Menstrual cycle and zidovudine pharmacokinetics in HIV-infected women

Objective: To examine the pharmacokinetics of zidovudine during each menstr ual cycle phase in women with HIV-1 infection. Design: Open-label, nonblinded study. Setting: The immunodeficiency clinic at the Eric County Medical Center, Buf falo, New York. Patients: 12 HIV-seropositive women started and completed the study. Inclus ion criteria for subjects included an acceptable medical history, chemistry profile, a complete blood count with differential, lymphocyte profile, uri nalysis and history of regular menstrual cycles. Interventions: All patients received a 100mg dose of zidovudine in a fasted state on three occasions. Main Outcome Measures: 8-hour pharmacokinetic profiles in the 12 women were obtained during the menstrual, follicular and luteal phases of the menstru al cycle. Serum estradiol, progesterone, cortisol and CD4+ and CD8+ cell co unts were also obtained during each menstrual cycle phase. Results: Considerable intra- and interpatient variability in zidovudine pha rmacokinetics was evident. Plasma concentrations declined in a linear fashi on over 4 hours following drug administration, with a prolonged elimination phase for most subjects. The overall mean elimination half-life of zidovud ine was 1.6 +/- 0.76h (range 0.56 to 3.7h) and the overall mean oral cleara nce was 2.32 +/- 0.88 L/h/kg (range 1 to 4.7 L/hfkg). Time to maximum plasm a drug concentration was 0.58 +/- 0.27h, 0.77 +/- 0.64h and 0.63 +/- 0.29h during the menstrual, follicular and lutcal phases, respectively (p = 0.555 ). The mean peak plasma concentration of zidovudine was 684 +/- 502 mug/L, 666 +/- 562 mug/L and 527 +/- 210 mug/L during the three phases, respective ly (p = 0.472). The mean zidovudine area under the concentration-time curve (AUC) during the studied phases was 781 +/- 279 mug.h/L, 875 +/- 391 mug.h /L and 693 +/- 176 mug.h/L, respectively (p = 0.128). The AUC was negativel y correlated with estradiol serum concentration (r = -0.329, p < 0.05). Conclusion: There were no significant differences in the zidovudine pharmac okinetic parameters across the menstrual cycle phases, but there was high i ntersubject variability during each cycle phase. A significant negative rel ationship was found between zidovudine AUC and estradiol concentrations, su ggesting that zidovudine glucuronidation may change in relation to the mens trual cycle phase. Hormonal variation and fluctuations in women, especially estrogen, may therefore explain in part the unpredictability of zidovudine exposure after an oral dose. This may impact plasma concentrations of zido vudine and potentially cause fluctuating concentrations throughout prolonge d periods each month.