Leveraging Digital Twins for Patient Stratification and Treatment Optimization in geriatric oncology A Breast Cancer Multivariate Clustering Analysis (Preprint)

Abstract

<sec> <title>UNSTRUCTURED</title> Purpose: Define optimal adjuvant therapeutic strategies for elderly breast cancer patients remains a challenge, given that this population is often overlooked and underserved in clinical research and decision-making tools. Patients and Methods: Data from women aged 70+ with HER2-negative early-stage breast cancer treated at the French Léon Bérard Cancer Center from 1997 to 2016 was retrospectively analyzed. Generative Manifold learning and machine learning algorithms were employed to understand complex data relationships and develop predictive models. Digital twins were synthetized for every breast cancer patient to establish their personalized normative values of biological characteristics. Results: From 1229 initial patients, 793 were included after data refinement. The unsupervised machine learning framework unveiled 6 clusters in the population, estimated chemotherapy benefits, and emphasized the overall biological profile over individual factors like comorbidities. The generative manifold learning model demonstrated high predictive efficacy, with mean AUC scores of 0.81 and 0.76 for Random Forest Classification and Support Vector Classifier, respectively. Conclusion: The study presents a novel prognostic tool for elderly breast cancer patients, enhancing treatment guidance through advanced AI techniques. This approach provides a nuanced understanding of disease dynamics and therapeutic strategies, underscoring the importance of tailored healthcare in oncology. </sec>

Ähnliche Arbeiten