Evaluation of caffeine as an in vivo probe for CYP1A2 using measurements in plasma, saliva, and urine.

Twenty-five healthy volunteers were given 100 mg caffeine orally and several estimates of cytochrome P450 1A2 (CYP1A2) activity were evaluated. The validation was performed by correlation of different parameters in plasma, saliva, and urine to two measures of caffeine clearance, CL(oral) and CL(137X-->17X) that served as standards of reference. Two subjects were excluded because of noncompliance with a caffeine-free diet. In the remaining 23 subjects, both plasma and saliva total clearances of caffeine were highly correlated with each other (r(s) = 0.97, p < 0.0001). The ratio 17X/137X restricted to one sampling point taken 4 hours after dose, showed a high correlation (r(s)) with CL(oral) and CL(137X-->17X) in plasma (0.84/0.83) and saliva (0.82/0.77) (p < 0.0001 for all the correlation values) where 17X is 1,7-dimethylxanthine (paraxanthine) and 137X is 1,3,7-trimethylxanthine (caffeine). Additionally, the ratio (AFMU + 1U + 1X + 17U + 17X)/137X in a 0-24 hours urine sampling showed the highest correlation with CL(137X-->17X) (r(s) = 0.85, p < 0.001) where AFMU is 5-acetylamino-6-formylamino-3-methyluracil, 1U is 1-methyluracil, 1X is 1-methylxanthine, and 17U is 1,7-dimethyluric acid. The major estimates of CYP1A2 activity were significantly less in nonsmoking females, and this probably was related to the use of oral contraceptives in this subpopulation. In summary, among caffeine-based approaches for CYP1A2, the authors recommend either plasma or saliva 17X/137X ratio and the urinary (AFMU + 1U + 1X + 17U + 17X)/137X ratio during a sampling interval of at least 8 hours, starting at time zero since caffeine intake. These indices are simple, reliable, and relatively inexpensive estimates of CYP1A2 activity to be used in the study of human populations.