Blood Gene Expression Profile Predicts Response to Antipsychotics

Area of Science:

• Pharmacogenomics
• Neuroscience
• Molecular Psychiatry

Background:

• Antipsychotic medications are widely prescribed for schizophrenia, yet exhibit significant variability in efficacy among individuals.
• Current methods lack the ability to predict patient response to specific antipsychotic treatments.
• Inter-individual differences in drug response necessitate predictive biomarkers.

Purpose of the Study:

• To identify a predictive biomarker for antipsychotic drug response in schizophrenia patients.
• To develop a methodology for predicting clinical efficacy of antipsychotics before treatment initiation.
• To enable personalized antipsychotic selection based on individual biological profiles.

Main Methods:

• Transcriptome sequencing of blood mRNA from antipsychotic-naïve schizophrenia patients categorized by treatment response (best vs. worst responders).
• Differential gene expression analysis to identify genes associated with clinical response.
• Random Forests machine learning algorithm employed to build a predictive model using gene expression data.

Main Results:

• Identified 130 differentially expressed genes (P_adj < 0.01) between best and worst responders.
• Developed a four-gene blood expression profile with high predictive accuracy (0.83 cross-validation accuracy, 0.97 AUC).
• Demonstrated the capability of this gene profile to predict antipsychotic medication response.

Conclusions:

• A four-gene signature in blood can predict antipsychotic treatment response in schizophrenia patients.
• This predictive profile facilitates the selection of the most effective antipsychotic for individual patients.
• This approach represents a significant step towards personalized medicine in psychiatric pharmacotherapy.