Clinical implications of antiretroviral drug interactions with warfarin: a case-control study.

Objectives: Warfarin, a frequently prescribed anticoagulant with a narrow therapeutic index, is susceptible to drug-drug interactions with antiretroviral therapy (ART). This study compared the warfarin maintenance dose (WMD) between patients receiving and not receiving ART and evaluated predictors of warfarin dosage among those on ART.

Methods: This was a case-control (1:2) study. Cases were HIV-infected patients receiving warfarin and protease inhibitor (PI)- and/or non-nucleoside reverse transcriptase inhibitor (NNRTI)-based ART. Controls were randomly selected HIV-uninfected patients receiving warfarin. The WMD was compared between cases and controls and between cases on varying ART regimens. Bivariate comparisons were performed and a linear regression model was developed to identify predictors of WMD.

Results: We identified 18 case and 36 control patients eligible for inclusion. Cases were younger than controls (mean age: 45.8 versus 63.1 years, P < 0.01), more often male (72.2% versus 36.1%, P=0.02) and more likely to be African American (50.0% versus 22.2%, P=0.04). ART was classified as PI-based (n=9), NNRTI-based (n=7) and PI + NNRTI-based (n=2). The WMD (mean ± SD) differed between cases and controls (8.6  ±  3.4 mg versus 5.1 ± 1.5 mg, P < 0.01), but not ART regimens (PI: 8.8  ±  4.5 mg; NNRTI: 8.6   ± 1.8 mg; PI + NNRTI: 7.3  ±  3.3 mg; P = 0.86). Race and ritonavir dose were independent predictors of WMD, predicting an increase of 3.9 mg (95% CI: 0.88-6.98, P = 0.02) if a patient was African American or 3.7 mg (95% CI: 0.53-6.89, P = 0.03) if the total daily ritonavir dose was 200 mg.

Conclusions: The required WMD was significantly higher in patients receiving ART. Prompt dose titration to achieve a higher WMD with vigilant monitoring may be required due to these drug-drug interactions.