Abstract Body

Background:

HPTN 083 demonstrated superiority for long-acting injectable cabotegravir (CAB) compared to daily oral TDF-FTC for HIV pre-exposure prophylaxis (PrEP) in cisgender men and transgender women who have sex with men (MSM, TGW). The study was conducted at 43 sites in North and South America, Asia and Africa. During the blinded and first unblinded year study periods, site-based HIV testing algorithms included a US FDA cleared rapid test (RT) with results prior to product administration, and a laboratory-based antigen/antibody assay (Ag/Ab) that was not resulted until after product administration. We present the PPV of these tests in people receiving CAB or TDF/FTC PrEP.

Methods:

All sites performed RTs and Ag/Ab tests at all study visits and required a non-detected HIV RNA within 14d of study entry; some sites performed two rapid tests prior to product administration based on local practice. HIV status was determined by an external, blinded adjudication committee based on site HIV testing and retrospective HIV testing at a central laboratory. Positive predictive value (PPV, 95% confidence intervals [CI]) for initial site-based reactive testing was assessed for permutations of site test results.

Results:

Of 4566 enrolled participants, 70 were excluded (results could not be adjudicated, reactive test results at enrollment, or no HIV testing data after enrollment), 48 had a false reactive test, 130 had a true reactive test, and 4322 had no reactive tests. The analysis included data from 113,316 visits, including 177 initial reactive visits with a reactive RT or Ag/Ab test. PPVs for one or two RTs (regardless of Ag/Ab result), one RT plus one Ag/Ab test, and one Ag/Ab test (regardless of RT result) are in the Table.

Conclusions:

The PPV of one reactive RT plus one reactive Ag/Ab was 100% for both CAB and TDF-FTC; The PPVs of two reactive RTs for TDF-FTC PrEP and CAB PrEP were 95% and 83% respectively. The PPVs of one reactive Ag/Ab with a negative RT performed were low for both groups and were lower for CAB. In the absence of more sensitive testing, a reactive RT plus a reactive Ag/Ab test, or two reactive RTs had sufficient PPV to warrant initiation of ART. In settings where RNA testing is unavailable or infeasible, algorithms using RTs and Ag/Ab tests had high PPV in the context of MSM/TGW PrEP. Lower PPVs of all tests in CAB cases are attributable to lower HIV incidence in CAB arm participants.