Theophylline has no advantages over caffeine as a putative model drug for assessing CYP1A2 activity in humans

Aims The cytochrome P4501A2 (CYP1A2) catalyses the metabolism of a number of clinically used drugs, and thus there is an interest in determining the activity of CYP1A2 in patients before treatment with CYP1A2 substrates. Caffeine is the most commonly used model drug to assess CYP1A2 function, but due to the complex metabolism of caffeine, there is a need for an alternative drug to use as an index of CYP1A2 activity. In this study the CYP1A2 substrate theophylline was tested as a possible alternative to caffeine as a model drug for CYP1A2. Methods Twelve healthy volunteers ingested 200 mg of caffeine, and the caffeine metabolic ratios (CMR), CMR(urine)=(AFMU+1MX+1MU)/17DMU and CMR(plasma)=17DMX/137TMX were determined 6 h after drug intake. After a period of about 2 months the volunteers ingested 257 mg theophylline and blood samples were drawn and urine was collected during the following 48 h. The oral and partial clearances of theophylline were calculated via N-demethylation and 8-hydroxylation. The theophylline metabolic ratios, 1MU/13DMX and 3MX/13DMX being evaluated as indices of CYP1A2 catalysed N-demethylation and 13DMU/13DMX as an index of party CYP1A2 catalysed 8-hydroxylation, were estimated in 0-12 h, 0-24 h and 0-48 h urine samples, and in plasma and spot urine samples 6 h after the intake of theophylline. Results The theophylline plasma ratios for the N-demethylation pathways correlated with the oral clearance of theophylline (r(s)=0.881-0.934, P<0.001) and with the respective formation clearances of the metabolites (r(s)=0.712-0.925, P<0.05). Furthermore, all of the theophylline plasma ratios correlated with the caffeine plasma ratio (r(s)=0.645-0.663, P<0.05). None of the caffeine metabolic ratios and none of the 6 h urinary theophylline ratios correlated with the oral or the partial clearances of theophylline (r(s)=0.042-0.556, P>0.05). The theophylline 0-12 h urine ratios correlated with the oral clearance of theophylline (r(s)=0.677-0.757, P<0.05) and with the respective formation clearances of the metabolites (r(s)=0.705-0.750, P<0.05). However, none of the theophylline urine ratios correlated with any of the caffeine metabolic ratios. Conclusions In summary the theophylline 6h plasma and 0-12 h urine ratios 1MU/13DMX and 3MX/13DMX, both reflecting N-demethylation seem to be predictors of the CYP1A2 mediated metabolism of theophylline, whereas only the plasma ratio correlated with the caffeine plasma 17DMX/137TMX ratio. Thus, it would appear that the plasma theophylline N-demethylation ratios are superior to the urine ratios as indices of CYP1A2 activity. However, because in some individuals the concentrations of theophylline metabolites in plasma were close to the limit of detection, it is concluded that theophylline does not have marked advantages over caffeine as a model drug for assessing CYP1A2 activity.