New option for PrEP – TAF/FTC is non-inferior to TDF/FTC: results of phase 3 DISCOVER study

Simon Collins, HIV i-Base

The results from the large international phase 3 DISCOVER study are notable showing that a new version of PrEP is at least as affective at protecting against HIV infection as the currently approved formulation.

The study was also able to do this despite much lower numbers of people becoming positive than were predicted when the research was first planned. [1, 2]

For ethical reasons, this study did not include a placebo control arm with all participants taking both one active and one placebo pill each day. These two active arms comparing the a new formulation of tenofovir alafenamide/emtricitabine (TAF/FTC) to already approved tenofovir DF/FTC (TDF/FTC). And while the study was running, rates of new HIV infections are likely to have become lower (due partly at least to the wider availability of PrEP).

The results were given as a late-breaking oral presentation at CROI 2019 by Charles Hare from the University of California San Francisco.

From September 2016 to May 2017, the study enrolled 5387 gay/bisexual men or transgender women who were randomised 1:1 to either daily TAF/FTC or TDF/FTC plus appropriate placebo.

Baseline demographics included median age 34 years (range 18 to 76), and 74 participants were transgender women. Ethnicity was reported as 84% white, 9% black and 5% Asian – but also referred to 24% of participants being Latin/Hispanic.

During the study 17% vs 16% (n = 452 vs 430) participants discontinued treatment, in the TAF vs TDF arms respectively. This was mainly because of loss to follow-up (LTFU) (n=201 vs 170) or participant decision (n=193 vs 175, with much fewer participants (36 vs 49) stopping due to side effects.

Entry criteria also included being at high HIV risk, with a definition that included one or more risk factors: having had receptive sex without condoms at least twice in the previous three months (60%), having had rectal chlamydia or gonorrhoea in the previous six months (each for >10% participants), recent syphilis, recreational drug use (67%), binge drinking (22%) or to be already taking PrEP at baseline (16%).

The primary endpoint was the number of HIV infections per 100 patient years of follow-up (PYFU) after all participants have a minimum of 48 weeks follow up and 50% of participants have follow-up for 96 weeks.

After a total of 8756 PYFU there were only 22 new HIV infections: 7 vs 15 in the TAF/FTC vs TDF/FTC arms respectively, none of which were in transgender women. This gave an incidence of 0.16 vs 0.34 per 100 PYFU and an incidence rate ratio (IRR) of 0.47 (95%CI: 0.19 to 1.5) numerically in favour of TAF/FTC (though not statistically significant) and meeting criteria for non-inferiority.

However, of the people diagnosed HIV positive during the study, 5 participants were found to have been HIV positive at baseline (1 vs 4) and 15 had low drug levels (5 vs 10) at the time of HIV diagnosis, suggesting suboptimal adherence. This left only one participant in each arm who became HIV positive with medium or high drug levels, suggesting higher adherence. Non-inferiority was also maintained in a sensitivity analysis that excluded participants who were HIV positive at baseline (0.55; 95% CI: 0.20 to 1.48).

In 18 samples that were genotyped (n=6 vs 13), the 4 cases of M184V were all in the TDF/FTC arm with no tenofovir resistance in either arm.

Overall, 95% of participants reported adverse events (AEs), with only 20% with a possible link to study drug, nearly all grade 1 or 2. Grade 8 side effects (6% in each arm) with 0.1% vs 0.2% linked to serious drug-related side effects. Only 1% vs 2% of the TAF/FTC vs TDF/FTC arms respectively stopped treatment due to side effects.

There were three deaths (n=1 vs 2: one case of traffic accident and carcinoma and one unknown. Most AEs were symptoms related to sexually transmitted infections (STIs) – reported in about 15% of participants in both arms at all 3-monthly study time points. These include rates of 1.45/100 and 1.38/100 PYFU for TAF/FTC vs TDF/FTC respectively.

Mean percentage change in bone density include + 0.50% vs –1.12 at the spine at spine and +0.18 vs 0.99 in the hip at week 48 in TAF/FTC vs TDF/FTC arms respectively, both (p<0.001).

Renal safety was also better in TAF/FTC group with eGFR changes of +1.8 vs –2.3 mL/min at week 48 (p<0.001). Similar benefits were reported for sub-clinical markers for proximal tubular protein to creatinine ratios.

Finally, given the low number of HIV diagnoses, even with very high rates of STIs linked to HIV risk, the group calculated the background HIV incidence for gay men who not on PrEP in the US cities where the study was running. This was estimated as 4.02 /100 PY (95%CI: 3.96 to 4.09) and was compared to observed rates of 0.08 (95%CI 0.01 to 0.28) and 0.45 (95%CI: 0.23 to 0.78) in the TAF/FTC and TDF/FTC group respectively.

The study concluded that TAF/FTC had high efficacy and good tolerability in high risk gay men and transgender women.


These results are positive in supporting that TAF/FTC is non-inferior to TDF/FTC with safety benefits that might be clinically important both for people <30 years old at a time when bone density is still growing and in older people who are more susceptible to loss of both BMD and renal function.

Drug pricing will be critical for both access and use in nearly every setting, especially with TDF/FTC now off-patent in the EU (though not, apparently, in the US for several years).


  1. Hare CB et al. The phase 3 discover study: daily F/TAF or F/TDF for HIV preexposure prophylaxis. CROI 2019, 4–7 March 2019, Seattle, Washington. Oral abstract LB 104LB. (abstract) (webcast)
  1. Safety and efficacy of emtricitabine and tenofovir alafenamide (F/TAF) fixed-dose combination once daily for pre-exposure prophylaxis in men and transgender women who have sex with men and are at risk of HIV-1 infection (DISCOVER).

Links to other websites are current at date of posting but not maintained.