CROI 2018 Abstract eBook

Abstract eBook

Poster Abstracts

Conclusion: Our data suggests that the CAP/CTM test is more useful in detecting HIV infection than plasma viral load in individuals on HAART, although verification of HIV infection status post initiation of HAART in acute and primary infection may be difficult. 574 QUANTIFICATION OF UNDETECTABLE PLASMA HIV RNA (<20 COPIES/ML) WITH SINGLE-COPY ASSAY Nicolas A. Margot 1 , Dianna L. Koontz 2 , Scott McCallister 1 , John W. Mellors 2 , Christian Callebaut 1 1 Gilead Sciences, Inc, Foster City, CA, USA, 2 University of Pittsburgh, Pittsburgh, PA, USA Background: Plasma HIV RNA or viral load (VL) is measured in clinical practice and trials of antiretrovirals using FDA-cleared assays such as COBAS TaqMan HIV-1 Assay v2.0. The TaqMan assay provides quantification of viremia at or above 20 copies/mL, but lower values are reported as “<20” or “Target Not Detected” (TND). Current “kick & kill” HIV eradication strategies may require more sensitive assays to detect changes in low-level persistent viremia. Here, the novel integrase single-copy assay (iSCA) (Cillo JClinMicro 2013) was evaluated for measurement of low-level persistent viremia in a large number of clinical trial samples. Methods: Plasma samples were fromWeek 24 visits of a Phase 2 study in previously treatment naïve HIV-1-infected patients treated with a single tablet regimen containing an integrase inhibitor. HIV-1 RNA levels were assessed at a central laboratory using HIV-1 TaqMan 2.0 Assay (Roche Diagnostics, Indianapolis, IN). The iSCA assay was performed in a blinded fashion on matched samples (University of Pittsburgh) and results from the 2 assays were compared. Results: Paired TaqMan/iSCA data were obtained for 151 HIV-infected adults. All participants were on treatment and virologically suppressed (mean 110 days) at week 24. Most samples (117/151, 77%) had non-quantifiable TaqMan result, either <20 copies/mL (n=44) or TND (n=73). Quantification was achieved with iSCA for all 117 samples (mean VL 2.6 copies/mL for 73 samples with TND; mean VL 8.2 copies/mL for 44 samples with <20 copies/mL). Zero copy control samples included with each assay run were all negative for HIV RNA (<1 copy). For samples quantified with both assays (n=34), iSCA values were slightly lower than TaqMan (mean VL of 29.5 copies/mL compared to 61.4 copies/mL, respectively). Conclusion: In this large sample collection from virologically suppressed HIV-infected adults, use of iSCA led to quantification of low-level viremia below the limit of detection of the TaqMan assay in 77% (117/151) of previously non- quantifiable plasma samples. This dataset emphasizes the value of the iSCA over classical HIV VL assays for measurement of low-level viremia and its potential for use in HIV cure studies to assess whether experimental interventions alter viremia. 575 REPLICATE APTIMA VL TESTING DETECTS RESIDUAL VIREMIA IN MOST ART-TREATED ADULTS Sonia Bakkour 1 , Sheila M. Keating 1 , Xutao Deng 1 , Mars Stone 1 , AndrewWorlock 2 , Steven G. Deeks 3 , Peter Bacchetti 3 , Melanie Dimapasoc 1 , Joseph Lau 1 , Leilani Montalvo 1 , Scott Hauenstein 2 , Douglas D. Richman 4 , Michael P. Busch 1 1 Blood Systems Research Institute, San Francisco, CA, USA, 2 Hologic Corporation, Bedford, MA, USA, 3 University of California San Francisco, San Francisco, CA, USA, 4 University of California San Diego, La Jolla, CA, USA Background: The ability to determine if a curative intervention reduces the reservoir or if a latency reversing agent is effective will depend on access to ultrasensitive, high-throughput measurements of residual viremia. Current viral quantification methods are limited by lack of sensitivity or the need for specialized, lengthy processing. Methods: The Aptima HIV-1 Quant Assay provides standard viral load (VL) measurements on a 0.5mL sample, but it also provides a reactive/non-reactive digital readout that may be reasonably sensitive even when only a single copy is present in the 0.5 mL sample. Readouts on multiple replicates (reps) on the system’s automated platform using the standard sample volume can provide ultrasensitive estimates of copies/mL (cp/mL) via Poisson analysis. An analytical panel comprised of 5 serial dilutions each of 4 HIV+ window period donation samples (2 clade B and 2 clade C) was blindly tested using the Aptima Assay in 45 reps on 25 mL per dilution, and 110 large volume samples from antiretroviral- suppressed RAVEN study participants with consistently negative standard VL assay results were subjected to rep testing.

detectable but not quantifiable by Xpert VL. Most had concordant results at the log 10 2.3 threshold except one had a log 10 VL <2.3 by Roche VL but ≥2.3 by Xpert VL while 9 specimens with a log 10 VL ≥2.3 using Roche VL were <2.3 with Xpert VL. Results were similar at the log 10 3.0 cut-off. Among uninfected specimens there were 8 errors (1.6%) using Xpert VL; 5 had insufficient volume to repeat. Specificity was 491/492 (99.8%; 95% CI 98.87%-99.99%). Conclusion: Xpert VL results were highly correlated and concordant with Roche VL, and the test performed with high specificity. Seroconversion specimens missed by Xpert VL had a Roche VL <100 copies/mL. FDA approval of the simple Xpert VL may allow laboratories that cannot bring on large, complex VL tests to conduct HIV monitoring.

Poster Abstracts

573 CELLULAR HIV-1 NUCLEIC ACID DETECTION IN CASES WITH UNDETECTABLE PLASMA VIRAL LOAD

Linda Jagodzinski 1 , Holly R. Hack 2 , Ying Liu 2 , Jennifer Malia 1 , Mark M. Manak 2 , Nittaya Phanuphak 3 , Mark de Souza 3 , Eugène Kroon 3 , Donn Colby 3 , Nitiya Chomchey 3 , Merlin L. Robb 2 , Nelson L. Michael 1 , Jintanat Ananworanich 2 , Sheila A. Peel 1 1 Walter Reed Army Institute of Research, Silver Spring, MD, USA, 2 US Military HIV Research Program, Silver Spring, MD, USA, 3 SEARCH, Bangkok, Thailand Background: HIV-1 diagnosis based solely on HIV serological tests can result in misclassification of HIV-1 infection status of HIV vaccinated individuals, infants born to HIV infected mothers and individuals that start highly-active antiretroviral therapy (HAART) during early infection. Individuals at-risk for HIV infection on pre-exposure prophylaxis (PrEP) may also demonstrate reduced or absent serological response to infection. The objective of this study was to determine if a HIV total nucleic acid (TNA) test could be used to determine HIV-1 infection status. Methods: The Roche COBAS AmpliPrep/COBAS TaqMan HIV-1 test v2.0 (CAP/ CTM), which detects RNA/DNA, was used to evaluate HIV TNA in peripheral blood mononuclear cells (PBMCs) in infection cases where plasma HIV-1 RNA was not detected. PBMCs from Thai individuals (N=21) who initiated HAART at Fiebig (F) stage I-VI of acute HIV infection (RV254/SEARCH010) were collected at 8 and 60 weeks post treatment and tested by CAP/CTM. In a separate US study, cell pellets from HIV-1 infected individuals on HAART (N=67) and from HIV-1 uninfected individuals (N=92) were also tested. All individuals demonstrated undetectable plasma HIV-1 RNA at the time of sample collection. Results: Five of six individuals who initiated HAART in early acute infection (F I) were undetectable by HIV-1 TNA at 8 and 60 weeks; one individual demonstrated a low threshold value in 1 of 3 replicates. Individuals who initiated treatment at F II (N=6) were detectable by HIV-1 TNA at week 8, but at week 60 were either not detected or approaching the lower limit of detection by TNA. Of the 3 individuals treated at F III only one had detectable TNA at week 8 and none by week 60. Six individuals who started treatment at Fiebig IV - VI had detectable HIV-1 TNA, but the relative values reported decreased two to fifteen- fold from 8 weeks to 60 weeks post HAART initiation. All chronically infected individuals from the US study were HIV-1 TNA positive. As expected, HIV-1 TNA was not detected in uninfected individuals.

CROI 2018 211