Comparison of Roche TaqMan v.2.0, Abbott RealTime and Modified Abbott Residual Viremia Assays in Testing of Routine Samples for Viremia Monitoring in HIV Care

BACKGROUND: Routine HIV RNA clinical monitoring frequently reports virermia (VL) as detectable, quantifiable <50 or blips >50 c/mL. To assess whether the results represent true increases or just assay variability, we designed comparisons of the Roche COBAS AmpliPrep / COBAS TaqMan HIV-1 Test version 2.0 (TM) and the Abbott RealTime HIV-1 assay (RT) to a residual viremia (RV) assay with a limit of detection of 3 c/mL, the Modified Abbott Residual Viremia assay (MARV). METHODS: HIV suppressed patients on stable therapy presenting for routine visits at the HIV clinic in Henry Ford Hospital (HFH) were prospectively enrolled. Blood samples and questionnaires were obtained at each visit. Paired plasma samples were processed and analyzed per manufacturer instructions; HFH clinical laboratory processed the samples using the TM assay and the McKinnon Research Laboratory performed the FDA approved RT assay and RV testing using the MARV assay. Parametric and non-parametric analyses were conducted as indicated. RESULTS: 124 HIV patients are reported with a mean age of 51, mostly male (84%) and African American (64%). Mean CD4 cell counts were 661 cells/mm(3). TM assay results for 196 plasma samples were not detectable (ND) 119 (61%), detectable (DT) 61 (31%) and quantifiable viremia (QV) >20 c/mL 16 (8%). 5 patients had QV over 50 c/mL. 60 tested patients (49.5%) had at least one sample either DT or QV. TM results were not correlated to CD4 cell counts, antiretrovirals, medical conditions or reported adherence (avg. 97%). 174 RT and 187 MARV paired tests were ND on 142 (81%) / 121 (65%), DT on 34 (17%) RT, and quantifiable in 0% / 34 (33%) of the samples respectively. Higher VL detection by the TM was shown compared with the RT (P = 0.005) and MARV (P < 0.001) assays. Mean QV by the MARV was 5.0 vs.. 55 c/mL on the TM assay (P < 0.001). VL for all assays trends correlated when compared for ND, DT and >20 copies/mL (P < 0.001). Bland-Altman plots show higher VL detected by TM as compared with MARV (P < 0.001) and good correlation between RT and MARV assays (P = 0.6). CONCLUSION: HIV viremia is more frequently reported by TM assay as compared with the RT assay and was significantly higher than VL detected by the MARV assay. Over detection of VL by the TM assay may impact clinical decision making and increase cost of care. DISCLOSURES: J. Mckinnon, Abbott Molecular: Speaker’s Bureau, Grant recipient, Research grant and Speaker honorarium D. Lucic, Abbott Molecular: Employee, Salary