利用患者来源类器官定义放射敏感性指标以预测口腔癌预后

Over 60% of patients with oral cancer are diagnosed at advanced stages, and the standard treatment was surgery followed by adjuvant radiotherapy. Nevertheless, 30-50% of patients experience recurrence within two years, whereas patients with radiosensitive tumors may be exposed to unnecessarily high radiation doses. Therefore, predicting radiotherapy response and stratifying patients is essential for personalized treatment approaches and improving clinical outcomes. Patient-derived organoids (PDOs) are robust preclinical models that recapitulate the characteristics of patients' tumors. In this study, we defined radiation sensitivity indicators using an oral cancer PDO library comprising 102 patient-derived organoids. Dose-dependent radiation responses were evaluated for 68 PDOs and correlated with clinical data. When organoids were classified based on the mean survival fraction at 2 Gy (SF2), patients in the high-SF2 group exhibited a significantly shorter recurrence-free survival. It suggests that PDO-based assessment of radiation response may serve as a useful predictor of clinical outcomes. In addition, a radiation sensitivity-related pathway score was generated from differentially expressed genes (DEGs) identified in radiosensitive PDOs and applied to The Cancer Genome Atlas (TCGA) oral cancer cohort. Patients were classified into low- and high-score groups according to the sensitivity score. The high-score group had a significantly more favorable survival prognosis. Furthermore, to identify a tissue-detectable radiation sensitivity marker (RSM), we selected the gene associated with the most enriched pathway in the low-SF2 organoid group, resulting in the identification of a candidate marker, RSM-1. Subsequent immunohistochemical (IHC) analysis demonstrated that RSM-1 was expressed at higher levels in tumors from patients without recurrence. Overall, this study utilized patient-derived organoids to evaluate radiation responses, derive a radiosensitivity score, and identify a clinically relevant biomarker. By integrating molecular and phenotypic tumor characteristics, this approach provides reliable prognostic indicators, enabling prediction of patients' radiation responses and supporting personalized therapeutic strategies to improve prognosis in oral cancer.