Prediction Methods for Nicotine Clearance Using Cotinine and 3-Hydroxy-Cotinine Spot Saliva Samples II. Model Application

Publisher Website
Google Scholar
Abstract
To develop and compare methods that predict individual nicotine (NIC) clearance, which reflects CYP2A6 activity, using random saliva cotinine (COT) and trans 3′-hydroxycotinine (3HC) measurements. COT and 3HC saliva concentrations in smokers were simulated utilizing a mechanistic population pharmacokinetic model of NIC metabolism that was adapted from the one described in a companion paper. Four methods to predict NIC clearance using the metabolites concentrations were compared. The precision bias, and the fraction of predictions that are made with an absolute error below 25% were the performance measures evaluated. Four prediction methods were compared: (M1) reference method, an intercept slope model of the metabolite concentration ratios ([3HC]/[COT]) (M2) an intercept slope model of the natural logarithm of the metabolite ratios (M3) a spline of the logarithm of the metabolite ratios (M4) Maximal Posteriori Bayesian estimate of NIC clearance conditioned on the model, COT and 3HC concentrations. In addition, the effect of smoking patterns on the concentrations of COT and 3HC was evaluated. The precision, accuracy, and the fraction of predictions with an absolute error below 25%, were higher for methods M2–M4 compared to method M1. However, the differences between M2 and M4 were small. Additionally, smoking pattern did not affect the metabolite concentration profiles. Predicting NIC clearance using an intercept slope model of the natural logarithm of the ratio of 3HC to COT appears to be a relatively simple method that is better than using the metabolite ratio directly. This method has a bias of approximately −10%, precision of approximately 60%. The fraction of estimates below an absolute error of 25% is 43%. These results support use of M2 to estimate CYP2A6 activity in smokers in the clinical setting.

This publication has 15 references indexed in Scilit: