A single-cell atlas linking intratumoral states to therapeutic vulnerabilities across cancers

Abstract

Intratumoral heterogeneity (ITH) is a major determinant of therapeutic failure, yet its impact on drug response across cancers remains incompletely understood. Here, we present the Therapeutic Cancer Cell Atlas (TCCA), a pan-cancer single-cell resource integrating [~]1.8 million transcriptomes from 537 patients and 183 cancer cell lines spanning 34 tumor types. By combining single-cell transcriptomics with copy-number alteration inference and computational drug-response prediction, we systematically map therapeutic heterogeneity at subclonal resolution across cancers. Using this framework, we identify ten recurrent therapeutic clusters that capture conserved and context-specific drug vulnerabilities across tumor lineages. Notably, therapeutic heterogeneity is largely decoupled from genomic and transcriptomic diversity and instead arises from distinct functional transcriptional programs and tumor microenvironment (TME) states. Integration with transcriptional metaprograms and TME archetypes reveals how stress responses, proliferative states, lineage programs, and immune context shape drug sensitivity beyond tissue of origin. We further demonstrate the translational relevance of TCCA by linking therapeutic clusters to patient outcomes and validating predicted vulnerabilities using pharmacogenomic datasets, including clinically actionable examples in aggressive tumor subtypes. Together, TCCA provides a multidimensional atlas connecting subclonal states, microenvironmental context, and drug response, offering a scalable framework to guide therapeutic prioritization, drug repurposing, and combination strategies in precision oncology.