence. The other two patients had the revision arthroplasty at another institution and had no evidence of recurrence at the time of the latest follow-up.
Nine of the eleven patients had no evidence of recurrence at an average of 11.3 years (range, 3.6 to 20.1 years). The two remaining patients had a recurrence two and twelve years after the arthroplasty and a concomitant synovectomy. In one of them (Case 1), the recurrence was noted at the time of the revision arthroplasty, as mentioned earlier. The other patient (Case 8) had a recurrence at the age of fifty-six years, two years after the total knee arthroplasty and a complete synovectomy. He had pain and swelling in both the anterior aspect of the thigh and the popliteal region of the knee. Anterior and posterior open synovectomy was performed through two incisions in an attempt to retain the well fixed components. Histological analysis of biopsy specimens showed that the disease had invaded some of the muscles in the posterior aspect of the knee (Fig. 1). The patient did well for one year; he then began to have pain in the same locations as before. In addition to recurrence of the disease, he had a severe pain syndrome. He was managed with an above-the-knee amputation; histologically, the margins were clear. Three years after the amputation, he was functioning reasonably well with a prosthesis. He had a functional score of 70 points and no evidence of recurrence.
Group IB: Inactive Diffuse Disease (Table I)
All three patients in this group had an intact prosthesis at the latest follow-up evaluation and were functioning well. The average postoperative knee score was 91 points (92 points for two patients and 90 points for one), and the average functional score was 72 points (60 points for two patients and 95 points for one).
No patient had manipulation under anesthesia. All three eventually achieved full extension, and one had persistent limitation of flexion to 75 degrees. The average flexion at the latest follow-up evaluation was 98 degrees.
All components were well fixed. There were radiolucent lines at the bone-cement interface adjacent to two tibial components. The lines were less than one millimeter in width, were not circumferential, and were non-progressive. One femoral component was associated with a non-progressive radiolucent line, less than one millimeter in width, at the bone-cement interface in zone 1.
All patients were free of disease at the latest follow-up evaluation.
Group II: Focal Disease (Table I)
All four patients in this group had an intact prosthesis at the latest follow-up evaluation and were functioning well. The average postoperative knee score was 89 points (range, 83 to 97 points), and the average functional score was 86 points (range, 75 to 100 points).
One patient had manipulation under anesthesia. All four patients eventually achieved full extension. The average flexion at the latest follow-up evaluation was 98 degrees.
All components were well fixed. Radiolucent lines were not seen adjacent to any tibial component. One femoral component was associated with a radiolucent line, which was less than one millimeter in width, at the bone-cement interface in zone 1. All patients were free of disease at the latest follow-up evaluation.
Pigmented villonodular synovitis of the knee is a difficult problem for both the patient and the surgeon. A subset of patients who have the disease will eventually have severe degenerative joint disease, and synovectomy alone may not be sufficient to treat the pain. Very little has been reported in the literature regarding total knee arthroplasty in patients who have pigmented villonodular synovitis. Schwartz et al. described the experience at our institution with patients who had pigmented villonodular synovitis of large joints. Four patients were managed with total knee arthroplasty, but no outcome data were given. To our knowledge, the current report is the first to describe the results of total knee arthroplasty in patients who had pigmented villonodular synovitis.
Ten of the fourteen patients in Group I (diffuse disease) had an intact prosthesis at the latest follow-up evaluation, an average of 10.6 years after the arthroplasty. All ten implants were well fixed and functioning satisfactorily. All four failures were in the eleven patients who had active disease. Three failures were due to aseptic loosening that necessitated revision. Two of the prostheses that failed had a polycentric design, which is known to be associated with poorer long-term results than is a condylar design16,25. If the implants with a polycentric design are excluded, nine of the eleven prostheses survived an average of 9.5 years and were functioning well, with an average knee score of 90 points and an average functional score of 71 points.
With regard to the patients in Group II, we believe that the focal disease was a coincidence and was not directly related to the degenerative joint disease. Patients who have focal pigmented villonodular synovitis do not usually have the same destructive changes as patients who have diffuse disease8,17,23. All four of these patients had a well functioning knee prosthesis without evidence of recurrence at the latest follow-up evaluation. Focal pigmented villonodular synovitis at the time of a total knee replacement does not preclude a well functioning prosthesis.
Total knee arthroplasty in patients who have pigmented villonodular synovitis is not without problems. There were two recurrences in the fourteen patients in Group I, who had diffuse disease. This rate is similar to those reported in the literature for patients who have had total synovectomy alone2,9,21,22. Both recurrences were in patients who had extensive diffuse disease. Each had had more than one recurrence before the arthroplasty despite having had a number of previous procedures. Although one patient did well after a revision total knee arthroplasty and repeat synovectomy, the other patient subsequently had an above-the-knee amputation.
There are several options for the treatment of a recurrence of pigmented villonodular synovitis after a total knee arthroplasty, depending on the functional status of the patient and the status of the extremity. If the components are stable and all of the involved synovial tissue can be removed without jeopardizing the neurovascular status of the extremity, a repeat synovectomy can be done. If the disease has invaded the surrounding soft tissues, external beam or intra-articular irradiation may be performed in an attempt to preserve the function of the extremity10,20,26. Other options include an arthrodesis or possibly an amputation.
Several of our patients had difficulty achieving an adequate functional range of motion in the early postoperative period. Ten of the fourteen patients in Group I (seven of the eleven in Group IA and all three in Group IB) had limitations in extension or flexion, or both. Two patients had manipulation under anesthesia, and one had a posterior capsular release. The reason for postoperative stiffness is not entirely clear. Other authors have noted that patients who have diffuse pigmented villonodular synovitis often have difficulty in achieving a satisfactory range of motion both initially and after total synovectomy2,9,14. This is probably due to the disease process combined with the fact that many of these patients have had numerous previous operations.
Total knee arthroplasty, which is often superior for the treatment of degenerative joint disease, also is a viable option for patients who have diffuse pigmented villonodular synovitis of the knee with destructive joint changes. Total knee replacement can provide a pain-free, well functioning extremity in a patient who would otherwise have stiffness and pain. An extensive exposure of the knee joint can be obtained by excision of the posterior cruciate ligament, thus allowing thorough débridement of the entire synovial membrane. It should be emphasized that the use of total knee arthroplasty for the treatment of this disease is not without problems. Despite extensive débridement, the disease may recur. Diligence must be maintained in the postoperative period to ensure a functional range of motion. We currently recommend excision and substitution of the posterior cruciate ligament, in order to adequately debride the entire synovial membrane and to attempt to limit postoperative stiffness, when a total knee arthroplasty is performed in patients who have diffuse pigmented villonodular synovitis.