Prediction of treatment response to antipsychotic drugs for precision medicine approach to schizophrenia: randomized trials and multiomics analysis

Abstract

Background: Choosing the appropriate antipsychotic drug (APD) treatment for patients with schizophrenia (SCZ) can be challenging, as the treatment response to APD is highly variable and difficult to predict due to the lack of effective biomarkers. Previous studies have indicated the association between treatment response and genetic and epigenetic factors, but no effective biomarkers have been identified. Hence, further research is imperative to enhance precision medicine in SCZ treatment.

Methods: Participants with SCZ were recruited from two randomized trials. The discovery cohort was recruited from the CAPOC trial (n = 2307) involved 6 weeks of treatment and equally randomized the participants to the Olanzapine, Risperidone, Quetiapine, Aripiprazole, Ziprasidone, and Haloperidol/Perphenazine (subsequently equally assigned to one or the other) groups. The external validation cohort was recruited from the CAPEC trial (n = 1379), which involved 8 weeks of treatment and equally randomized the participants to the Olanzapine, Risperidone, and Aripiprazole groups. Additionally, healthy controls (n = 275) from the local community were utilized as a genetic/epigenetic reference. The genetic and epigenetic (DNA methylation) risks of SCZ were assessed using the polygenic risk score (PRS) and polymethylation score, respectively. The study also examined the genetic-epigenetic interactions with treatment response through differential methylation analysis, methylation quantitative trait loci, colocalization, and promoter-anchored chromatin interaction. Machine learning was used to develop a prediction model for treatment response, which was evaluated for accuracy and clinical benefit using the area under curve (AUC) for classification, R² for regression, and decision curve analysis.

Results: Six risk genes for SCZ (LINC01795, DDHD2, SBNO1, KCNG2, SEMA7A, and RUFY1) involved in cortical morphology were identified as having a genetic-epigenetic interaction associated with treatment response. The developed and externally validated prediction model, which incorporated clinical information, PRS, genetic risk score (GRS), and proxy methylation level (proxyDNAm), demonstrated positive benefits for a wide range of patients receiving different APDs, regardless of sex [discovery cohort: AUC = 0.874 (95% CI 0.867-0.881), R² = 0.478; external validation cohort: AUC = 0.851 (95% CI 0.841-0.861), R² = 0.507].

Conclusions: This study presents a promising precision medicine approach to evaluate treatment response, which has the potential to aid clinicians in making informed decisions about APD treatment for patients with SCZ. Trial registration Chinese Clinical Trial Registry ( https://www.chictr.org.cn/ ), 18. Aug 2009 retrospectively registered: CAPOC-ChiCTR-RNC-09000521 ( https://www.chictr.org.cn/showproj.aspx?proj=9014 ), CAPEC-ChiCTR-RNC-09000522 ( https://www.chictr.org.cn/showproj.aspx?proj=9013 ).

Keywords: Antipsychotic drug; Epigenetics; Genetics; Prediction model; Schizophrenia; Treatment response.

Fig. 1

Research flow diagram. The flow diagrams illustrate the trial profiles and research design. a This study used participants from two randomized trials: CAPOC and CAPEC. The flow diagram describes the detailed profile of the trials. b Multiomics analyses were conducted to investigate the relationship between genetic/epigenetic risks of SCZ and treatment response to APDs, and developed a prediction model for treatment response in the discovery cohort (CAPOC). We used an external validation cohort (CAPEC) to validate the prediction models. CAPOC Chinese Antipsychotics Pharmacogenetics Consortium; CAPEC Chinese Antipsychotics Pharmacogenomics Consortium

Fig. 2

Multiomics analyses. Genetic and epigenetic factors reflecting treatment response were investigated by meQTL and DMR, and the results were refined using three different approaches. a Upon comparison of meQTLs, risk-DMRs, and RES-DMRs, 324 genes were identified as ASM genes associated with SCZ risk and treatment response. b Within the ASM genes, colocalization identified four signals, including rs11125746 in LINC01795, rs12674515 in DDHD2, rs28759130 in SBNO1, and rs498541 in KCNG2. c PAI identified 14 genes showing significant differences in overall PAI strength under the case‒control condition and non-RES versus RES condition. Among them, RUFY1 displayed brain‒blood consistency in methylation, transcription, and chromatin interaction, along with a significant difference in PAI strength. d EWAS identified one genome-wide significant (P_adj < 1 × 10^–8) signal located in the SEMA7A gene. Solid and dashed gray lines represent genome-wide and suggestive significance, respectively. meQTL. Methylation quantitative trait loci; DMR. Differentially methylated region; PAI. Promoter-anchored chromatin interaction; SCZ. Schizophrenia; RES. Response; ASM. Allele-specific methylated; GWAS. Genome-wide association study; PP4. Posterior probability for a shared signal; SMR. Summary-based Mendelian randomization; LINC01795. Long intergenic non-protein coding RNA 1795; DDHD2. DDHD domain containing 2; SBNO1. Strawberry notch homolog 1; KCNG2. Potassium voltage-gated channel modifier subfamily G member 2; RUFY1. RUN and FYVE domain containing 1; MIR885. microRNA 885, BEGAIN brain enriched guanylate kinase associated; SLC7A7. Solute carrier family 7 member 7; KLF5. KLF transcription factor 5; SEMA7A. Semaphorin 7A

Fig. 3

Performance of the optimal prediction model for treatment response in the discovery and validation cohorts. Visualization of regression performance of four RES-prediction models in a discovery cohort and b external validation cohort. The solid line represents the linear relationship between the scaled prediction value and scaled PANSS reduction rate. c, d illustrate the AUCs of four RES-prediction models. Solid lines in different colors represent receiver operating curves for different RES-prediction models, and the ribbons in different colors represent the confidence intervals of different models in the c discovery cohort and d external validation cohort. e Decision curves for four RES-prediction models in the discovery cohort (solid line) and external validation cohort (dashed line); lines in different colors represent different models. PANSS. Positive and negative syndrome scale; RES. Response; AUC. Area under the curve; C + P. Clinical information + PRS; C + G. Clinical information + GRS; C + M. Clinical information + proxyDNAm; C + PGM. Clinical information + PRS + GRS + proxyDNAm; PRS. Polygenic risk score; GRS. Genetic risk score