Adherence challenges with daily oral pre-exposure prophylaxis during pregnancy and the postpartum period in South African women: a cohort study.

Abstract

Daily oral pre-exposure prophylaxis (PrEP) can reduce HIV acquisition. However, prevention effectiveness requires daily adherence prior to and during periods of sexual activity. Little is known about pharmacologic measures of PrEP adherence during pregnancy and postpartum and the factors related to optimal adherence during periods of sexual activity in this population. Between August 2019 and October 2021, we enrolled pregnant women without HIV at their first antenatal care visit followed-up through 12 months postpartum. Eligible women ≥16 years old received HIV prevention counselling and were offered oral PrEP (TDF-FTC). We quantified tenofovir-diphosphate (TFV-DP) in dried blood spots in women who reported taking PrEP in the past 30 days (at quarterly follow-up visits). We used regression models with generalized estimating equations to evaluate correlates of TFV-DP (any vs. none, and ≥2 vs. <2 doses/week), adjusting for maternal age and pregnancy status. In 382 women who started PrEP in pregnancy, returned for follow-up and reported PrEP use in the past 30 days, the median age was 27 years (interquartile range [IQR] = 23-32), and the median time on PrEP was 168 days (IQR = 84-252 days). Half of the samples had quantifiable TFV-DP at any time point (52%), declining from 67% of pregnant women 3 months post-initiation to 31% of postpartum women by 12 months. Overall, 72% had concentrations corresponding to <2 doses/week; 25% ≥2 doses/week; 3% 7 doses/week. Concentrations were lower in postpartum versus pregnancy (age-adjusted odds ratio [aOR] = 0.44; 95% confidence interval [CI] = 0.35-0.54). The correlation of self-reported adherence and TFV-DP ranged from -0.07 in pregnancy to 0.25 in postpartum women. Variables associated with having quantifiable TFV-DP included partner living with HIV/unknown serostatus (aOR = 1.50; 95% CI = 1.01-2.22), and reported frequency of sexual activity in the past month (aOR sex >5/month vs. no sex or <5 times/month = 2.11; 95% CI = 1.58-2.82) adjusting for age and pregnancy versus postpartum status. TFV-DP concentrations declined over follow-up time (aOR for 6 vs. 3 months = 0.49; 95% CI = 0.36-0.67). Objectively measured adherence to PrEP was low overall and did not correlate with self-reported use. There is an urgent need for objective adherence measures to support clinical decision-making as well as adherence support interventions as part of PrEP services for pregnant and postpartum women at risk of HIV.