Little has been reported on the factors, genetic or other, that underlie the variability in individual response, particularly for
autism. In this study we simultaneously explored the effects of multiple candidate genes on clinical improvement and occurrence of
adverse drug reactions, in 45 autistic patients who received monotherapy with
risperidone up to 1 year. Candidate genes involved in the pharmacokinetics (
CYP2D6 and ABCB1) and pharmacodynamics (HTR2A, HTR2C, DRD2, DRD3, HTR6) of the
drug, and the
brain-derived neurotrophic factor (
BDNF) gene, were analysed. Using the generalized estimating equation method these genes were tested for association with
drug efficacy, assessed with the
Autism Treatment Evaluation Checklist, and with safety and tolerability measures, such as
prolactin levels, body mass index (BMI), waist circumference and neurological adverse effects, including extrapyramidal movements. Our results confirm that
risperidone therapy was very effective in reducing some
autism symptoms and caused few serious adverse effects. After adjusting for confounding factors, the HTR2A c.-1438G>A, DRD3 Ser9Gly, HTR2C c.995G>A and ABCB1 1236C>T polymorphisms were predictors for clinical improvement with
risperidone therapy. The HTR2A c.-1438G>A, HTR2C c.68G>C (p.C33S), HTR6 c.7154-2542C>T and
BDNF c.196G>A (p.V66M) polymorphisms influenced
prolactin elevation. HTR2C c.68G>C and
CYP2D6 polymorphisms were associated with
risperidone-induced increase in BMI or waist circumference. We thus identified for the first time several genes implicated in
risperidone efficacy and safety in
autism patients. Although association results require replication, given the small sample size, the study makes a preliminary contribution to the personalized
therapy of
risperidone in
autism.