Background: PrEP adherence was low in efficacy trials among African adolescent girls and young women (AGYW). Adherence among African AGYW after PrEP is known to be efficacious and with additional adherence support are unknown.
Methods: HPTN 082 enrolled sexually-active AGYW ages 16-25 in Cape Town and Johannesburg, South Africa and Harare, Zimbabwe. AGYW were randomized to standard adherence support (counseling, 2-way SMS, and adherence clubs) or standard support plus drug level feedback at 2 and 3 months (M) with follow-up at M 6, 9 and 12. Adherence was assessed by tenofovir-diphosphate (TFV-DP) in dried blood spots (DBS) measuring use in the prior month, and plasma tenofovir (TFV) measuring use in the prior week; high adherence is defined as TFV-DP >700 fmol/punch and plasma TFV>40 ng/ml, given association with HIV protection in trials.
Results: 427 AGYW started PrEP; median age was 21 and median VOICE risk score was 7 (score ≥5 associated with >6% HIV incidence in prior cohorts). 212 AGYW were randomized to standard and 215 to drug level feedback; 74 (17%) discontinued PrEP by M12, most commonly due to pregnancy (n=13) and participant preference (n=19). At M3, 85% took PrEP (detectable TFV-DP; 66% detectable TFV by plasma), and 25% had high adherence by DBS and 48% by plasma. There were no differences by arm in proportions with detectable TFV-DP or high adherence by DBS at M3 and M6 or plasma TFV at M6 and 12 (all p>0.3). Adherence decreased significantly from M3-M12 when visits decreased to quarterly (p< 0.0001). Four acquired HIV, all of whom had undetectable plasma tenofovir at seroconversion.
Conclusions: Most African AGYW were taking PrEP in the first 3 months, and a substantial minority had high adherence by DBS and plasma, which was >2-fold higher than in PrEP efficacy trials. Adherence declined significantly from M3-12, similar to other PrEP studies in youth. PrEP adherence did not increase with addition of drug level feedback. The four (1%) who acquired HIV were not taking PrEP. This combination prevention package that included PrEP achieved high protection. Research is needed to determine effective adherence support to sustain PrEP use among African AGYW.