Identification of genetic factors predisposing to chemotherapy and immunotherapy toxicity

Chin, Ik Shin ORCID: 0000-0002-7659-8753 (2024). Identification of genetic factors predisposing to chemotherapy and immunotherapy toxicity. University of Birmingham. Ph.D.

Chin2024PhD.pdf
Text
Restricted to Repository staff only until 31 July 2025.
Available under License All rights reserved.
Download (8MB) | Request a copy

Abstract

Introduction
Serious Common Terminology Criteria for Adverse Events (CTCAE) grade 3-4 toxicities can cause substantial morbidity and negatively impact patients’ quality of life. The ability to predict patients’ toxicity risk from systemic anti-cancer therapy can help personalise treatment. Germline genetic variation has been shown to be a key determinant of toxicity risk. My thesis aimed to identify genetic variants predisposing towards chemotherapy and immune checkpoint inhibitor (ICI) toxicity. I performed genome wide association studies (GWAS) and fine mapping studies of both fluoropyrimidine-induced toxicity and ICI-induced immune related adverse events (irAEs). I also analysed candidate variants and performed functional analysis for one locus.
Methods
Per cycle CTCAE graded toxicity data was obtained from 6459 colorectal cancer patients who received fluoropyrimidine chemotherapy from four clinical trials. CTCAE graded irAE were analysed from a discovery cohort of 599 patients on ICIs. 103 cases of serious irAE and 2497 controls from UK Biobank formed a replication cohort. Tagging single nucleotide polymorphism (SNP) arrays were used for genotyping and genome-wide imputation was performed. Association testing and meta-analyses were conducted for common capecitabine and ICI induced toxicity variables. Conditional and joint association analysis was done to identify top independent SNPs that were tested in polygenic risk score (PRS) analyses. Preliminary functional experiments were also conducted for a gene implicated in fluoropyrimidine induced cardiotoxicity.
Results
In relation to toxicities induced by fluoropyrimidines, I identified four SNPs associated with early CTCAE grade 3-4 capecitabine toxicity at GWAS significance (P<5x10-8) and a further two novel SNPs associated with cardiovascular toxicity under the additive genetic model. I also generated data that did not validate the current literature that MIR27A variant rs895819 is only associated with toxicity in DPYD no function/low function allele carriers.
In relation to ICI-induced irAEs, I identified eight independent SNPs associated with any serious irAE at suggestive significance (P<5x10-6) under the additive model. Seven and 12 SNPs were associated with ≥grade 3 GI and hepatitis irAEs at P<5x10-6. A diabetes PRS was significantly associated with hyperglycaemic irAEs events. Four SNPs associated with ≥grade 3 endocrine, GI and hepatitis irAEs were replicated in the validation phase at nominal significance (P<0.05), although none remained genome-wide significant following meta-analysis. SNPs associated with any serious irAE, immune-related hepatitis and a novel toxicity score is mapped to the ZNF701 gene at suggestive significance.
Conclusion
We identified potential genetic variants that can contribute towards predictive risk models of toxicity. Use of these markers in practice should be balanced against markers of efficacy to maximise treatment benefit.

Type of Work:

Thesis (Doctorates > Ph.D.)

Award Type:

Doctorates > Ph.D.

Supervisor(s):