The diagnostic landscape of bone and soft tissue sarcoma: the pan-sarcoma era and AI-driven classification

Abstract

Bone and soft tissue sarcomas are a diagnostic challenge due to their histological complexity, variety, and mutational heterogeneity. This has led to continued refinements in classification and diagnostic criteria, presented and soon to be updated by the WHO. This commentary features articles published in this journal and elsewhere that have ultilised pan-sarcoma datasets and artificial intelligence-based classifiers. A notable entry is the DKFZ Sarcoma Classifier, which is based on methylation profiles and available for use via a commercial platform. This work has created hope that one day a comprehensive sarcoma classifier will be available, though it is also a reminder of the many challenges apparent in studying rare cancer types.