Abstract Body

Background:

Safety studies of prenatal PrEP use to date have relied on maternal self-report of PrEP adherence which may not accurately measure infant PrEP exposure. We evaluated perinatal outcomes following maternal PrEP use confirmed with tenofovir-diphosphate (TFV-DP) concentrations in dried blood spots (DBS).

Methods:

Data were analyzed from a subset of women enrolled in a cluster RCT (NCT03070600) evaluating PrEP delivery strategies at 20 clinics in Western Kenya. HIV-negative women were enrolled and offered PrEP during pregnancy and followed through 9 months postpartum regardless of PrEP status. DBS samples were collected at every visit from women who initiated PrEP. A subset of women who self-reported PrEP use in the prior 30 days at antenatal visits were randomly selected and TFV-DP levels were measured in DBS using liquid chromatography/tandem mass spectrometry. Women who were randomly selected but did not have detectable TFV-DP were excluded. PrEP-exposure during pregnancy was defined as having detectable TFV-DP in DBS. Outcomes among women with and without prenatal PrEP exposure were compared, adjusting for partner HIV status, maternal age, gestational age, and syphilis status.

Results:

Overall, 3608 women were included in the analysis; 3505 women were PrEP-unexposed and 103 PrEP initiators were randomly selected and had detectable TFV-DP during pregnancy (18% of all PrEP initiators). Median maternal age at enrollment was 24 years (IQR: 21-28), median gestational age was 24 weeks (IQR: 21-28) and 27% were primigravida. Compared to PrEP-unexposed women, women with confirmed PrEP exposure during pregnancy experienced similar frequencies of stillbirth (4% v 3%, aPR=1.03, 95% CI:0.12-8.93, p: 0.98), preterm birth (16% v 19%, aPR=0.92, 95% CI:0.58-1.47, p: 0.73), low birthweight (0% vs. 2%, p: 0.18), small-for-gestational-age (13% v 10%, aPR=1.41, 95% CI:0.78-2.54, p=0.26), and neonatal death (1% v 2% aPR=0.69, 95% CI:0.09-5.26, p: 0.72). At 9 months, there was no association between prenatal PrEP exposure and frequency of underweight (p=0.68), stunting (p=0.39), or of wasting (p=0.79); results were similar at 6 weeks and 6 months. Among women with any detectable TFV-DP, frequency of adverse perinatal outcomes were similar by TFV-DP levels based on PrEP dosing benchmarks (Table 1).

Conclusions:

Similar to prior safety data that relied on self-reported PrEP use, we found no differences in adverse perinatal outcomes among Kenyan women with prenatal PrEP exposure confirmed with a biologic measure.

Table 1. Birth and infant growth outcomes by confirmed prenatal PrEP exposure status