Pharmacogenetic biomarkers for chemotherapy-induced adverse drug reactions

Sammanfattning: Cancer is a serious disease expected to be the world-leading cause of death in the 21st century. The use of harsh chemotherapies is motivated and accepted but, unfortunately, is often accompanied by severe toxicity and adverse drug reactions (ADRs). These occur because the classical chemotherapies’ common modes of action effectively kill and/or reduce the growth rate not only of tumour cells, but also of many other rapidly dividing healthy cells in the body. There are also considerable interindividual differences in ADRs, even between patients with similar cancers and disease stage treated with equal doses; some experience severe to life-threatening ADRs after one dose, leading to treatment delays, adjustments, or even discontinuation resulting in suboptimal treatment, while others remain unaffected through all treatment cycles. Being able to predict which patients are at high or low risk of ADRs, and to adjust doses accordingly before treatment, would probably decrease toxicity and patient suffering while also increasing treatment tolerability and effects. In this thesis, we have used next-generation sequencing (NGS) and bioinformatics for the prediction of myelosuppressive ADRs in lung and ovarian cancer patients treated with gemcitabine/carboplatin and paclitaxel/carboplatin.Paper I shows that ABCB1 and CYP2C8 genotypes have small effects inadequate for stratification of paclitaxel/carboplatin toxicity. This supports the transition to whole-exome sequencing (WES) and whole-genome sequencing (WGS). Papers II and IV, respectively, use WES and WGS, and demonstrate that genetic variation in or around genes involved in blood cell regulation and proliferation, or genes differentially expressed at chemotherapy exposure, can be used in polygenic prediction models for stratification of gemcitabine/carboplatininduced myelosuppression. Paper III reassuringly shows that WES and WGS are concordant and mostly yield comparable genotypes across the exome. Paper V proves that single-cell RNA sequencing of hematopoietic stem cells is a feasible method for elucidating differential transcriptional effects induced as a response to in vitro chemotherapy treatment.In conclusion, our results supports the transition to genome-wide approaches using WES, WGS, and RNA sequencing to establish polygenic models that combine effects of multiple pharmacogenetic biomarkers for predicting chemotherapy-induced ADRs. This approach could be applied to improve risk stratification and our understanding of toxicity and ADRs related to other drugs and diseases. We hope that our myelosuppression prediction models can be refined and validated to facilitate personalized treatments, leading to increased patient wellbeing and quality of life.