Bone and Renal Health in Infants With or Without Breastmilk Exposure to Tenofovir-Based Maternal Antiretroviral Treatment in the PROMISE Randomized Trial

We assessed bone and kidney outcomes in infants randomized postdelivery as mother–infant pairs within the IMPAACT PROMISE trial to maternal tenofovir disoproxil fumarate–based antiretroviral treatment (mART) or infant nevirapine prophylaxis (iNVP) to prevent breastfeeding HIV transmission. METHODS: Infants were coenrolled in the P1084s substudy on randomization day and followed through Week 74. Lumbar spine bone mineral content (LS-BMC) was assessed at entry (6–21 age days) and Week 26 by dual-energy x-ray absorptiometry. Creatinine clearance (CrCl) was calculated at entry; Weeks 10, 26, and 74. Student t tests compared mean LS-BMC and CrCl at Week 26 and mean change from entry between arms. RESULTS: Of 400 enrolled infants, the mean (SD; n) for entry LS-BMC was 1.68 g (0.35; n = 363) and CrCl was 64.2 mL/min/1.73 m(2) (24.6; n = 357). At Week 26, 98% of infants were breastfeeding and 96% on their assigned HIV prevention strategy. The mean (SD) Week 26 LS-BMC was 2.64 g (0.48) for mART and 2.77 g (0.44) for iNVP; mean difference (95% confidence interval [CI]) −0.13 g (−0.22 to −0.04), P = 0.007, n = 375/398 (94%). Mean absolute (−0.14 g [−0.23 to −0.06]) and percent (−10.88% [−18.53 to −3.23]) increase in LS-BMC from entry was smaller for mART than iNVP. At Week 26, the mean (SD) CrCl was 130.0 mL/min/1.73 m(2) (34.9) for mART vs. 126.1 mL/min/1.73 m(2) (30.0) for iNVP; mean difference (95% CI) 3.8 (−3.0 to 10.7), P = 0.27, n = 349/398 (88%). CONCLUSION: Week 26 mean LS-BMC was lower in infants in the mART group compared with the iNVP group. However, this difference (∼0.23 g) was less than one-half SD, considered potentially clinically relevant. No infant renal safety concerns were observed.