Abstract C051: Impact of population pharmacogenomics on cisplatin-induced neurotoxicities
Swetha Nakshatri,Paul C. Dinh,Darren R. Feldman,Robert J. Hamilton,Chunkit Fung,David J. Vaughn,Christian Kollmannsberger,Robert Huddart,Lawrence H. Einhorn,Nancy J. Cox,Lois B. Travis,Eileen Dolan
DOI: https://doi.org/10.1158/1538-7755.disp24-c051
2024-09-23
Cancer Epidemiology Biomarkers & Prevention
Abstract:Background: Studies have shown that taxane-treated breast cancer patients with African (AFR) ancestry experience worse peripheral sensory neuropathy (PSN), while African American vincristine-treated leukemia patients have less PSN than Europeans (EUR). We hypothesized that ancestral disparities may exist for cisplatin-induced neurotoxicities and may relate to differences in allele frequency and resultant gene expression of SNPs associated with these toxicities. Methods: In our Platinum Study of 1680 genotyped testicular cancer survivors (TCS), multidimensional scaling scores were anchored by the 1000 Genomes population to classify geographic ancestry. PSN (EORTC-CIPN20) and other toxicities plus clinical and lifestyle variables were examined for associations between geographic ancestry and phenotypes using logistic regression and Wilcoxon rank sum test. Genotyped SNPs with Fst > 0.25 in 1000 Genomes were identified and evaluated with GTEx data to find expression or splicing quantitative trait loci in nerve/brain tissue or for cisplatin-associated candidate genes, across all tissues. LDmatrix was used to find linkage disequilibrium. If a block of SNPs had R2 > 0.75, one independent SNP was kept for association analysis. Logistic regression with age at questionnaire completion and 10 genetic principal components as covariates was used to calculate the association between genotype and toxicity. Results: In an analysis of patients who received 400-450 mg/m2 of cisplatin including EUR (n=681), Asian (ASN) axis (n=44) and AFR (n=13), the odds ratio (OR) for any neuropathy vs. none in the AFR vs. EUR was 7.8 (P=0.049) and in the AFR vs. ASN axis (including East Asian 1000 Genomes populations) was 10.0 (P=0.034). TCS with AFR ancestry had significantly more vertigo. There were 19,874 SNPs for analysis from the genome-wide approach and 118 SNPs from the candidate gene approach. While associations between genotype and toxicity did not reach Bonferroni threshold, some SNPs had p-values in the 10-5 range. For PSN, one SNP had increased frequency of a risk allele in the AFR population while the other had increased frequency of a protective allele, while for vertigo, four SNPs had increased frequency of risk alleles in the AFR population. Of interest was rs34904346, in which the TT vs. AA genotype had an OR of 1.9 (p=2x10-5) for any vs. no PSN, with the T allele having 90% frequency in AFR ancestry patients versus ∼60% in other populations. This SNP and three other independent SNPs associated with PSN incidence (p∼10-3) are eQTLs for RNF24 in nerve tissue, which interacts with membrane proteins that regulate intracellular calcium levels. Conclusions: We show preliminary evidence for the role of differing risk allele frequencies across populations in explaining disparities in cisplatin-induced PSN and vertigo, likely in combination with structural and environmental factors. Genotyping risk alleles could customize dosing, treatment, and post-treatment monitoring to benefit all cisplatin-treated patients. Citation Format: Swetha Nakshatri, Paul C. Dinh, Darren R. Feldman, Robert J. Hamilton, Chunkit Fung, David J. Vaughn, Christian Kollmannsberger, Robert Huddart, Lawrence H. Einhorn, Nancy J. Cox, Lois B. Travis, Eileen Dolan. Impact of population pharmacogenomics on cisplatin-induced neurotoxicities [abstract]. In: Proceedings of the 17th AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2024 Sep 21-24; Los Angeles, CA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2024;33(9 Suppl) nr C051.
oncology,public, environmental & occupational health