Impact of Population Pharmacogenomics on Cisplatin-Induced Neurotoxicities in Testicular Cancer Survivors

Abstract

Background: Cisplatin is a commonly used chemotherapeutic across numerous cancer types that can cause neurotoxicities in patients, including peripheral sensory neuropathy, tinnitus, hearing loss, and vertigo.

Objective: We aimed to evaluate, for the first time, how genetic ancestry impacts cisplatin-induced neurotoxicities and if disparities are related to population differences in allele frequency.

Methods: In a cohort of cisplatin-treated testicular cancer survivors, relationships between genetic ancestry and neurotoxicities, medications, and lifestyle factors were assessed using logistic regression and Kruskal-Wallis tests and multiple pairwise comparisons using the Wilcoxon rank-sum test (Benjamini-Hochberg adjustment). Associations between single nucleotide polymorphism (SNP) genotypes and neurotoxicities with significant inter-population disparities were calculated to identify independent, functional variants with population allele frequency differentiation associated with toxicities.

Results: Following four cycles of cisplatin-based chemotherapy, African ancestry survivors were significantly more likely to have neuropathy and vertigo versus European and Asian-axis ancestry survivors, although Asian axis survivors were significantly younger at evaluation than other ancestries. Following filtering for population allele frequency differentiation, functional relevance, and independence, 19,992 SNPs were tested for association with toxicities. Although none passed the Bonferroni threshold, two and four SNPs were associated with neuropathy and vertigo, respectively, at suggestively significant p < 1.0 × 10^-4. For neuropathy, rs34904346 (p = 2.0 × 10^-5) was an expression quantitative trait locus (eQTL) for RNF24 in nerve tissue, with three other RNF24 eQTLs associated with neuropathy (p < 0.01). For vertigo, rs3777909 (p = 3.1 × 10^-5) was an eQTL for MFSD4B in nerve and REV3L in brain tissue, along with three other eQTLs for MFSD4B and four for REV3L associated with vertigo (p < 0.05). In silico, higher MFSD4B and REV3L expression in cancer cell lines were associated with significantly greater cisplatin sensitivity.

Conclusion: African ancestry was associated with increased cisplatin-induced peripheral sensory neuropathy and vertigo versus European ancestry. Population allele frequency differences and expression levels of RNF24, MFSD4B, and REV3L were potentially implicated.

Keywords: cisplatin; disparities; genetic variants; neurotoxicity; pharmacogenomics; vertigo.

FIGURE 1

Association between genetic ancestry and toxicity phenotypes. (A) Association between PSN and ancestry for survivors receiving four cycles of cisplatin. (B) Association between vertigo and ancestry for survivors receiving four cycles of cisplatin. (C) Association between tinnitus and ancestry for survivors receiving four cycles of cisplatin. (D) Association between hearing thresholds and ancestry for survivors receiving four cycles of cisplatin. (E) Association between self‐reported health and ancestry for all survivors across cisplatin doses. AFRAFR, African ancestry; ASN axis, Asian axis ancestry; EUR, European ancestry; PSN, Peripheral Sensory Neuropathy.

FIGURE 2

Filtering SNPs for population pharmacogenomic analysis. (A) Genome‐wide approach to select SNPs for genotype–phenotype analysis between independent, functional, ancestry‐informative SNPs and neurotoxicity phenotype. (B) Candidate gene approach to select cisplatin‐associated SNPs for genotype–phenotype analysis between independent, functional, ancestry‐informative, cisplatin‐associated SNPs and neurotoxicity phenotype. F _ST, fixation index; GTEx, Genotype‐Tissue Expression Project; LD, linkage disequilibrium; SNP, Single Nucleotide Polymorphism; eQTL, expression quantitative loci; sQTL, splicing quantitative trait loci.

FIGURE 3

eQTLs for RNF24 are associated with peripheral sensory neuropathy. (A) T allele frequency for rs34904346 in testicular cancer survivors for AFRAFR, AFR axis, ASN axis, and EUR populations. (B) Normalized gene expression and p‐value in GTEx for eQTLs for RNF24 in tibial nerve tissue (V8 data). (C) Normalized gene expression and p‐value in GTEx for eQTLs for RNF24 in brain cortex tissue (V8 data). (D) Heatmap matrix of pairwise linkage disequilibrium statistics (R ² labeled) for eQTLs for RNF24 with association between SNP genotype and PSN phenotype (p < 0.01 for all). (B, C) were generated using GTEx V8 data from the GTEx portal between 10/2023 and 12/2023, with visualization created in 05/2024. (D) was adapted from NCI LDmatrix. eQTL, expression quantitative loci; AFRAFR, African ancestry; AFR axis, African axis ancestry; ASN axis, Asian axis ancestry; EUR, European ancestry; GTEx, Genotype‐Tissue Expression Project; SNP, single nucleotide polymorphism.

FIGURE 4

eQTLs for MFSD4B and REV3L are associated with vertigo. (A) A allele frequency for rs3777909 in testicular cancer survivors for AFRAFR, AFR axis, ASN axis, and EUR populations. (B) Normalized gene expression and p‐value in GTEx for eQTLs for MFSD4B in tibial nerve tissue (V8 data). (C) Normalized gene expression and p‐value in GTEx for eQTLs for REV3L in brain cerebellum tissue (V8 data). (D) Heatmap matrix of pairwise linkage disequilibrium statistics (R ² labeled) for eQTLs for MFSD4B/REV3L with association between SNP genotype and vertigo phenotype (p < 0.05 for all). (E) Correlation between MFSD4B expression in cancer cell lines and cisplatin AUC. (F) Correlation between REV3L expression in cancer cell lines and cisplatin AUC. (B, C) were generated using GTEx V8 data from the GTEx portal between 10/2023 and 12/2023, with visualization created in 05/2024. (D) was adapted from NCI LDmatrix. AFRAFR, African ancestry; AFR axis, African axis ancestry; ASN axis, Asian axis ancestry; EUR, European ancestry; GTEx, Genotype‐Tissue Expression Project; SNP, single nucleotide polymorphism.