Determining the clinically optimal dose in methadone maintenance therapy (MMT) is a time-consuming procedure, which considers clinical signs and symptoms. To perform a quantitative trait locus association for identifying genetic variants for MMT dosage that underlie heroin addiction and methadone metabolism and then integrate several genotypic and phenotypic factors are potential predictors for clinically optimal MMT dose for personalized prescription. In total, 316 heroin-dependent patients undergoing MMT were recruited at the Addiction Center of the China Medical University Hospital. A multinomial logistic regression model was used to assess associations between genetic polymorphisms and MMT dosing. The data were randomly separated into training and testing sets. In order to enhance the prediction accuracy and the reliability of the prediction model, we used areas under the receiver operating characteristic curves to evaluate optimal MMT dose in both training and testing sets. Four single nucleotide polymorphisms, namely rs806368 inCNR1, s1386493 inTPH2, s16974799 inCYP2B6, and rs2229205 inOPRL1, were significantly associated with the maximum MMT dose (P < .05). The genetic risk score (GRS) was associated with maximum MMT dose, and after adjustments for age, sex, and body mass index, the GRS remained independently associated with the maximum MMT dose. The area under the receiver operating characteristic curve of the combined GRS and craving score was 0.77 for maximum MMT dose, with 75% sensitivity and 60% specificity. Integrating the GRS and craving scores may be useful in the evaluation of individual MMT dose requirements at treatment initiation. Optimal dose prediction allows clinicians to tailor MMT to each patient´s needs.
Palabras clave:
Puntuación de riesgo genético; Tratamiento del mantenimiento con metadona; Predicción; Locus de un carácter cuantitativo