Leveraging large-scale datasets and single cell omics data to develop a polygenic score for cisplatin-induced...

Area of Science:

• Pharmacogenomics
• Genetics
• Ototoxicity Research

Background:

• Cisplatin-induced ototoxicity (CIO) causes irreversible hearing loss and is a significant adverse effect of cisplatin chemotherapy.
• Genetic factors and polygenicity contribute to CIO risk, necessitating advanced predictive tools.
• Polygenic scores (PGS) integrating multiple genetic variants show promise for identifying individuals at high risk for CIO.

Purpose of the Study:

• To develop and evaluate a biologically informed polygenic score (PGS_CIO) for predicting cisplatin-induced ototoxicity (CIO).
• To compare the predictive performance of PGS_CIO against a general hearing loss polygenic score (PGS_HL).
• To leverage large-scale hearing loss GWAS and single-cell omics data for improved pharmacogenomic trait prediction.

Main Methods:

• Developed two polygenic scores: PGS_HL using hearing loss GWAS data and PGS_CIO incorporating biologically relevant variants from cisplatin-treated murine inner ear snRNA-seq data.
• PGS_CIO was refined using variants mapped to differentially expressed genes in specific cochlear cell populations.
• Evaluated the association of both PGSs with CIO in independent patient cohorts.

Main Results:

• The biologically informed PGS_CIO showed significantly superior performance in predicting CIO in the target cohort (P = 5.54 × 10^-5) compared to PGS_HL (P = 2.93 × 10^-3).
• PGS_CIO was significantly associated with CIO in the test cohort (P = 0.04), whereas PGS_HL did not show a significant association (P = 0.52).
• Identified specific cochlear cell types potentially involved in CIO pathogenesis.

Conclusions:

• The study presents the first PGS specifically developed for CIO, utilizing a novel integration of GWAS and single-cell omics data.
• The developed PGS_CIO offers a promising tool for personalized medicine, potentially enabling tailored therapeutic strategies to prevent ototoxicity.
• This approach advances the development of PGSs for pharmacogenomic traits and identifies key cellular players in CIO.