Abstract Body

Background: Prior to promoting tenofovir gel for HIV prevention in women, it is important to know whether tenofovir gel users who acquire HIV infection can be treated with tenofovir containing antiretroviral treatment (ART) with good clinical outcomes. The CAPRISA009 randomised clinical trial assessed ART outcomes in women who acquired HIV during tenofovir gel trials.

Methods: Seroconvertors from the CAPRISA004 tenofovir gel effectiveness trial and CAPRISA008 tenofovir gel implementation trial were recruited into the CAPRISA009 study when eligible for ART initiation (CD4 count <350 cells/µl, pregnancy or AIDS-defining illness). Women were randomised to a tenofovir-containing or tenofovir-sparing (zidovudine-containing) regimen. CD4 counts and viral loads (VL) were monitored at baseline and follow-up visits. Clinical outcomes, virological suppression (VL <50 copies/ml), CD4 counts, adverse events and drug switches were compared between the two groups using Fisher’s exact and Wilcoxon rank sum tests.

Results: From June 2011 to August 2014, 59 women were enrolled and followed-up for a median of 13 months (IQR 6-19). Twenty-nine women (7 tenofovir gel exposed) were randomized to a tenofovir-containing and 30 (9 tenofovir gel exposed) to a tenofovir-sparing regimen. Median baseline CD4 count and VL were 361 cells/µl (IQR 275–420) and 4.64 log copies/ml (IQR 4.01-5.11), and did not differ by ART assignment. Overall VL suppression rates were 85.7% and 76.2% at 6 months (p=0.70) and 84.2% and 76.5% at 12 months (p=0.68) in the tenofovir-containing and tenofovir-sparing groups. In women with prior tenofovir exposure, VL suppression rates were 100% and 50% at 6 months (p=0.14) and 100% and 25% at 12 months (p=0.17) in the tenofovir-containing versus tenofovir-sparing groups. Overall, median CD4 counts at 12 months were 577 cells/µl and 541 cells/µl (p=0.45) and 388 cells/µl and 549 cells/µl (p= 0.35) in tenofovir gel exposed women in the tenofovir-containing and tenofovir-sparing groups. Women randomised to a tenofovir-sparing regimen experienced a higher frequency of grade 3 or 4 AEs (33.3% vs 10.3%, p=0.06) and had more toxicity-related drug switches (26.7% vs. 0.0%, p< 0.01) compared to women receiving a tenofovir-containing regimen. There were no drug switches in women with prior tenofovir gel exposure.

Conclusions: Tenofovir-containing ART was effective and safer in tenofovir gel trial seroconvertors and should be recommended as the preferred treatment option for women with prior exposure to tenofovir gel.