CROI 2020 Abstract eBook

Abstract eBook

Poster Abstracts

Results: OraQuick had a sensitivity of 92.6%, (95% CI: 75.7 - 99.1) on pre and post-embalmed samples when compared to the gold standard. The specificity was 97.1% (95% CI: 91.9 - 99.4) and 95.2%, (95% CI: 89.2 - 98.4) pre and post-embalming respectively (Table1). The pre-embalming PPV of OraQuick® was 89.3% (95% CI: 71.8 - 97.7) and 83.3%, 95% (95% CI: 65.3 - 94.4) post- embalming. FDR was lower on pre-embalming compared to post embalming at 10.7% (95% CI 2.3 – 28.2) and 16.7%, (95% CI: 5.6 - 34.7) respectively. Only 2/27 (7%) were false negative. FOR pre-embalming (1.92%) and post- embalming (1.96%) were similar. Conclusion: OraQuick® was found to more specific than sensitive on oral specimens from cadavers. Similar performance has been reported among living subjects. It is a convenient less invasive screening test for surveillance of HIV among cadavers within a mortuary setting.

970 AUTOMATED HIGH-THROUGHPUT QUANTIFICATION OF LOW-LEVEL HIV-1 PLASMA VIREMIA Jana L. Jacobs 1 , Melissa A.Tosiano 1 , Dianna L. Koontz 1 , AndrewWorlock 2 , Karen Harrington 2 , Sonia Bakkour 3 , Michael P. Busch 3 , John W. Mellors 1 1 University of Pittsburgh, Pittsburgh, PA, USA, 2 Hologic Corporation, Bedford, MA, USA, 3 Vitalant Research Institute, San Francisco, CA, USA Background: Low-level plasma HIV-1 viremia persists in the majority of HIV-1 positive individuals despite long-term clinically-effective ART. Clearance of HIV-1 viremia remains a critical goal towards an HIV cure, but complex and low-throughput single copy assays (SCA) limit the capacity to monitor the effects of interventions on persistent viremia. Here we report the evaluation of two high-throughput methods on the Hologic Panther platform to automate quantitation of low-level viremia in comparison with a SCA targeting integrase (iSCA2.0; Tosiano, et al. J Clin Micro 2019). Methods: The assay methods performed on the Hologic Panther platform were: 1) testing of nine 0.5mL replicates (Panther 9x) with estimation of HIV-1 RNA concentration using statistical inference based on binary outcome; and 2) concentration of 5mL plasma to one 0.7 mL replicate by centrifugation (Panther spun). Plasma HIV-1 RNA standards (20, 5, 2.5, 1.25, 0.625, and 0 copies/ml) from the Quality Assurance (VQA) at Rush University were tested in 5 independent runs of 5 replicates. Both Panther methods were compared to the manual iSCA 2.0. Mean, standard deviation and percent positive assays were calculated for each run and the 95% LOD was assessed using maximum likelihood estimation. Results: Assay results are summarized in the Table. The 95% LODs (95% CI) were 2.3 (1.6, 3), 3.0 (2.1, 3.8), 3.9 (2.8, 5) for iSCA2.0, Panther 9x and Panther spun, respectively, indicating that iSCA2.0 was most sensitive but that Panther 9x was only marginally less sensitive. Panther spun had reduced sensitivity compared to the other methods. Each assay had 100% specificity across 25 replicates of 0 copies/ml. The weekly estimated throughput for the Panther 9x method is 5-10 times that of iSCA 2.0. Conclusion: Although the manual single copy assay targeting HIV-1 integrase (iSCA 2.0) has the lowest 95% limit of detection for plasma HIV-1 RNA, multiple replicate testing (9x) on the Hologic Panther platform has similar sensitivity and could be used as a screening tool for higher throughput monitoring in clinical trials of interventions aimed at clearing persistent viremia towards a functional cure of HIV-1 infection.

Poster Abstracts

969 DRIED BLOOD SPOTS PROVIDE SIMPLIFIED ACCURATE MEASUREMENT OF HIV VIRAL LOAD Torin Schaafsma 1 , Katherine Thomas 1 , Heidi Van Rooyen 2 , Maryam Shahmanesh 3 , Jared Baeten 1 , Connie L. Celum 1 , Ruanne V. Barnabas 1 , for the DO ART Study Team 1 University of Washington, Seattle, WA, USA, 2 Human Sciences Research Council, Pretoria, South Africa, 3 Africa Health Research Institute, Mtubatuba, South Africa Background: HIV viral load (VL) is a robust measure of adherence and treatment efficacy when monitoring antiretroviral therapy (ART), facilitating timely switching to second line regimens. However, collection of plasma for VL requires phlebotomy, controlled transport conditions, and is costly, thereby limiting its use in community-based settings. The use of dried blood spot (DBS) cards of finger-prick blood transported at ambient temperature to a central laboratory for VL testing would simplify monitoring, but requires validation against the gold standard method of plasma VL. Methods: In a randomized study of community-based delivery of ART in KwaZulu-Natal, South Africa (the DO ART Study) persons living with HIV provided concurrent EDTA plasma and DBS specimens. Specimens were transported 100 - 250 km to Global Labs (Durban), which used the bioMérieux NucliSENS EasyQ HIV-1 assay to measure plasma VL (limit of quantification [LOQ]: 20 copies/mL) and a modified version of the same assay for DBS whole blood specimens (LOQ: 100 copies/mL). Values below the LOQ are represented as 1 (0 when log transformed). We compared 856 pairs of results from 678 DO ART participants using intra-class correlation, two by two tables, scatterplots, and mean-difference plots in R. Results: There was high correlation between log-transformed DBS and plasma VL results, with an intra-class correlation coefficient of 0.93 (95% CI: 0.93 - 0.94). DBS viral loads tended to be lower overall with a mean difference between DBS and plasma results of -0.17 log 10 copies/mL (95% of the differences were from -1.42 to 1.09 log 10 copies/mL). Using the WHO threshold for viral suppression of 1000 copies/mL, in a population with 13% virological failure, the sensitivity and specificity of DBS were 91% and 99%, respectively, with positive and negative predictive values of 93% and 99% for detecting treatment failures. In this population, for every 100 persons undergoing viral load monitoring via DBS, an estimated one treatment failure would be miscategorized. There were no clinically meaningful differences by sex or CD4 count. Conclusion: DBS provides a highly accurate result compared to plasma VL and could be used in a simplified approach to population-based ART monitoring in resource-limited settings. Self-collection of DBS cards should be evaluated as a means of further simplifying specimen collection and ART monitoring.

CROI 2020 364