Grading: 1C The choice of third agent should be based on safety, tolerability and efficacy in pregnancy. Based on

non-pregnant adults, BHIVA guidelines for the treatment of HIV-1 positive adults with antiretroviral therapy 2012 (www.bhiva.org/PublishedandApproved.aspx) recommended an NNRTI, with efavirenz preferred to nevirapine, or a boosted PI of which lopinavir or atazanavir have been most widely prescribed. For the pregnant woman, there is more experience with nevirapine as efavirenz has until recently been avoided in pregnancy. The Writing Group consider there to be insufficient evidence to recommend the learn more avoidance of efavirenz in the first trimester of pregnancy, and include efavirenz in the list of compounds that may be initiated during pregnancy. Despite the well-documented cutaneous, mucosal and hepatotoxicity with nevirapine at higher CD4 T-lymphocyte counts, nevirapine remains an option for women with a CD4 T-lymphocyte count <250 cells/μL. Nevirapine is well tolerated in pregnancy, with several studies suggesting this to be the case even above the stated ICG-001 ic50 CD4 cell count cut-off [68-71]; has favourable pharmacokinetics in pregnancy [72-74] and has been shown to reduce the risk of MTCT even when given as a single dose in labour, alone or supplementing zidovudine monotherapy or dual therapy [75-77]. Despite some concerns regarding diabetes, PTD (see below)

and pharmacokinetics during the third trimester (discussed separately) several ritonavir-boosted PIs have been shown to be effective as the third agent in HAART in pregnancy (lopinavir [66],[78], atazanavir [79], saquinavir [80],[81]). In the European Collaborative Study, time to undetectable VL was longer in women initiating PI-based HAART; however, in this study 80% of these women were taking nelfinavir [82]. In a more recent study, treatment with a boosted PI resulted in more rapid viral suppression (to <50 HIV RNA copies/mL) than nevirapine, except in the highest

VL quartile [83]. In another multicentre study nevirapine-based HAART reduced VL more rapidly during the first 2 weeks of therapy than PI-based HAART with nelfinavir, Methocarbamol atazanavir or lopinavir, but time to undetectable was influenced by baseline VL rather than choice of HAART [84]. The role of newer PIs (e.g. darunavir), integrase inhibitors and entry inhibitors in the treatment-naïve pregnant patient has yet to be determined; therefore other, more established, options should preferentially be initiated. The data on the association of HAART and PTD are conflicting. Some studies implicate boosted PIs, others do not. The data are summarized below. The association between HAART and PTD was first reported by the Swiss Cohort in 1998 [60],[85], and subsequently by a number of other European studies, including three analyses from the ECS [60],[86-88]. Analysis of the NSHPC UK and Ireland data in 2007 found there to be a 1.5-fold increased risk of PTD when comparing women on HAART with those on mono- or dual therapy [89].