Network-based clustering unveils interconnected landscapes of genomic and clinical features across myeloid malignancies

Abstract

Myeloid malignancies exhibit considerable heterogeneity with overlapping clinical and genetic features among different subtypes. Current classification schemes, predominantly based on clinical features, fall short of capturing the complex genomic landscapes of these malignancies. Here, we present a data-driven approach that integrates mutational features and clinical covariates within networks of their probabilistic relationships, enabling the discovery of de novo cancer subgroups. In a cohort of 1323 patients across acute myeloid leukemia, myelodysplastic syndromes, chronic myelomonocytic leukemia and myeloproliferative neoplasms, we identified novel subgroups that outperform established risk classifications in prognostic accuracy. Our findings suggest that mutational patterns are often shared across different types of myeloid malignancies, with distinct subtypes potentially representing evolutionary stages en route to leukemia. Within the novel subgroups, our integrative method discerns unique patterns combining genomic and clinical features to provide a comprehensive view of the multifaceted genomic and clinical landscape of myeloid malignancies. This in turn may guide the development of targeted therapeutic strategies and offers a pathway to enhanced patient stratification.