Pharmacokinetic model of zidovudine and its glucuronide metabolite in pregnancy

Background/Aims Zidovudine (AZT) drastically reduces mother to infant transmission of HIV. Fetal drug activity is likely critical for effect. We developed a pharmacokinetic (PK) model to predict fetal levels. Methods The model is defined by 6 differential equations and predicts compartmental shifts of AZT and AZT-glucuronide (GAZT) in mother, fetus and amniotic fluid. Assumptions were: oral administration, passive placental transfer and equal amniotic fluid volume for AZT and GAZT. Existing data from 5 baboons were fit using the Nelder-Mead simplex algorithm (WinNonlin). Derived parameters (mean±SD) were compared to independent measures and assessed for physiological validity. Results As expected, metabolic clearance from the mother (91±20 L/h) was higher than with IV infusion. GAZT clearance (35±7.1 L/h) compared well with direct measures. Placental and direct fetal clearances of AZT (1.2±0.65 and 0.18±0.14 L/h) were slightly less than those from paired maternal and fetal infusions. Other parameters (placental GAZT clearance, amniotic fluid clearance) followed physiologic principles. Conclusion A relatively simple PK model can predict disposition of AZT in pregnancy. Applications include minimal sampling strategy to capture PK/PD targets and modeling other drug products used during pregnancy. Clinical Pharmacology & Therapeutics (2005) 77, P41–P41; doi: 10.1016/j.clpt.2004.12.049