A Phase II component of the NRG-HN009 trial finds weekly cisplatin does not significantly reduce acute toxicity versus every-three-week dosing in p16-negative locally advanced head and neck cancer, directly informing how clinicians can balance scheduling and tolerability. In esophageal cancer, CT-based AI radiomics showed mixed results: a proof-of-concept model predicting early radiotherapy response appeared unstable under fully nested validation, cautioning against premature deployment. Across real-world practice, an esophageal cancer study suggests residual tumor burden after immunotherapy is a strong predictor of survival, pointing to practical post-treatment stratification signals that can complement biomarker strategies.