CROI 2016 Abstract eBook

Abstract Listing

Poster Abstracts

521 4th Generation Rapid Tests Improve Detection of Acute Infection in MTN-003 (VOICE) Edward Livant 1 ; Amy L. Opest 2 ; CliftonW. Kelly 3 ; Rashika Maharaj 4 ; Natasha Samsunder 5 ; Rosetta Lindiwe Nhlangulela 6 ; Patrick Karugaba 7 ; Jeanne Marrazzo 8 ; Z Michael Chirenje 9 ; Urvi M. Parikh 2 ; for theVOICE StudyTeam 1 Magee-Womens Hosp of the Univ of Pittsburgh Med Cntr, Pittsburgh, PA, USA; 2 Univ of Pittsburgh, Pittsburgh, PA, USA; 3 Fred Hutchinson Cancer Rsr Cntr, Seattle, WA, USA; 4 South African Med Rsr Council, Durban, South Africa; 5 Cntr for the AIDS Prog of Rsr in South Africa, Durban, South Africa; 6 Aurum Inst for Hlth Rsr, Johannesburg, South Africa; 7 Makerere Univ Johns Hopkins Univ Rsr Collab Lab, Kampala, Uganda; 8 Univ of Washington, Seattle, WA, USA; 9 Univ of Zimbabwe, Univ of California San Francisco Collaborative Rsr Prog, Harare, Zimbabwe Background: The greatest risk of antiretroviral resistance from pre-exposure prophylaxis is conferred by use during undetected acute infection. Early, accurate identification of infection in HIV prevention trials is critical for protection of human subjects, data quality and study efficiency. We evaluated pre-seroconversion plasma from the VOICE Study (MTN-003) to determine if 4th generation antigen/antibody (Ag/Ab) rapid tests would have detected HIV infection earlier than the 3rd gen rapids used in the trial. Methods: VOICE evaluated the safety and effectiveness of oral tenofovir disoproxil fumarate (TDF), oral TDF-emtricitabine (FTC) and vaginal tenofovir 1% gel for prevention of HIV infection in 5029 women from 15 clinical sites in South Africa, Uganda and Zimbabwe. Participants were monitored monthly for seroconversion with 1 or 2 3rd gen HIV rapids (Alere Determine™, OraQuick Advance® and/or Trinity Biotech™ Unigold®). Pre-seroconversion plasma collected closest to the date of the 1st positive rapid was tested for HIV-1 RNA (Roche TaqMan) and with Bio-Rad GS HIV Combo Ag/Ab EIA and re-tested with Unigold and OraQuick. The 4th gen FDA-approved Alere Determine™ HIV-1/2 Ag/Ab Combo (FDA- Combo) was compared with the Conformité Européene (CE)-Marked Alere™ (CE-Combo) which is marketed to be optimized for non-B subtypes. We also evaluated Multispot® (MS), Western Blot (WB), and Geenius™ (all Bio-Rad) as confirmatory tests. McNemar’s test was used to calculate statistical significance. Results: Of 231 plasma samples collected 12-91 days (median 55) before detection of seroconversion, 68 had RNA >200 cp/mL, and of those 57 tested negative by both Unigold and OraQuick. Of these 57, 30(53%) were reactive by 4th gen EIA. 16(28%) were reactive on the CE-Combo (1 Ab; 9 Ag; 6 Ag/Ab reactive). MS confirmed only 1 of 16 acute infections while WB (9 indeterminate [ind]) and Geenius (2 HIV-2 ind) confirmed none. 4(7%) samples were reactive by FDA-Combo (2 Ab; 2 Ag; 0 Ag/Ab reactive) of which none were confirmed by MS, WB (3 ind) or Geenius (1 HIV-2 ind). CE-Combo detected significantly more infections than FDA-Combo (p<0.002). Conclusions: In VOICE, 28% of infections missed by current 3rd gen rapid tests would have been identified with the use of CE-marked Alere HIV Combo. Geenius, MS and WB were all insensitive (<10%) in confirming infections detected by 4th gen assays. An improved diagnostic algorithm that includes 4th gen rapids with HIV RNA testing will be essential for efficiently identifying seroconverters. 522 Usefulness of Rapid Tests for HIV Diagnosis in the ANRS IPERGAY PrEP Trial Constance Delaugerre 1 ; Isabelle Charreau 2 ; Nadia Mahjoub 3 ; Eric Cua 4 ; Armelle Pasquet 5 ; Nolween Hall 6 ; Marie Laure Chaix 7 ; Guillemette Antoni 2 ; Jean-Michel Molina 8 ; for the IPERGAY Study Group 1 Hosp Saint Louis et Univ Paris Diderot, Paris, France; 2 INSERM SC10-US19, Villejuif, France; 3 Hosp Saint-Louis, Paris, France; 4 Hosp de l’Archet, Paris, France; 5 DRON Hosp, Tourcoing, France; 6 CISIH, Nantes, France; 7 INSERM U941, Univ Paris Diderot, Paris, France; 8 Hopital Saint-Louis, Paris, France Background: Pre-exposure prophylaxis (PrEP) implementation will lead to more frequent HIV testing. Rapid tests are likely to be used especially in resource limited countries, and our aimwas to assess their usefulness in the setting of the ANRS IPERGAY PrEP trial. Methods: In the ANRS IPERGAY trial, a 4 th generation (4thG) antigen/antibody immunoassay (ARCHITECT HIV Ag/Ab Combo®, Abbott) and/or plasma Viral Load (pVL) (AmpliPrep/ COBAS® TaqMan® HIV-1 Test, v2.0) were used for HIV diagnosis at screening and during follow-up. We used stored sera to perform the following tests at the date of diagnosis: pVL, rapid test (VIKIA® HIV1/2, Biomérieux) and HIV-1 western blot (WB, GS HIV-1 Western Blot®, Biorad). We defined 3 stages of HIV infection according to the number of WB antibodies (Ab): chronic (> 7 Ab), recent (1 < Ab < 7) and acute (0 Ab). HIV-1 subtype was determined after phylogenetic analysis of the RT sequence. Results: Overall, 31 HIV-1-infected patients were diagnosed during the ANRS IPERGAY trial. Stored sera were available for 27 cases of HIV infection. Overall, the 4thG was positive in 25 (93%) (median index 52), rapid tests in 15 (56%) and positive WB (> 1 Ab) in 16 (59%) patients. Median pVL was 5.16 log 10 copies/ml. HIV-1 subtype B was identified in 16/25 (64%) cases. Sensitivity of rapid tests was 100% (95%CI: 54-100) for chronic infection, 70% (95%CI: 35-93) for recent infection and only 18% (95%CI: 3-52) for acute infection (p < 0.002). Of note, among the 12 positive sera with the 4thG assay but with negative rapid tests, 8 were retested at the follow-up visit (median: 5 days) and 4/8 became positive with rapid tests. Conclusions: Rapid tests were able to adequately detect chronic infection at screening but largely failed to diagnose acute HIV infection in people at high risk enrolled in a PrEP trial. 4 th G assays should be used in settings where PrEP is implemented to avoid missing acute HIV infection, with the risk of selecting drug resistance and of ongoing HIV transmission.

Poster Abstracts

523 Use of an Automated Assay to Identify HIV Infections in a Population Survey in Africa

Estelle Piwowar-Manning 1 ; ShaunaWolf 1 ; Kelsey Donohue 1 ; Anelet James 2 ; Barry Kosloff 3 ; AliciaYoung 4 ; Gert U. van Zyl 5 ; Helen Ayles 6 ; Richard Hayes 6 ; Susan H. Eshleman 1 1 Johns Hopkins Univ Sch of Med, Baltimore, MD, USA; 2 Desmond Tutu TB Cntr, Stellenbosch Univ, Cape Town, South Africa; 3 ZAMBART, Lusaka, Zambia; 4 SCHARP, Seattle, WA, USA; 5 Stellenbosch Univ and NHLS Tygerberg Business Unit, Tygerberg, South Africa; 6 London Sch of Hygiene & Trop Med, London, UK Background: HPTN 071 (PopART) is a community-randomized HIV prevention trial in Zambia and South Africa (SA) that includes a cohort of >30,000 adults. Initial HIV testing is performed in-country using a 4 th generation test (Abbott ARCHITECT HIV Ag/Ab Combo Test, one run per sample); special procedures for sample and data management are used in-country to optimize data quality. HIV status is confirmed in the United States (US). We evaluated performance of the Abbott test as an HIV screening test. Methods: Samples that were initially reactive with the Abbott test (Abbott POS) were tested in the US with a second 4 th generation test (BioRad Combo); 10% of samples that were initially non-reactive with the Abbott test (Abbott NEG) were retested in the US using the Abbott test. Samples with discordant results (in-country and US results did not agree) were tested with the BioRad Multispot test, the BioRad Geenius test, and an HIV RNA test (cutoff of 40 or 400 copies/mL). Results: 10,390 samples were tested in the US (5,563 from Zambia, 4,827 from SA; 7,695 Abbott POS; 2,695 Abbott NEG). 56 samples had discordant results. For the Abbott POS samples: 52/7,695 (0.62%) were non-reactive with the BioRad Combo test (25 from Zambia, 27 from SA). Further testing revealed that 4 were HIV-positive (the original Abbott test result was correct), 48 were HIV-negative (the original Abbott test results were false positive results; 1 case involved a sample mixup). For the Abbott NEG samples: 4/2,695 (0.15%) were reactive when retested with the same assay. One had a signal/cutoff (S/C) ratio of 1.25 and was non-reactive on a second repeat test. The other two had S/C ratios of 4.77 and 10.38; both were reactive with the BioRad Combo test, but had undetectable HIV RNA and non-reactive Multispot and Geenius tests. All 4 were classified as HIV-negative. Overall, 48/7,695 Abbott POS samples had an initial false positive test result and 0/2,695 Abbott NEG samples had a false negative result (sensitivity=100%; specificity=98.3%).

203

CROI 2016