PhD Scientific Days 2025

Pharmaceutical Sciences and Health Technologies III.

Statistical methods for processing family history - potential risk calculating methods for individuals with family history of cancer

Name of the presenter

Székely Flóra

Institute/workplace of the presenter

Semmelweis University Institute of Genomic Medicine and Rare Disorders

Authors

Dr. Flóra Székely¹

1: Semmelweis University Institute of Genomic Medicine and Rare Disorders

Text of the abstract

Introduction: The systematic collection and evaluation of family medical history may open new avenues of research, particularly in diseases where familial patterns play a significant role - such as hereditary cancer syndromes. Harmonizing individual clinical and molecular features within a unified data model can provide quantitative input for clinical decision-making.
Aims: Through the Semmelweis Federated Data Warehouse project, - which aims to connect clinical, laboratory, and genomic profiles by identifying mutual synergies - we built a structured database enriched with family history information obtained from the medical records of patients undergoing oncogenetic counseling. Our goal was to create a transparent, analyzable dataset that could reveal potential inheritance patterns and quantify familial aggregation of disease.
Methods: Within enrolled 140 patients, detailed family histories were extracted from their medical documentation, and by using Human Phenotype Ontology (HPO) terms, we annotated and encoded phenotypic data. The resulting database encompassed approximately 2,500 family members. Mathematical models was applied to estimate cancer risk scores based on familial data, including the number of known family members, their degree of relatedness, and the presence or absence of specific cancer types across their lifetimes.
Results: We compared the two obtained risk scores with each other, and in cases where a pathogenic genetic mutation was known, with other widely used cancer risk score calculators.
Conclusion: Transforming family history into a standardized variable holds significant potential, particularly for enabling clearer comparative analysis of familial disease burdens. Rich familial history data also support gene-specific penetrance estimation, which could help in clinical decision-making.
Funding: Supported by the TKP2021-NVA-15 grant of the Thematic Area Excellence Program of the National Research, Development and Innovation Office.