Performance characteristics of the COBAS AmpliScreen HIV-1 test, version 1.5, an assay designed for screening plasma mini-pools

BACKGROUND: The COBAS AmpliScreen HIV-1 test, version 1.5 (v1.5) (Roche Mol ecular Systems), is designed for screening pools composed of samples from 2 4 individual units of blood or plasma. A specimen-processing procedure (Mul tiprep) simultaneously concentrates and extracts HIV-1, HCV, and HBV partic les from plasma and incorporates an HIV-1 internal control (IC) RNA. Proces sed samples are amplified by RT-PCR using HIV-1-specific primers and detect ed by hybridization of the amplified products to HIV-1- and IC-specific oli gonucleotide probes. STUDY DESIGN AND METHODS: Plasma samples containing known quantities of HIV -1 were used to evaluate analytical sensitivity and precision and to valida te a pool testing algorithm. Analytical specificity was evaluated by adding various viruses and bacteria to HIV-1-negative plasma. Seroconversion pane ls were tested to estimate the window-period reduction achieved by RNA test ing. RESULTS: The analytical sensitivity of the test (concentration that yields greater than or equal to 95% positive results in a set of replicate tests) was 25 copies of HIV-1 RNA per mt of pooled plasma. Representative strains from all HIV-1 group M subtypes were reproducibly detected (> 95% positive results) at concentrations of 20 to 200 viral particles per mt. The test di d not cross-react with a set of 31 viral and 5 bacterial isolates, and it y ielded negative results on a panel of 500 blood samples from HIV-1-seronega tive donors. Plasma samples containing abnormally high levels of Hb, albumi n, triglycerides, or bilirubin did not interfere with the test. HIV-1 RNA w as detected 2 to 14 days before HIV-1 antibody and 0 to 28 days before p24 antigen. The test specifically detected pools containing a single positive unit with 2400 HIV-1 RNA copies per mt and correctly identified the positiv e unit. CONCLUSION: The COBAS AmpliScreen HIV-1 test, v1.5, has sufficient sensitiv ity to detect a single infected unit containing 600 copies of HIV-1 per mt in a pool with 23 uninfected units and should reduce the window period betw een infection and seroconversion by at least 2 to 14 days.