Issing, Christian, Menche, Constantin, Richter, Mara Romero, Mosa, Mohammed H., von der Grün, Jens, Fleischmann, Maximilian, Thoenissen, Philipp, Winkelmann, Ria, Darvishi, Tahmineh, Loth, Andreas G., Ghanaati, Shahram, Rödel, Franz, Wild, Peter J., Brandts, Christian H., Stöver, Timo, and Farin, Henner F.
Background: Head and neck cancers (HNC) represent an extremely heterogeneous group of diseases with a poorly predictable therapy outcome. Patient-derived tumor organoids (PDTO) offer enormous potential for individualized therapy testing and a better mechanistic understanding of the main HNC drivers. Methods: Here, we have established a comprehensive molecularly and functionally characterized head and neck organoid biobank (HNOB) recapitulating the clinically relevant subtypes of TP53 mutant and human papillomavirus type 16 (HPV 16) infection-driven HNC. Organoids were exposed to radiotherapy, and responses were correlated with clinical data. Genetically engineered normal and tumor organoids were used for testing the direct functional consequences of TP53-loss and HPV infection. Results: The HNOB consisting of 18 organoid models, including 15 tumor models, was generated. We identified subtype-associated transcriptomic signatures and pathological features, including sensitivity to TP53 stabilization by the MDM2 inhibitor Nutlin-3. Furthermore, we describe an in vitro radio response assay revealing phenotypic heterogeneity linked to the individual patient's treatment outcome, including relapse probability. Using genetically engineered organoids, the possibility of co-existence of both cancer drivers was confirmed. TP53 loss, as well as HPV, increased growth in normal and tumor organoids. TP53 loss-of-function alone was insufficient to promote radiation resistance, whereas HPV 16 oncogenes E6/E7 mediated radiosensitivity via induction of cell cycle arrest. Conclusion: Our results highlight the translational value of the head and neck organoid models not only for patient stratification but also for mechanistic validation of therapy responsiveness of specific cancer drivers. [ABSTRACT FROM AUTHOR]