Comprehensive, up-to-date information on HIV/AIDS treatment, prevention, and policy from the University of California San Francisco

A brief summary of the performance of fourth-generation assays, as reported in the literature(1-19), follows. A comparison of four of the commercial fourth-generation assays versus four third-generation antibody ELISAs using seroconversion panels reported that all four of the fourth-generation assays detected infection in fewer days than all the third-generation antibody ELISAs, thus supporting that these assays show promise in reducing the window period and detecting early infection. The VIDAS HIV DUO (bioMerieux), an earlier version of the VIDAS HIV DUO Ultra, demonstrated the highest sensitivity by detecting infection in the shortest number of days. Additional promising performances by fourth-generation ELISAs have been reported. These assays have demonstrated sensitivities ranging from 99.5 - 100%. Several studies with large sample sizes have reported sensitivities of 98 - 99.87% by the Vironostika HIV Uni-Form II Ag/AB, the Enzymun_-Test HIV Combi, AxSYM_ HIV Ag/AB, and the Enzygnost HIV Integral. Using seroconversion panels, several evaluations have calculated a reduction in the diagnostic window by a minimum of four days and some have even reported a reduction of up to nine days using the fourth-generation assays. The specificities have ranged from 98 - 100%, excluding one study that reported a specificity of only 88.2%. This latter study tested the performance of the HIV DUO with a sample size of only 250 potentially cross-reactive samples such as specimens from pregnant women, patients with autoimmune disorders or rheumatoid factor, patients reactive for IgM antibodies to cytomegalovirus, herpes simplex virus, rubella virus, or toxoplasmosis, and patients infected with hepatitis C or Epstein-Barr virus. Three of the four studies have reported sensitivities of 100% when testing samples from individuals with early and established infections. The only study reporting a sensitivity of 91.4% used 17 seroconversion panels, and thus assessed analytical sensitivity, not epidemiologic sensitivity (ability to detect established infection). In addition to the study noted above that reported a specificity of 88.2%, another study using the HIV DUO demonstrated a specificity ranging from 98.2 - 100% when small cohorts of potentially cross-reactive samples were tested. The fourth study reported a specificity of 99.5% in a very large, multi-center evaluation using samples from 29,657 patients. In addition to detecting all 453 HIV-infected individuals, the HIV DUO had the ability to detect early infection in 17 samples containing p24 antigen that were non-reactive by the reference third-generation assays and by Western blot, further supporting the benefits of adding the antigen testing component to a screening test. Interestingly, only 11 of the 17 samples had detectable levels of RNA, and all 17 resolved as acute infections upon follow-up testing. Our laboratory was the first to evaluate a new version, the VIDAS HIV DUO Ultra, with 2,847 samples and to utilize the largest variety of sample populations including hospitalized patients, outpatients, blood donors, HIV-2 samples, HIV-1 Group O samples, HIV-1 Group M clades, patients from three geographical locations, and seroconversion panels.(20) The VIDAS HIV DUO Ultra demonstrated 100% sensitivity and at least a 99.5% specificity overall, with a 99.7 % specificity in low-risk individuals. The analytical sensitivity, as assessed by seroconversion panels and p24 antigen in samples, was equivalent to, or better than, the reference assays used to characterize these panels. This assay, and others, not only detect antibody and antigen simultaneously, but also can discriminate by indicating which of the two analytes is positive.