Background:  Antiretroviral exposure is a key determinate of treatment outcome, but significant interindividual PK variability complicates its estimation. Intensive PK studies are traditionally performed in a small number of highly selected patients to minimize variability, a strategy that limits study of factors that influence drug exposure under conditions of actual use.

Methods:  The Women’s Interagency HIV Study (WIHS) performed intensive PK studies to identify factors influencing exposure to atazanavir (ATV) in a representative sample. The 24 hour PK studies were completed for 122 women on ATV-based regimens under conditions of actual use including concomitant medications, cigarette smoking if habitual, and usual diet. Factors that could influence PK parameters were measured, including race, age, ritonavir (RTV) use, pregnancy, oral contraceptive (OCP) or hormone replacement (HRT), menstrual stage, interacting medications, recreational drugs, smoking, liver and renal function, weight, and dietary fat intake.  Variables were examined via univariate models and multivariable models were constructed by forward stepwise selection, including only variables with P <0.05. 

Results:  Area under the curve (AUC) measurements were calculated from each participants’ PK study to estimate exposure. RTV use increased exposure by 3.46 (95%CI 2.5 to 4.8, P <0.001)-fold. Patients who used oral or injectable hormones had 0.36 (0.19 to 0.70, p 0.003)-fold lower AUC. A creatinine clearance of <60 mL/min compared to >60 resulted in increases in ATV AUC of 1.69 (1.06 to 2.71, P <0.029)-fold. Diarrhea with 3 or more soft/liquid stools per day reduced ATV exposure by 0.58 (0.38 to 0.90, P<0.016)-fold. For every doubling in bilirubin level, AUC increased by 1.86 (1.14 to 3.03, P <0.013)-fold. Exogenous hormones, RTV, impaired renal function and bilirubin were all independently associated with ATV AUC in multivariate models (Table) controlling for significant variables, as well as age, race, and weight.

Conclusions:  These 24-hour intensive PK studies in 122 women on ATV under conditions of actual use demonstrated the expected boosting of exposure by RTV and the association between ATV levels and bilirubin. We also found that commonly used exogenous hormones (OCP/HRT) had significant effects on ATV exposure in women. Moderate renal impairment also influenced ATV exposure in multivariate models. Further examination of the impact of these parameters on ATV outcomes in chronically treated patients is needed.