HPV acquisition and clearance in women randomized to the contraceptives depot medroxyprogesterone acetate, copper intrauterine device, or levonorgestrel implant: a secondary analysis of a randomized clinical trial

Background Intramuscular depot medroxyprogesterone acetate (DMPA-IM), levonorgestrel implants (LNG-implant), and copper intrauterine devices (Cu-IUD) are widely used in Africa. Their impact on human papillomavirus (HPV) incidence and clearance remains unclear. We compared these outcomes in a subsample from a randomised trial. Methods Women (16–35 years) in South Africa and Kenya, randomized to Cu-IUD, LNG-implant, or DMPA-IM (NCT02550067), were followed quarterly for up to 18 months (ECHO Biological Mechanisms Sub-study, [ECHO-BMS], 2016–2018). Endocervical swabs were tested for 36 HPV genotypes. The primary outcome was incident HPV infection among HPV-negative women who contributed follow-up data. Secondary outcomes included HPV clearance, incident persistent HPV infections, and continuous-use analyses. Findings Among the 311 women who constituted the HPV sub-study sample (of the 430 in the parent ECHO trial from sub-study sites), 81 women were HPV-negative at enrolment, 57/81 acquired HPV: 14 (0·75/person-year [0·41–1·23]) in DMPA-IM, 21 (1·94/person-year [1·20–2·96]) in Cu-IUD, and 22 (1·00/person-year [0·62–1·51]) in LNG-implant users. Cu-IUD users had a significantly higher risk of acquiring any HPV than DMPA-IM (HR 2·98, 95% CI 1·50–5·92, p=0·002) and LNG-implant (HR 2·36, 1·27–4·38, p=0·007) users. The difference between LNG-implant and DMPA-IM was insignificant (HR 1·26, 0·64–2·49, p=0·50). These findings were confirmed in supplementary analyses. No difference in HPV clearance was observed. Interpretation HPV acquisition risk differed by contraceptive method, suggesting that HPV outcomes should be considered in contraceptive counselling and the potential benefit of combining HPV vaccination with Cu-IUD use.