RESUMO
Objective To establish the population pharmacokinetics(PPK)model of Lamotrigine(LTG)in children with epilepsy in China for promoting individualized dosage regimen. Methods The sparse data of LTG serum concentrations from 60 pediatric patients with epilepsy were collected. One hundred and fourteen serum concentration points were divided into LTG+valproic acid (VPA) group(n=56),LTG+enzymatic inducer(E1)group(n=26),LTG+EI+VPA group(n=16)and single LTG group(n=16).The serum drug concentrations were the clinical routinely tested steadv state concentrations.The LTG PPK parameters were calculated using the non-parametric expectation maximization(NPEM) Program of USC*PACK software,and then a PPK model was established. Based on this model,LTG serum concentrations were predicted with Bayesian fitting program of USC*PACK software.Mean prediction error(MPE)and mean squared prediction error(MSPE) were calculated to evaluate the accuracy and precision of the concentration prediction and to valid the PPK model.Results The greatest likelihood was-192.87.Optimum PPK parameters were:Ka=(1.97 1.66)h~(-1);Vs=(1.07±0.89)L/kg;Kel=(0.05±0.05)h~(-1).The linear regression function Y_(OBS)=-0.09+1.05 Y_(PRED)(R~2=0.98,P<0.001),and determination of coefficient was 0.98.MPE was-0.16 g/mL,and MSPE was 0.28(μg/mL)~2.Conclusion A PPK model of LTG in children with epilepsy in China can be successfully established using the USC*PACK software, based on which LTG serum concentrations can be predicted accurately with a Bayesian approach.