乌鸦传媒

An extended testing protocol helped reclassify incident HIV infections as baseline infections in a randomized trial involving patients using long-acting cabotegravir as injectable pre-exposure prophylaxis (PrEP), a researcher said.

In post-hoc analyses, annual HIV incidence within the cabotegravir arm of declined to 0.37 from 0.41, and the hazard ratio declined from 0.34 to 0.32, reported Raphael Landovitz, MD, of University of California Los Angeles, and colleagues.

Notably, extended testing did not reveal any changes in the tenofovir/emtricitabine (Truvada) arm, with 39 incident infections, nearly all due to suboptimal or no adherence to study drug.

Landovitz's group performed a post-hoc analysis of the phase IIb/III trial presented at the virtual International AIDS Conference in July 2020, and shared their results at a late-breaking session at the virtual Conference on Retroviruses and Opportunistic Infections (CROI). The trial compared PrEP with injectable cabotegravir versus daily oral PrEP among HIV-negative men who have sex with men (MSM) and transgender women. At the time, Landovitz reported 13 incident infections in the cabotegravir arm and two baseline infections. The study was stopped early due to exceeding the superiority boundary.

Landovitz said discordant results in a standard HIV testing algorithm within the trial prompted additional nucleic acid-based testing at sites, which led to retrospective testing of stored samples. The post-hoc analysis found that one of the incident infections was actually a baseline infection, and identified another baseline infection, meaning there were 12 incident infections and four baseline infections.

"The level of viremia was often low at the first HIV positive visit and there was often a prolonged period of low-level viremia or viral suppression after infection," Landovitz said.

For the open-label part of the trial, he said they are currently exploring using viral load testing as primary screening for HIV infection, and oral lead-in will be optional as well. The current screening algorithm included real-time point-of-care antibody testing and antigen antibody testing at every visit, including an obligatory negative RNA test within 14 days of enrollment.

During a question-and-answer session, Landovitz framed this as essentially a testing issue, as the currently available diagnostics are less sensitive than nucleic acid-based testing.

"This is a call to action for the expedited development of more sensitive, potentially point-of-care, affordable, implementable diagnostics that can identify incident HIV infections at earlier time points than our current diagnostic algorithms do," he said, adding this could have implications for other long-acting antivirals currently in development.

Revised classifications resulted in Landovitz and colleagues sorting these infections into four groups: baseline infections prior to any study product (four infections), infections that occurred when a patient had a long hiatus from last administration of cabotegravir at the time of detection (five infections), infections for which detection occurred during oral lead-in phase (three infections), and infections that were identified despite on-time cabotegravir injections (four infections).

Interestingly, when examining these cases by groups, the four cases identified despite adherence to cabotegravir had evidence of testing positive for HIV weeks earlier than time of detection at the site.

Evaluation of the correlates of protection is ongoing, but long-acting cabotegravir can delay detection of infection using standard HIV testing algorithms, Landovitz said. But he also pointed out some potential disadvantages to using more sensitive diagnostics for early identification of infection, noting that diagnostics that are too sensitive can "send us down a pathway that isn't resource-efficient and causes more distress for consumers and providers because of time awaiting testing results."

Comment