Iatrogenic unscheduled (breakthrough) endometrial bleeding

Bleeding patterns of HIV-infected women using an etonogestrel-releasing contraceptive implant and efavirenz-based or lopinavir/ritonavir-based antiretroviral therapy

OBJECTIVES

The use of progestogen-only contraceptives may cause a change in bleeding pattern, which is a common cause of discontinuation of these methods. Co-administration with some antiretroviral therapies (ART) changes the bioavailability of the etonogestrel (ENG)-releasing contraceptive implant, possibly affecting the bleeding pattern. Bleeding patterns were evaluated in HIV-positive users of the ENG implant co-administered with two common ART regimens.

METHODS

Forty-five HIV-positive women who wished to use an ENG implant were included in this study: 15 had received zidovudine/lamivudine (AZT/3TC) + lopinavir/ritonavir (LPV/r) for ≥3 months (LPV/r-based ART group), 15 had received AZT/3TC + efavirenz (EFV) for ≥3 months (EFV-based ART group), and 15 had not received ART (non-ART group). Bleeding patterns were evaluated at 3 and 6 months after implant placement using a standard bleeding calendar.

RESULTS

Amenorrhoea and infrequent bleeding rates were higher in the LPV/r-based ART group (50% and 36%, respectively) than in the other groups (non-ART group, 36% and 29%, respectively; EFV-based ART group, 7% and 14.5%, respectively; p = 0.01). The EFV-based ART group more frequently had regular bleeding (71.5%) compared with the other groups (LPV/r-based ART group, 7%; non-ART group, 21%; p = 0.01). The proportions of women with frequent and prolonged bleeding were similar (p > 0.05) in the three groups.

CONCLUSIONS

The co-administration of EFV-based or LPV/r-based ART with the ENG implant affected the expected bleeding patterns during use of the implant, although unfavourable bleeding (frequent and prolonged) was not associated with the medications under evaluation.