Efavirenz use during pregnancy and for women of child-bearing potential.

Background: Efavirenz is the preferred non-nucleoside reverse transcriptase inhibitor for first-line antiretroviral treatment in many countries. For women of childbearing potential, advantages of efavirenz are balanced by concerns that it is teratogenic. This paper reviews evidence of efavirenz teratogenicity and considers implications in common clinical scenarios. Findings: Concerns of efavirenz-induced fetal effects stem from animal studies, although the predictive value of animal data for humans is unknown. Four retrospective cases of central nervous system birth defects in infants with first trimester exposure to efavirenz have been interpreted as being consistent with animal data. In a prospective pregnancy registry, which is subject to fewer potential biases, no increase was detected in overall risk of birth defects following exposure to efavirenz in the first-trimester. Discussion: For women planning a pregnancy or not using contraception, efavirenz should be avoided if alternatives are available. According to WHO guidelines for resource-constrained settings, benefits of efavirenz are likely to outweigh risks for women using contraception. Women who become pregnant while receiving efavirenz often consider drug substitution or temporarily suspending treatment. Both options have substantial risks for maternal and fetal health which, we argue, appear unjustified after the critical period of organogenesis (3-8 weeks post-conception). Efavirenz-based triple regimens, initiated after the first trimester of pregnancy and discontinued after childbirth, are potentially an important alternative for reducing mother-to-child transmission in pregnant women who do not yet require antiretroviral treatment. Conclusion: Current recommendations for care for women who become pregnant while receiving efavirenz may need to be re-considered, particularly in settings with limited alternative drugs and laboratory monitoring. With current data limitations, additional adequately powered prospective studies are needed. [ABSTRACT FROM AUTHOR]