Assessing Benefits of Pre-Exposure Prophylaxis Relative to Risks of Drug Resistance

Contributing Editor
Robert Grant, M.D., M.P.H.
Robert M. Grant, M.D., M.P.H.

"Resistant infections during PrEP should be weighed against the numbers of infections averted," stated renowned HIV prevention and PrEP (pre-exposure prophylaxis) researcher Robert Grant, M.D., M.P.H., of the Gladstone Institute for Virology and Immunology in a PrEP-focused session at AIDS 2016 in Durban, South Africa. "Each of which will require lifelong therapy and the attendant annual risk of virologic treatment failure with drug resistance," he noted.

At CROI earlier this year, researchers presented the case of the first known HIV acquisition by a person prescribed PrEP and adhering to the regimen, which led to an upsurge of community concern regarding the efficacy of PrEP. Experts have persisted in reassuring would-be PrEP takers and prescribers that these cases will continue to be extremely rare. But just how rare?

Putting this concern into perspective, Grant, who has been a strong, knowledgeable voice in the PrEP conversation since its inception, presented findings from a systematic review of resistance testing results from people who became HIV positive while participating in randomized, placebo-controlled PrEP studies through May 2015.

The analysis comprised results from six clinical trials and one demonstration project. All trials included in the review used rapid antibody tests prior to PrEP initiation and retrospectively tested baseline specimens for HIV RNA among those who became HIV positive. Resistance in all the trials was measured using clinical genotypic tests, which can detect drug resistance present in 20% or more of the virus population, according to Grant; further, four of the trials used sensitive genotypic assays that are able to detect resistant viral variants present in low abundance (greater than about 1% of the virus population).

Grant reminded the Durban audience that drug resistance can be a cause of PrEP failing to be effective or it can be a consequence of said failure. The risk of drug resistance with tenofovir/emtricitabine (Truvada) PrEP tends to occur in people who started or restarted PrEP unaware that they were already experiencing acute HIV infection. Starting PrEP during this period selects for resistance to emtricitabine (FTC, Emtriva) alone. In the single known case of PrEP failure, the patient was exposed to an uncommon HIV strain that was resistant to both tenofovir (TDF, Viread) and emtricitabine.

The absolute risk of excess drug resistance associated with incident or emergent infection in all six trials was 0.05% -- low risk, in other words. Grant added that PrEP would need to be provided to roughly 1,844 people (known as the "number needed to harm," or NNH) to get one drug-resistant infection. Conversely, the number of people who would need to get PrEP in order to prevent one HIV infection is -- depending on the study and adherence results therein -- between 13 and 60 individuals.

"The benefit of PrEP in this overall analysis far exceeds the risk of drug resistance," Grant stated, "and the risk of drug resistance is rare."

Drilling Down the Results

Grant and colleagues compared the active-drug and placebo arms of all studies and divided the analysis according to those who were found to be in acute HIV infection when PrEP was started and those who had emergent infections occur after the study began.

In this analysis, almost all resistance detected was to emtricitabine alone -- a mutation that, as Grant and colleagues noted and Paul Sax, M.D., corroborated in his summary of key AIDS 2016 studies, is treatable.

In the three studies with tenofovir-only arms, resistance to tenofovir did occur in one participant who received tenofovir PrEP. This was a person who was experiencing acute infection at PrEP initiation; no tenofovir resistance was detected in any of the individuals with emergent infection in the active arm.

In the studies that compared Truvada with placebo, there were nine excess drug-resistant infections: eleven in the Truvada arm and two in the placebo arm. However, subtracting the total number of infections in the placebo arm from the total in the active arm yields the number of infections averted by Truvada PrEP in these five studies: Ninety-two infections, or eight (92/9) infections were averted per drug-resistant infection overall. Excluding those who were found to have been experiencing acute HIV infection at PrEP initiation yielded 22 (90/4) infections averted per case of drug-resistant virus.

The four trials that used more sensitive resistance tests turned up more cases of resistance in low abundance -- however, these additional cases did not greatly alter the results of the risk-benefit ratios. There were five excess cases of emtricitabine drug resistance (15 in the active arm, including those who were in acute infection at PrEP initiation, minus 10 in the placebo arm). However, again employing the simple calculation above, there were 101 infections averted in all -- (194+17)-(103+7) -- and 20 (101/5) infections averted per case of drug-resistant infection. Excluding those in acute infection at the start of PrEP, then 91 incident infections were averted for every case of drug-resistant infection.

So, then, is more accurate RNA testing warranted at PrEP initiation? Grant and his co-authors concluded that their analysis supported accepted guidance around use of rapid third-generation antibody tests when PrEP is started. HIV RNA screening for acute infection would certainly increase benefits relative to risk of drug resistance, they noted. However, it would increase costs, is not available in many resource-limited settings and is not required to achieve a favorable PrEP risk-benefit assessment.

As Grant said in a post-CROI interview with BETA: "One chooses whether to focus on the glass 99% full or 1% empty."