Share
Title
Positive Predictive Value of HIV Serological Tests in HPTN 084 Trial
Presenter
Mina Hosseinipour
Authors
M. Hosseinipour1,2, E. Voldal3, B. Hanscom3, S. Eshleman4, E. Piwowar-Manning4, Y. Agyei4, K Nyarota5, F Angira6, S. Dadabhai7, D. Gadama2, BG Mirembe8, P. Hunidzarira5, S. Innes9, D Kalonji10, J. Makhema11, P Mandima5, A. Marais12, J Mpendo13, P Mukwekwerere5, N Mgodi5, V Naidoo14, P Nahirya Ntege15, H Nuwagaba-Biribonwoha16, E Roos17, N Singh10, B Siziba9, E Spooner10, J. Farrior18, S. Rose18, L. Soto-Torres19, J Rooney20, A.R. Rinehart21, R Landovitz22, M.S. Cohen1, S. Delany-Moretlwe17, on behalf of the HPTN 084 study team
Institutions
1University of North Carolina at Chapel Hill School of Medicine, Chapel Hill, United States, 2UNC Project Malawi, Lilongwe, Malawi, 3Statistical Centre for HIV/AIDS Research and Prevention Fred Hutchinson Cancer Center, Seattle, United States, 4Johns Hopkins University School of Medicine, Department of Pathology, Baltimore, United States, 5University of Zimbabwe Clinical Trials Research Centre, Harare, Zimbabwe, 6Kisumu CRS, KEMRI, Kisumu, Kenya, 7Johns Hopkins Research Project, Kamuzu University of Health Sciences, Blantyre, Malawi, 8Makerere University - Johns Hopkins University Research Collaboration, Kampala, Uganda, 9Desmond Tutu Health Foundation, University of Cape Town, Cape Town, South Africa, 10HIV and other Infectious Diseases Research Unit, South African Medical Research Council, Durban, South Africa, 11Botswana Harvard AIDS Institute Partnership, Gaborone, Botswana, 12Perinatal HIV Research Unit, University of the Witwatersrand, Johannesburg, South Africa, 13UVRI-IAVI, Entebbe, Uganda, 14Desmond Tutu TB Centre, University of Stellenbosch, Stellenbosch, South Africa, 15Baylor College of Medicine Children’s Foundation Uganda, Kampala, Uganda, 16Eswatini Prevention Center, ICAP at Columbia University Mailman School of Public Health, New York, United States, 17Wits RHI, University of the Witwatersrand, Johannesburg, South Africa, 18FHI 360, Durham, United States, 19Division of AIDS, National Institute for Allergy and infectious diseases, Rockville, United States, 20Gilead Sciences, Foster City, United States, 21ViiV Healthcare, Durham, United States, 22UCLA, Los Angeles, United States

BACKGROUND: HPTN 084 showed that injectable cabotegravir (CAB) is effective for PrEP in women. HIV diagnosis in the context of PrEP use may be complicated by both false negative and false positive tests results. We evaluated the positive predictive value (PPV) of the HPTN 084 testing algorithm to guide HIV treatment initiation decisions in women on PrEP.
METHODS: Site HIV testing included rapid tests (RT) and a lab-based antigen/antibody (Ag/Ab) test. Any reactive test prompted study product hold and further HIV status evaluation. Final HIV status was adjudicated based on study site test results and a centralized laboratory. PPV (95% confidence intervals [CI]) for initial site-based reactive visits were assessed for different testing algorithms compared to adjudicated results.
RESULTS: Of 20 sites, 14 used two parallel RT and 6 used one RT. 3180 participants contributed 67314 HIV testing visits. Overall, 5% (162/3180) had 1 reactive test and 74/162 were adjudicated HIV positive by November 2022 (8 CAB, 66 TDF/FTC) while 86 participants did not acquire HIV. The PPV for any reactive RT or Ag/Ab test was 46% (CI: 38%, 54%). Two parallel reactive RT had a PPV of 100% (CI: 91%, 100%). Oraquick Advance as a single test had high PPV (90%, CI: 78%, 97%). Ag/Ab testing detected 73/74 HIV diagnoses but had low PPV (54%, CI: 45%, 62%). PPV of all tests assessed was higher in those randomized to TDF/FTC compared to CAB (Table 1).

Table 1.PPV for initial reactive visits with adjudicated HIV status (162 total, 74 confirmed positive) by treatment arm from HPTN 084
Rapid test (All types)HIV-positive/ total reactive (CAB)
HIV-positive/ total reactive (TDF/FTC)
PPV (95% CI)
(CAB)
PPV (95% CI) (TDF/FTC
Difference in PPV (95% CI) (CAB vs. TDF/FTC)
Overall PPV (95% CI) (CAB and TDF/FTC)
Rapid Test (All types)8/2081/9440% (19%,64%)
86% (78%, 92%)
-46% (-72%, -21%)
78% (69%, 85%)
Alere Determine
(3rd gen)
4/1236/4433% (10%,65%)
82% (67%, 92%)
-48% (-83%, -14%)
71% (58%, 83%)
OraQuick ADVANCE Rapid HIV-1/2 (3rd gen)
4/540/44Insufficient sample size
91% (78%, 97%)
Insufficient sample size
90% (78%, 97%)
Ag/Ab test
7/4266/9417% (7%,31%)
70% (60%, 79%)
-54% (-70%, -37%)
54% (45%, 62%)
Two reactive rapid tests
4/436/36Insufficient sample size
100% (90%, 100%)
Insufficient sample size
100% (91%, 100%)
Any reactive rapid or Ag/Ab test
8/5566/10715% (6%, 27%)
62% (52%, 71%)
-47% (-62%, -33%)
46% (38%, 54%)

CONCLUSIONS: In this study, two reactive RT were sufficient to confirm HIV diagnosis and recommend treatment initiation. With a single reactive HIV test and high frequency of false positive testing, PrEP programs should anticipate the need for further testing, counseling about false positivity, and plans to resume PrEP after excluding HIV. More data is needed to determine if additional testing may be required in the setting of CAB.