SUMMARY Inflammatory infiltration of the synovial membrane has been described in a proportion of cases of osteoarthritis (OA). Using conventional histology, lymphoid follicles, diffuse fibrosis, and perivascular fibrosis were shown to be present to a significantly greater extent and in more synovial membranes in osteoarthritis than in those cases where there was a mechanical or traumatic background to the joint disease. Calcium pyrophosphate dihydrate crystals (five patients) and detritic fragments of bone and cartilage (seven patients) were present in small numbers of the total cases of OA (38) studied. Neither of these features was related to the presence of an inflammatory infiltrate. Examination of 20 osteoarthritic synovial membranes using monoclonal antibodies showed the presence of lymphoid follicles containing T helper and T suppressor lymphocytes, B lymphocytes, and macrophages expressing HLA-DR in five cases. The T helper:suppressor ratio varied between 1:1 and 2-5:1 in these follicles. In addition, half of the OA samples, including these five cases, showed the presence of a diffuse cellular infiltrate containing T and B lymphocytes and macrophages, which were HLA-DR positive. Granulocytes were present in this diffuse infiltrate in those cases containing lymphoid follicles. The results confirm the presence of an inflammatory form of osteoarthritis but also show that the proportions of lymphoid cells are not the same as those considered to be typical of rheumatoid arthritis.

SUMMARY The histological features of the synovial membrane of peripheral joints in ankylosing spondylitis are similar to those seen in rheumatoid arthritis. There is intimal cell hyperplasia, a diffuse lymphocyte and plasma cell infiltrate, and formation of lymphoid follicles. Peroxidaseantiperoxidase staining shows the presence of IgG-, IgA-, and IgM-containing plasma cells in ankylosing spondylitis. The percentage of IgM-containing cells is significantly lower in ankylosing spondylitis than in rheumatoid arthritis.The histology of the synovial membrane of peripheral joints in ankylosing spondylitis has been examined in a small series of cases by Cruickshank,' Julkunen,2 and Wagner3 and the appearances found to be basically similar to those of rheumatoid arthritis. Geiler4 and Fassbender5 also gave descriptions, but did not formally examine a series of biopsies.Immunohistochemical methods have previously been applied to the rheumatoid synovium by numerous workers,6-"7 but these methods have not been used in the other arthritides except for occasional cases.9 17 18We have obtained histological sections of synovial membrane from ankylosing spondylitics and wish to report our findings using routine staining techniques and the peroxidase-antiperoxidase (PAP) immunohistochemical method for the detection of immunoglobulin subclasses in plasma cells present in the synovial membrane. A series of rheumatoid synovial membranes have been examined for comparison.The PAP method enabled accurate identification of the site of immunoglobulin, which is not possible with the fluorescent methods used by previous workers. Materials and methodsA series of 14 samples of synovial tissue from ankylosing spondylitis was obtained from the London Hospital or kindly sent by colleagues at other centres. These samples were derived from synovial biopsies at the time of synovectomy or joint replacement. Synovial membrane has also been studied from 14 cases of

HospitalSUMMARY Immunohistochemical procedures were used to analyse the subpopulations of mononuclear cells in muscle biopsies from 24 patients with polymyositis. The character of the cellular infiltrate was similar at the perivascular, perimysial, and endomysial sites, with cytotoxic-suppressor T lymphocytes (T8 +) and macrophages being the dominant elements. Helper T lymphocytes (T4+) and B lymphocytes were present in smaller numbers. A control series of 17 muscle biopsies from normal subjects and patients with non-inflammatory myopathies and neurogenic conditions was also studied: the numbers of mononuclear cells present were much smaller than in polymyositis, but the ratio of T4:T8 lymphocytes was similar to that found in biopsies affected by polymyositis. We conclude that both cytotoxic-suppressor T lymphocytes and macrophages are important in the pathogenesis of inflammatory myopathy. We also used the same techniques to investigate the T cell subsets in various non-inflammatory myopathies. We also investigated any possible correlation between serum creatine kinase concentrations and the intensity of inflammatory infiltrate at different sites in the muscle biopsy. Material and methodsCases of polymyositis during the years [1978][1979][1980][1981][1982][1983][1984] were taken from the surgical pathology files of the London Hospital, and 24 cases fulfilled the diagnostic criteria of Bohan and Peter.' There were five cases of dermatomyositis, and two cases of polymyositis associated with neoplasia (one thymoma and one gastric carcinoma).Seventeen biopsies from patients with noninflammatory myopathies were chosen at random, including seven cases that showed no histological abnormality. Table 1 summarises the patient details. All the biopsies were originally reported histologically by one of us (MS).The biopsy specimens had been stored at -160°C in liquid nitrogen; cryostat sections were cut at 5 microns and left overnight at room temperature. The sections were fixed in acetone for 20 minutes at room temperature, transferred to Tris buffered saline (TBS)

No abstract

The localisation of fibronectin in the synovial membrane of rheumatoid arthritis (RA) and other chronic inflammatory joint diseases has been studied using a peroxidase-antiperoxidase immunohistochemical method. Synovia were studied from seven cases of seropositive RA three cases of seronegative RA, six cases of ankylosing spondylitis, four cases of Reiter's syndrome and five of psoriatic arthritis. Six were small biopsies and the remaining tissues were obtained at open surgery for orthopaedic procedures or biopsies. Fibronectin was demonstrated in all of the synovia examined and was present in intimal cells, synovial giant cells, the walls of small blood vessels, basement membrane of larger vessels and deposits of fibrin. No difference in this distribution of fibronectin was found in seropositive and seronegative RA, ankylosing spondylitis, Reiter's syndrome or psoriatic arthritis, neither was there any difference in the amount of fibronectin at various sites.

SUMMARYThe results of immunofluorescence studies in nine cases of rheumatoid arthritis and nine cases of ankylosing spondylitis are presented. Antiglobulin activity was sought with fluorescein labelled, heat aggregated human IgG and rabbit immunoglobulin (indirect immunofluorescence). A similar degree of increased staining for rheumatoid factor activity was seen after pepsin digestion in all cases of rheumatoid arthritis and three of the ankylosing spondylitic cases. This staining for rheumatoid factor could also be shown in control tissue after pepsin digestion, suggesting that the enhanced staining was an artefact.

Correspondence 991Vimentin as a tumour marker Sir: In the paper by Leader et al. (1987) evaluating vimentin, in the January issue of Histopathology, we were interested to read of a breast carcinoma that gave positive staining with vimentin.In May 1985, a 52-year-old lady had a mastectomy for a typical poorly differentiated infiltrating duct carcinoma of the breast. In November 1986 she developed a tumour in a lung. The radiological and bronchoscopic appearances favoured a primary lung neoplasm. For our own interest a small battery of immunocytochemical stains was done on several bronchoscopic biopsies containing carcinoma; the tumour cells were very strongly vimentin-positive.Sections of the previous breast carcinoma were then subjected to the same immunocytochemical stains and gave identical results. The tumour too proved to be very strikingly vimentin-positive (Dako vimentin). Six other infiltrating duct carcinomas of the breast, stained at the same time, were vimentin-negative. This unusual staining of the breast carcinoma provided strong evidence that the lung lesion was a metastasis. The patient has been started on Tamoxifen.