EVALUATION OF 14 COMMERCIAL HIV-1 HIV-2 ANTIBODY-ASSAYS USING SERUM PANELS OF DIFFERENT GEOGRAPHICAL ORIGIN AND CLINICAL STAGE INCLUDING A UNIQUE SEROCONVERSION PANEL/

The performance of 14 commercially available HIV-1/2 antibody assays w ere compared using well-characterized serum panels containing in total 1500-1800 sera. The panels included consecutive HIV-negative blood do nor sera from Sweden, unselected blood donor and patient sera from Tan zania and unselected sera from outpatient clinics in Guinea-Bissau. Fu rthermore selected HIV-1 antibody positive sera from Sweden and Tanzan ia and HIV-2 antibody positive sera from Guinea-Bissau were included i n the panels. The HIV-1 antibody positive sera were from individuals a t various stages of HIV infection, from primary infection, to asymptom atic phase and late stage disease. 12 of the 14 assays identified corr ectly all HIV-1 and HIV-2 antibody positive sera. One Tanzanian HIV-1 antibody positive sample with complete banding pattern on Western blot was not detected by two of the ELISAs employing synthetic peptides. T here were small differences in sensitivity between the assays when use d for analysis of seroconversion panels. The most sensitive assay, Abb ott IMx HIV-1/HIV-2 III Plus detected antibodies in all nine samples c ollected from four individuals during the first week after onset of sy mptoms of primary HIV-1 infection. Most of the assays became reactive during the second week after onset of symptoms and the least sensitive assays were reactive from the third week. The assays showed a high sp ecificity ranging from 99.2 to 100% when used for analysis of Swedish blood donor sera, while most of the assays showed a significantly lowe r specificity, 91.9-99.6%, when used for testing African specimens. (C ) 1998 Elsevier Science B.V. All rights reserved.