Analysis of polymorphic variation in drug metabolism: III. Glucuronidation and sulfation of diflunisal in man.

The urinary excretion of diflunisal (D) and its metabolites diflunisal sulfate (DS), diflunisal phenolic glucuronide (DPG), and diflunisal acyl glucuronide (DAG) were measured in 110 healthy, drug-free Caucasian volunteers given 50 mg of diflunisal by mouth. When expressed as fractional recoveries, DS, DPG, and DAG were strongly negatively correlated with one another. Metabolic ratios, on the other hand, correlated positively and tended to localize variability within a single enzyme pathway. Thus, females using estrogen-containing oral contraceptives were shown to excrete 50% less DS and 20% more DAG than non-users, and recoveries of DS were reduced by about 30% in cigarette smokers. Kernel density analyses of the log metabolic ratios of DS and DPG were broad-based and unimodal. However, kernel density estimates of the distribution of log metabolic ratios of DAG showed 3 peaks, 1 of which (an extensive metabolizer polymorph) could be removed by excluding contraceptive-using females. Similarly, there were 2 poor metabolizer peaks in the distribution of log metabolic ratios of DS attributable to cigarette smoking and, in females, use of an oral contraceptive. Thus, we conclude that the metabolism of diflunisal is altered by cigarette smoking and oral contraceptives, and that kernel density estimation, as applied to log metabolic ratios, is a sensitive and specific method for detection of polymorphic variation in drug metabolism.