14 Habenicht A J R, Goerig M, Grulich J, et al. Human platelet-derived growth factor stimulates prostaglandin synthesis by activation and by rapid de novo.
210
Bentley, McAuliffe 23 Gordon D, Bray M A, Morley J. Control of lymphokine secretion by prostaglandins. Nature 1976; 262: 401-2. 24 Knudsen P J, Dinarello C A, Strom T B. Prostaglandins posttranscriptionally inhibit monocyte expression of interleukin 1 activity by increasing intracellular cyclic adenosine monophosphate. J Immunol 1986; 137: 3189-94. 25 Kunkel S L, Spengler M, May M A, Spengler R, Larrick J, Rennick D. Prostaglandin E2 regulates macrophage-derived tumor necrosis factor gene expression. Biol Chem 1988; 263: 5380-4. 26 Rola-Pleszczynski M, Lemaire I. Leukotrienes augment interleukin 1 production by human monocytes. J Immunol 1985; 135: 3958-61. 27 Rola-Pleszczynski M. Immunoregulation by leukotrienes and other lipoxygenase metabolites. Immunology Today 1985; 6: 302-6. 28 Shirakawa F, Yamashita U, Chedid M, Mizel S B. Cyclic AMP-an intracellular second messenger for interleukin 1. Proc Natl Acad Sci USA 1988; 85: 8201-5. 29 Zhang Y, Lin J X, Yip Y K, Vilcek J. Enhancement of cAMP levels and of protein kinase activity by tumor necrosis factor and interleukin 1 in human fibroblasts: role in the induction of interleukin 6. Proc Natl Acad S'ci USA 1988; 85: 6802-5. 30 Dayer J M, Goldring S R, Robinson D R, Krane S M. Effects of human mononuclear cell factor on cultured rheumatoid synovial cells. Interactions of prostaglandin E2 and cyclic adenosine 3'5-monophosphate. Biochem Biophys Acta 1979; 586: 87-105. 31 Johnson H M, Russell J K, Torres B A. Second messenger role of arachidonic acid and its metabolites in interferon-y production. J Immunol 1986; 137: 3053-6. 32 Bourne H R. Who carries what message? Nature 1989; 337: 504-5.
14 Habenicht A J R, Goerig M, Grulich J, et al. Human platelet-derived growth factor stimulates prostaglandin synthesis by activation and by rapid de novo synthesis of cyclooxygenase. J Clin Invest 1985; 75: 1381-7. 15 Heldin C H, Backstrom G, Ostman A, et al. Binding of different dimeric forms of PDGF to human fibroblasts: evidence for two separate receptor types. EMBO7 1988; 7: 1387-94. 16 Nister M, Hammacher A, Melistrom K, et al. A glioma-derived PDGF A chain homodimer has different functional activities than a PDGF AB heterodimer purified from human platelets. Cell 1988; 52: 791-9. 17 Brom J, Knoller J, Koller M, Konig W. Tumour necrosis factors modulate the affinity state of the leukotriene B4 receptor on human neutrophils. Immunology 1988; 65: 647-9. 18 Dewhirst F E, Ago J M, Peros W J, Stashenko P. Synergism between parathyroid hormone and interleukin 1 in stimulating bone resorption in organ culture. 7ournal of Bone Mineral Research 1987; 2: 127-34. 19 Akahoshi T, Oppenheim J J, Matsushima K. Interleukin- 1 stimulates its own receptor expression on human fibroblasts through the endogenous production of prostaglandins. J Glin Invest 1988; 82: 1219-24. 20 Elias J A. Tumor necrosis factor interacts with interleukin- 1 and interferons to inhibit fibroblast proliferation via fibroblast prostaglandin-dependent and prostaglandin-independent mechanisms. Am Rev Respir Dis 1988; 138: 652-8. 21 Browning J. Interferons and rheumatoid arthritis: Insight into interferon biology? Immunology Today 1987; 8: 3724. 22 Hart P H, Whitty G A, Piccoli D S, Hamilton J A. Control by IFN-y and PGE2 of TNF( and IL-l production by human monocytes. Immunology 1989; 66: 376-83.
Pigmented villonodular synovitis In 1941 Jaffe, Lichtenstein, and Sutro introduced the term pigmented villonodular synovitis (PVNS) to describe a 'yellow-brown tumour-like tenosynovial lesion'.' Before their paper this condition had gone under a variety of names, many of which implied a neoplastic disorder. Indeed the earliest descriptions assumed it was a malignant disease, and amputation was commonly practised. The first report was by Chassaignac in 1852, who described a nodular lesion occurring in the flexor tendons of the hand.2 In 1909 Moser reported a diffuse type of lesion in the ankle.3 It was not until the paper of Jaffe et al, however, that a clear overall description of the condition was given. Reporting on 20 of their own cases affecting joints, tendon sheaths, and bursae, Jaffe et al defined them as either circumscribed or diffuse and showed that the histological appearances were similar in both types of articular synovitis and also in pigmented villonodular tenosynovitis and bursitis. They described the salient histological features: deposition of haemosiderin and infiltration of histiocytes and giant cells in a fibrous stroma within the synovium of tendon sheaths and large joints. They suggested that the condition represented an inflammatory response to an unknown agent. In their experience the treatment of choice was complete excision, and recurrence was due to inadequate removal of diseased tissue. Recurrences were never malignant and could be adequately treated by radiotherapy. Much has been written about this condition since 1941, but we know little more now than that written in this classic paper. Research into PVNS is difficult because of the rarity of the disorder and most series have had very few patients or have grouped cases from different anatomical sites. In a rare epidemiological study of this condition Myers et al found the incidence to be 1-8 per million population.4 This is probably an underestimate as they looked only at patients who had undergone surgery. They found a higher incidence in men, though other investigators have not always found this sex difference. The age range was from 11 to 84, but as with other series young men were most often affected and the knee was the most commonly involved joint. They were unable to find any clear aetiological associations. The cause and pathogenesis of PVNS remain obscure.
There have been many attempts to incriminate trauma as the cause, but no study clearly validates this claim. Many varied substances have been injected into the joints of experimental animals, but none has ever reproduced the typical clinical or histological features of PVNS. Myers et al felt that chronic repetitive trauma and haemarthrosis were important factors. The condition is not seen in haemophiliacs, however, nor in patients with Charcot's joints. No organisms have ever been consistently isolated from a lesion of PVNS, though there is one isolated report of the finding of a myxovirus-like structure from the knee in an unusual bilateral case.5 The likelihood of an infective cause seems remote. The most likely cause seems to be inflammatory, but the noxious agent remains unidentified. Lipids have been suggested as they are seen in increased concentration intracellularly in PVNS. There is no evidence for any systemic lipid abnormality, but Hirohata suggested a localised metabolic disturbance as the source of the lipids.6 Intra-articular injection of lipids does not induce PVNS, however, and probably, as Jaffe pointed out, the presence of lipids in the lesions is a secondary phenomenon as seen in other inflammatory lesions around bone. Most authors since Jaffe have thought that PVNS is a non-neoplastic process, but there are still some dissenters.7 Certainly the condition is non-malignant clinically and there are no recorded deaths from PVNS nor any proved incidents of metastasis.
Clinical features The classical clinical picture of PVNS is that of a young man with monoarticular involvement of the knee with the diffuse form of the disease. Other leg joints may be involved but virtually always in isolation. The nodular form tends to affect the fingers but can occur in the knee, where it may mimic a meniscal lesion.8 The course of the disease is insidious with slow progression of symptoms being the rule. Discomfort rather than pain is a common complaint and swelling is invariably present. In patients with longstanding disease stiffness is a feature. Routine haematological, biochemical, and immunological
Pigmented villonodular synovitis
tests of peripheral blood have always failed to disclose any consistent abnormality.4 9 Aspiration of the knee usually shows xanthochromic fluid, though on occasions bloody fluid may be obtained. Analysis of the synovial fluid from involved joints has never shown any distinctive changes to differentiate PVNS from other causes of an effusion.4 The earliest report on radiological features of PVNS was by Lewis in 1947,10 but the most substantial series remains that of Smith and Pugh from the Mayo clinic, who reported on 38 patients with histologically proved PVNS." In patients with moderate involvement soft tissue synovial swelling is seen and the nodular nature of this may give a clue to the diagnosis. In patients with more advanced disease smooth pressure defects develop in the adjacent para-articular region, and only in these patients is loss of joint space seen with irregularity of the articular surfaces. Extra-articular extension and cortical erosion are rare but may occur. Calcification is very rare and is more consistent with a synovial sarcoma. Osteoporosis is surprisingly uncommon even in patients with advanced disease. These changes are non-specific and plain radiography cannot therefore confirm the diagnosis. In the past arthrography has been widely used in diagnosis of PVNS and shows filling defects. Johansson et al, however, found that arthrography was diagnostic in only one third of cases and in most patients was normal. One might hope that newer diagnostic imaging methods might improve upon this low success rate but to date there have been few reports of the use of computed tomographic scanning in PVNS and none of magnetic resonance imaging. The latter probably holds most promise for the future. Despite the widespread acceptance of the use of arthroscopy as a diagnostic tool, in recent years there have been surprisingly few reports of the use of the arthroscope in PVNS. 2 3 Obviously, when combined with synovial biopsy it offers the best method of diagnosis.
21
Apparently, the best strategy for the patient with diffuse involvement of the knee is to treat them on a symptomatic basis, to accept that complete synovectomy is difficult to achieve and carries a high morbidity when performed as an open procedure, and to perform a lesser, closed procedure, which can easily be repeated as necessary if symptoms are severe enough. It is to be expected that in the future, with the use of the newly available powered arthroscopy tools, that closed synovectomy will be used. This combined with the use of continuous passive motion after operation should reduce the morbidity of this procedure. Arthroscopic synovectomy may not be as complete as that performed by an arthrotomy but is of much lower morbidity and can be more readily repeated if necessary. In fact there is some evidence that PVNS may be a self-limiting condition, with remission occasionally following biopsy alone. 14 In other joints apart from the knee similar principles apply. In the hip and shoulder, however, diagnosis is often made late, at the stage when secondary degenerative changes have already occurred, and the only option then available is arthrodesis or arthroplasty. These have also been used as salvage procedures in the knee. As most of these patients are young, arthrodesis is generally preferred. Perhaps the most important factors to bear in mind when treating PVNS are that it is not malignant and should not be treated as such and that early diagnosis and treatment are essential to reduce the incidence of recurrence and possibly of joint destruction. We can only hope that in the future our knowledge will advance further than it has in the last half century since Jaffe's classic account. Royal National Orthopaedic Hospital, Brockley Hill, Stanmore, Middlesex HA7 4LP Orthopaedic Department, Whipps Cross Hospital,
GEORGE BENTLEY
T McAULIFFE*
Leytonstone,
London Eli
Treatment In the absence of any clear knowledge of the pathogenesis, or of any large prospective studies, treatment of PVNS in the past has been predominantly surgical and largely pragmatic. It is clear that this is a non-malignant disorder and therefore radical treatment is not indicated. On the other hand, suboptimal treatment will lead to recurrence. There is general consensus among most authors that simple excision of the localised form of the condition is adequate, producing good function postoperatively and a low rate of recurrence. Treatment of diffuse PVNS has been less satisfactory. Most authors advocate total synovectomy in the knee, but even this may not be curative, presumably owing to the difficulty of performing a truly total synovectomy. Johansson reported 33% recurrence after synovectomy in the knee and in the Royal National Orthopaedic Hospital series it was nearly 50%. 14 In addition, the price to be paid in terms of function of the knee may be high as stiffness and pain are common sequelae after operation. Radiotherapy has been commonly used in the past, mostly for recurrences but also for primary disease.'5 Although there have never been any reported cases of malignancy following irradiation for PVNS, most recent commentators have felt that the potential risks of radiation treatment were too high in young patients with a benign condition. There may, however, be a place for it in recurrent, severely symptomatic disease in the elderly. Radiation synovectomy has been widely used in rheumatoid arthritis,'6 but there is only a solitary case report of its use in PVNS. 7 Potentially it carries the same risks as radiotherapy and should probably be avoided in the younger patient.
*Correspondence to Mr McAuliffe.
1 Jaffe H L, Lichtenstein L, Sutro C J. Pigmented villonodular synovitis, bursitis and tenosynovitis. Archives of Pathology 1941; 31: 731-65. 2 Chassaignac M. Cancer de la gaine des tendons. Gazette Hopital Civil Militaire 1852; 47: 185-6. 3 Moser E. Primares sarkom der fussgelenkkapsel. Exstirpaton Dauerheilung Deutsche Zeitschrifte fur Chirurgie 1909; 98: 306-10. 4 Myers B W, Masi A 1, Feigenbaum S L. Pigmented villonodular synovitis and tenosynovitis. A clinical epidemiological study of 166 cases and literature review. Medicine (Baltimore) 1980; 59: 223-38. 5 Molnar Z, Stern W H, Stoltzner G H. Cytoplasmic tubular structure in pigmented villonodular synovitis. Arthritis Rheum 1971; 14: 784. 6 Hirohata K. Light microscopic and electron microscopic studies of individual cells in pigmented villonodular synovitis and bursitis. KobeJ Med Sci 1968; 14: 251-79. 7 Rao A S, Vigorita V J. Pigmented villonodular synovitis (giant cell tumour of tendon sheath and synovial membrane). A review of 81 cases. J Bone Joint Surg [Am] 1984; 66: 76-94. 8 Granowitz S P, Mankin H J. Localised pigmented villonodular synovitis of the knee. Report of five cases. J Bone joint Surg [Am] 1967; 49: 122-8. 9 Johansson J E, Ajjouls S, Coughlin L P, Wener J A. Cruess R L. Pigmented villonodular synovitis of joints. Clin Orthop 1982; 163: 159-66. 10 Lewis R W. Roentgen diagnosis of pigmented villonodular synovitis and synovial sarcoma of the knee joint. Preliminary report. Radiology 1947; 49: 26-38. 11 Smith J H, Pugh D G. Roentgenographic aspects of articular pigmented villonodular synovitis. AIR 1962; 87: 1146-56. 12 Flandry F C, Jacobson K E, Andrews J R. Localised pigmented villonodular synovitis of the knee mimicking meniscal injury. Arthroscopy 1986; 2: 217-21. 13 Dorfman H. Arthroscopic detection of synovial disorders. Contemporary Orthopedics 1985; 10: 19-29. 14 Byers P D, Cotton R E, Deacon 0 W, et al. The diagnosis and treatment of pigmented villonodular synovitis. J Bone Joint Surg [Br] 1968; 50: 290-305. 15 Friedmann M, Schwartz E E. Irradiation therapy of pigmented villonodular synovitis. Bulletin of the Hospital for Joint Diseases 1957; 18: 19-32. 16 Sledge C B, Atcher R W, Shortkroff S, Anderson R J, Bloomer W D, Hurson B J. Intra-articular radiation synovectomy. Clin Orthop 1984; 182: 37-40. 17 Wiss D A. Recurrent villonodular synovitis of the knee. Successful treatment with yttrium 90. Clin Orthop 1982; 169: 139-44.
