Abstracts

Rationale: Mesial temporal lobe epilepsy (MTLE) is the most common form of adult epilepsy. Currently, the only characteristic, which seems to predict poor response to clinical treatment in these patients, is the presence of hippocampal sclerosis (HS). Studies showed that single nucleotide polymorphisms (SNPs) in genes encoding drug transporter and metabolism proteins could influence response to drug therapy. The objective of the study was to evaluate whether combining information from clinical variables and SNPs in candidate genes could improve the discriminatory accuracy of predicting response to drug therapy in patients with MTLE.Methods: We evaluated 73 patients with MTLE who were responsive to antiepileptic drug (AED) therapy and 163 patients refractory to drug therapy. We genotyped 119 SNPs within the ABCB1, ABCC2, CYP1A1, CYP1A2, CYP1B1, CYP2C9, CYP2C19, CYP2D6, CYP2E1, CYP3A4, and CYP3A5 genes. We assessed a first scenario using only clinical variables and a second one including SNP information. We used random forests combined with leave-one-out cross-validation to identify the best predictive model in each scenario and compared their discriminatory accuracies using the area under the curve statistic, by R software. Additionally, we built a variable importance plot to present the set of most relevant predictors on the best model.Results: The selected best model included the presence of HS and 53 SNPs among all genes evaluated. Additionally, including SNPs in the model improved the discriminatory accuracy from 0.5865 to 0.8244.Conclusions: To the best of our knowledge, this study demonstrated for the first time that genetic information improves the prediction of response to drug treatment in patients with MTLE. This may allow patients more likely to develop resistance to AED treatment to have a timely referral for epilepsy surgery. Supported by CEPID-BRAINN FAPESP, São Paulo, Brazil.