Background:  Data are limited on the use of atazanavir (ATV) in pregnancy, and there are concerns regarding the possibility of neonatal hyperbilirubinemia. Despite this, ATV is increasingly used in pregnancy, either when women conceive on therapy, or when it is initiated in pregnancy due to adherence, resistance, or tolerability issues.

Methods:  We reviewed case notes of all women attending 9 HIV units in London, who were receiving ATV in pregnancy. Data were collected on maternal CD4 count, viral load, toxicity, and plasma ATV concentrations. Infant data included serum bilirubin, use of phototherapy, and, where available, HIV status at 3 months.

Results:  We identified 31 women with a total of 33 pregnancies:  20 women had conceived when on ATV. At first antenatal visit, 16 had an undetectable viral load and 4 had viral loads of 98, 111, 673, and 60,300 copies/mL, respectively; their median CD4 count was 270 (range 63 to 930) x 10⁹/L; 13 women initiated ATV at a median of 22 weeks (range 5 to 33) gestation. They had a median viral load of 255 copies/mL (range 50 to 101,454) and a median CD4 count of 130 (range 61 to 660) x 10⁹/L at baseline. One mother was co-infected with hepatitis B virus (HBV); none with hepatitis C virus (HCV). Pre-treatment serum bilirubin and alanine aminotransferase (ALT) were within normal limits. At delivery, median viral load and CD4 counts were <50 copies/mL (range 50 to 28,189 copies/mL) and 260 (range 67 to 910) x 10⁹/L, respectively. Median maternal bilirubin level was 30 (range 4 to 76) µmol/l [1.8 (range 0.2-4.4) mg/dL] and 23/25 mothers had a raised bilirubin (>17µM/L [>1 mg/dL]). For 15 mothers, we measured ATV concentrations, performed at a median of 30 weeks’ gestation; 14 were taken in the third trimester. Mean trough ATV concentration was 421 mg/L and all but 1 were above the recommended therapeutic concentration of 100 mg/L. Of the total, 26 infants were born at a median of 38 weeks gestation (17 planned cesarian sections, 7 emergency cesarian sections, 2 planned spontaneous vaginal delivery). Mean birth weight was 2894 g and median infant bilirubin levels were 71 µM/L (range 10 to 191) (4.2 mg/dL, range 0.6 to 11.2). No infants required phototherapy and no birth defects were detected.

All infants are thus far HIV uninfected. There were 2 miscarriages, at 12 and 16 weeks, respectively, 5 women were still pregnant.

Conclusions:  To our knowledge, this is the largest reported case series of the use of ATV in pregnancy. Maternal plasma levels, where measured, were adequate. No early infant morbidity as a result of in utero exposure to ATV was demonstrated.