189 Head and neck squamous cell carcinomas (HNSCC) are the seventh most prevalent form of cancer and are associated with human papilloma virus infection (HPV-positive) or with tobacco and alcohol use (HPV-negative). HPV-negative HNSCCs have a high recurrence rate, and patients’ responses to treatment vary greatly because tumours and their microenvironments are highly heterogeneous. In a new study from Boston University Chobanian & Avedisian School of Medicine, researchers have assembled and publicly released a large single-cell atlas that maps the many cell types in head and neck tumours and shows how specific cell mixtures and interactions relate to tumour behaviour. These findings appear online in the journal Communications Medicine. “By mapping the diverse cells and communication networks inside head & neck cancer at such an unprecedented scale, we aim to support the identification of the cellular drivers of disease progression and treatment failure. This atlas provides publicly available roadmaps to guide new diagnostics and therapies that could improve survival and quality of life across head and neck cancer,” explains corresponding author Stefano Monti, PhD, professor of medicine at the school. The team analysed samples from 54 patients, totalling approximately 232,000 cells, by gathering six published single-cell RNA-sequence datasets of treatment-naive, HPV-negative HNSCC tumours. They filtered out low-quality cells and doublets, normalised and annotated the data with reference databases and marker genes, and integrated the datasets into a single atlas. Analyses included clustering and detailed cell-type annotation, signature scoring (e.g., cytotoxic vs dysfunctional T cells), hierarchical taxonomy discovery, cell–cell communication inference, and malignancy calling. They also tested associations between cell states or types and clinical variables (stage, sex) using mixed models and compositional analysis, and validated spatial localisation using an external dataset. Currently, there are three approved immune checkpoint inhibitors for treating HNSCC. The atlas highlights candidate targets for improving immunotherapy responses and for overcoming chemotherapy resistance linked to myeloid cytokines, and it may eventually inform therapies for other cancer types that share similar microenvironmental features. According to the researchers, single-cell atlases move us closer to characterising the cellular fingerprint of a patient’s tumour. “In the future, clinicians could use such fingerprints to select treatments that best target a tumour’s microenvironment, not just its genetics,” adds first author Lina Kroehling, a PhD student in bioinformatics.