ANTINUCLEAR, ANTINEUTROPHIL CYTOPLASMIC AND ANTIGLOMERULAR BASEMENT-MEMBRANE ANTIBODIES IN HIV-INFECTED INDIVIDUALS

Many autoantibodies have been described in HIV-infected individuals. W e have examined the incidence, associations and prognostic significanc e of anti-nuclear antibodies (ANA), anti-neutrophil cytoplasmic antibo dies (ANCA) and anti-glomerular basement membrane (GBM) antibodies in individuals with HIV infections.One hundred and five patients, with as ymptomatic infections (n=37), AIDS-related complex (n=32) or AIDS (n=3 6) were studied. Plasma from 24 of these (23%) were positive for ANA: most demonstrated speckled fluorescence (n=21) and were of low titre ( 1+ in 18). ANCA were demonstrated by IIF in 18 individuals (17%) and a ll fluorescent patterns were seen; 6 of these plasma were also positiv e in the ELISAs for antibodies to proteinase 3, myeloperoxidase or ela stase. Thirteen plasma were positive for ANCA in the neutrophil cytopl asm ELISA; 10 of these were also positive in the specific ELISAs. A to tal of 30 plasma bound to proteinase 3, myeloperoxidase or elastase in specific ELISAs, in 6 cases with 2 specificities. Finally, 18 plasma (17%) contained anti-GBM antibodies by ELISA, but none of 4 plasma tes ted in inhibition assays was specific. ANA, ANCA and anti-GBM antibodi es were not uncommon in HIV-infected individuals but the presence of t hese antibodies was not associated with the clinical manifestations of the corresponding autoimmune diseases. In addition, there was no corr elation between the demonstration of these antibodies and the immunolo gical status of the individual (apart from a correlation between CD4 c ounts less than 400/mu l with anti-GBM antibodies), the presence of an opportunistic infection, the development of malignancy or reduced sur vival. Some of these antibodies may arise from polyclonal activation, or be due to ''sticky'' serum since we have shown that the presence of anti-GBM antibodies correlated with the demonstration of ANCA by ELIS A. These antibodies are not more common in hypergammaglobulinemic plas ma but some may be due to heat-treatment of the plasma. The clinician caring for HIV-infected individuals needs to be aware of these ''false -positive'' antibody results.