ESTRO 2024 - Abstract Book

S51 ESTRO 2024 disease site. Whilst there were some exceptions for specific trials, people aged 70 and above, women, and people living in the most deprived areas were underrepresented in the overall trial population. For example, whilst 58% of bladder cancer patients treated with radiotherapy are aged 80 or older, only 22% of participants in bladder cancer trials were in this age group. Ethnicity was only recorded in one trial; non-white ethnic groups were underrepresented. Not all factors were collected in every trial, and even when trials collected these factors, they often had incomplete datasets. These findings are consistent with a lack of consensus regarding the collection of demographic data from trial participants across the UK. We hope that this case study encourages those designing and delivering trials to assess inclusivity, and that the progression towards considering Equality, Diversity and Inclusion in clinical research results in the development of evidence-based interventions to increase the representation of underserved groups in radiotherapy trials. Invited Speaker

3362

Incorporating genetic information in dose-response relationships

Tiziana Rancati

Fondazione IRCCS Istituto Nazionale dei Tumori, Data Science, Milan, Italy

Abstract:

Traditionally, the prediction of radiotherapy-induced side effects has predominantly relied on clinical patient-specific risk factors and dosimetry. While invaluable, these factors offer a limited understanding of the interindividual variability observed in patient responses to radiotherapy. Both genetic makeup and environmental factors influence the risk of experiencing side effects from radiotherapy. Based on this hypothesis, recent efforts have tried to use patients' genetic information to improve models that predict the likelihood of normal tissue complications (NTCP models). When it comes to genetics, most people's sensitivity to radiation is a complex trait that varies along a continuous spectrum and isn't solely determined by one gene. For NTCP models to be helpful in clinics, they must include enough common genetic variants, each of which might only slightly affect the risk of complications. Studies have found many common genetic differences (SNPs) that can be added to create a risk score. This score combines the effects of different genetic variations linked to specific traits. Scientists are also realizing that interactions between these genetic differences, known as epistasis, affect how likely someone is to have particular health problems. Epistasis means that when certain genetic variations come together, they can significantly impact health more than if they were considered individually. Some studies have identified potential biomarkers correlating susceptibility to radiation-induced side effects. These genetic markers can potentially stratify patients into distinct risk groups, allowing for tailored treatment regimens and proactive management strategies. Moreover, incorporating genetic data into predictive models enables a comprehensive evaluation of multifactorial interactions underlying radiation toxicity, paving the way for more nuanced risk assessment algorithms.

Despite the immense potential, challenges persist in translating genetic discoveries into clinically actionable tools for radiotherapy optimization. Issues such as sample size limitations, genetic heterogeneity, and reproducibility concerns