SENSITIVITY OF 2 ENZYME-LINKED-IMMUNOSORBENT-ASSAY TESTS IN RELATION TO WESTERN-BLOT IN DETECTING HUMAN T-CELL LYMPHOTROPIC VIRUS TYPE-I AND TYPE-II INFECTION AMONG HIV-1-INFECTED PATIENTS FROM SAO-PAULO, BRAZIL

We investigated the presence of human T-cell lymphotropic virus types I and II (HTLV-I and HTLV-II) infections, first searching for specific antibodies in 553 serum samples obtained from HIV-l-infected patients from Sao Paulo, Brazil. Sera were screened using two enzyme-linked im munosorbent assays (ELISAs): the ELISA-EM (ELISA HTLV-I/II, EMBRABIO, EX), which contains HTLV-I and HTLV-II lysates, and the ELISA-DB [ELIS A HTLV-I/II, Diagnostic Biotechnology (DB), Singapore], which contains HTLV-I lysate, and HTLV-I and HTLV-II recombinant env proteins (MTA-1 and K55, respectively). Serum samples showing two positive and/or bor derline results were confirmed by Western blot (WE 2.3, DB), which dis criminates HTLV-I from HTLV-II. WE analyses disclosed 22 cases (4.0%) of HTLV-I and 34 (6.1%) of HTLV-II seroreactivity; 24 sera had indeter minate antibody profile (4.3%) and 2 specimens showed reactivity to bo th MTA-1 and K55 env proteins. Using stringent WB criteria and analyzi ng the population according to risk factors, the prevalence rates of H TLV-I and HTLV-II infections were 11.2% and 16.8% in IV drug users, 3. 4% and 5.5% in heterosexual individuals, and 1.4% and 2.2% in homosexu al/bisexual men, respectively. A comparison of ELISA and WE results di sclosed that both ELISAs were highly sensitive in detecting HTLV-I ant ibodies, whereas the ELISA-DB showed 82% sensitivity and the ELISA-Ehl 100% sensitivity in detecting HTLV-II antibodies. PCR analyses conduc ted on 37 representative cells samples confirmed the presence of HTLV proviral DNA in the majority of concordant serological cases, except i n one, which was HTLV-I infected and seroreacted with K55 protein of H TLV-II. Indeed, after PCR, one case of HTLV-I infection and HTLV-II co infection, and 30% of WB-seroindeterminate or inconclusive cases infec ted with HTLV-II could be detected. Our data stress high prevalences o f both HTLV-I and HTLV-II infections in HIV-1 coinfected i.v. drug use rs from Sao Paulo, and suggests that ELISA kits containing only K55 pr otein as the HTLV-II-specific antigen, may not have the appropriate se nsitivity for the detection of HTLV-II infection in this geographic re gion, pointing out the need of improved screening tests to be used in Brazil. (C) 1998 Elsevier Science Inc.