SUMMARYIL-17 is a T cell cytokine with a complex and important role in the immune system. It has been detected in rheumatoid arthritis (RA) synovial membrane and found to stimulate the production of the proinflammatory cytokines IL-6, IL-8, tumour necrosis factor-alpha (TNF-a ) and granulocytemacrophage colony-stimulating factor (GM-CSF) in vitro. To date, there are few data available on the agents that stimulate IL-17 production. We therefore investigated the in vitro IL-17 response to a variety of mitogens and antigens, and compared the IL-17 response to interferon-gamma (IFN-g), IL-4, IL-10 and TNF-a . In this study we used a type-0 antigen, tetanus toxoid (TT), a type-1 antigen, PPD from Mycobacterium tuberculosis, a potential type-2 rye grass (RG) antigen (Lol I) and an autoantigen SS.B (La), to stimulate PBMC from healthy controls. Cytokine mRNA was measured using semiquantitative reverse transcriptase-polymerase chain reaction and cytokine protein measured using specific ELISA techniques, while the frequency of IL-17-producing T cells was determined by flow cytometry. The mitogens concanavalin A, phytohaemagglutinin and phorbol myristate acetate/ ionomycin induced a significant increase in IL-17, with the highest levels being produced by anti-CD3/anti-CD28 stimulation. The antigens TT and PPD significantly increased IL-17 mRNA expression over time, but failed to have such an effect at the protein level. IL-17 protein was also detectable in both antigen-specific (TT, SS.B) and non-specific T cell clones, but at levels lower than IFN-g . IL-17 production did not correlate with either the type-1 cytokine IFN-g or TNF-a or the type-2 cytokine IL-4 or IL-10 at either the mRNA or protein level.

Antineutrophil cytoplasmic antibodies (ANCA) are well described in Wegener's granulomatosis and some forms of vasculitis. They have also been described in patients with arthritis, but the specificity of these ANCA and their relationship to the presence of vasculitis, antinuclear antibodies (ANA) and granulocyte-specific ANA (GS-ANA), and to disease activity are uncertain. We studied 101 patients with forms of inflammatory arthritis and detected four cytoplasmic ANCA, eight perinuclear ANCA and 16 atypical ANCA. There was no association between the presence of ANCA and ANA or rheumatoid factor. No anti-PR3 antibodies were found and no strong anti-myeloperoxidase antibodies were detected. Four GS-ANA were detected and were distinct from ANCA. There was no association between rheumatoid arthritis disease activity or disability and ANCA status. ANCA did not predict vasculitis over a 3 yr follow-up. These ANCA appear to be epiphenomena. Their importance lies in their potential to mislead physicians towards a misdiagnosis of vasculitis.