Objective: Antibodies directed against components of neutrophil cytoplasm have been detected in various systemic vasculitides and especially in Wegener's granulomatosis. In polyarteritis nodosa (PAN) and Churg-Strauss syndrome, few data are available and correlation between clinical manifestations and antineutrophil cytoplasm antibodies (ANCA) has not been established. Therefore, we tested, before treatment of vasculitis, 62 consecutive patients suffering from PAN with hepatitis B virus (HBV) markers, PAN of unknown etiology or Churg-Strauss syndrome.
Methods: Only patients with PAN and Churg-Strauss syndrome were included in the study. The diseases were histologically and/or angiographically proven. Every patient's serum was tested by an indirect immunofluorescence assay (IFA) and, in 37 cases, by an enzyme linked immunosorbent assay (ELISA).
Results: ANCA detected by IFA were observed in 10.7% of the patients with PAN with HBV markers, in 27.3% of the patients with PAN without HBV markers and in 66.7% of the patients with Churg-Strauss syndrome. When ELISA was performed, 11.1% of the patients with PAN associated with HBV infection, 20% of the patients with PAN without HBV markers and 55.6% of the patients with Churg-Strauss syndrome were positive. ANCA were positively correlated with asthma and purpura and negatively correlated with HBV markers.
Conclusion: Regardless of the technique used, Churg-Strauss syndrome was associated with ANCA in about 60% of the cases while, in PAN of unknown etiology, ANCA were found in about 25% of cases. In contrast, IFA and ELISA only detected ANCA in a limited number of cases of PAN related to HBV infection. ELISA positivity in patients with PAN and Churg-Strauss syndrome was usually associated with antimyeloperoxidase antibodies. In our cases of PAN, ANCA and purpura were significantly correlated, suggesting that, in these cases, small vessels are involved and therefore macroscopic and microscopic PAN coexist. Thus it seems that ANCA are essentially present in the cases of small vessel vasculitis, as has been described, and are not a marker of pure macroscopic PAN, at least at our present level of understanding of these antibodies.