Patient-derived tumor organoids predict responses to irinotecan-based neoadjuvant chemoradiotherapy in patients with locally advanced rectal cancer.

ABSTRACT

Adding irinotecan to neoadjuvant chemoradiotherapy (nCRT) in locally advanced rectal cancer (LARC) increases the pathologic complete response (pCR) rate but brings more toxicities. Robust biomarkers to predict response to irinotecan-based nCRT are extremely necessary for selecting the right patients. Our previous study suggests that patient-derived tumor organoids (PDTOs) sensitivity to chemoradiotherapy matches patient responses. In this study, we investigated whether PDTOs sensitivity to irinotecan can predict complete response (CR) and survival. Eligible patients receiving irinotecan-based nCRT between April 5, 2017 and December 11, 2020 were enrolled in the training cohort (n = 91) for response prediction and survival analysis. Patients receiving nCRT between February 21, 2021 and September 17, 2021 were included in the validation cohort (n = 27). Predictive performances of irinotecan organoid size ratio (OSR) for CR or pCR were evaluated. The irinotecan-sensitive groups had higher response rates compared with the insensitive groups (training cohort 71.8% vs 24.4%, P < .0001; validation cohort, 81.8% vs 18.8%, P = .002). Moreover, the irinotecan-sensitive group had higher rates of 3-year disease-free survival (DFS 71.6% vs 55.5%, P = .034) and distant metastasis-free survival (DMFS, 77.9% vs 57.2%, P = .015) than the irinotecan-insensitive group. 5-FU and irradiation sensitivities failed to predict 3-year DFS (5-FU 65.4% vs 61.9%, P = .643; irradiation 84.8% vs 57.8%; P = .072). Performances of irinotecan OSR to predict CR or pCR were good in the training cohort (CR AUC = 0.828; 95% CI = 0.723-0.932; pCR AUC = 0.864; 95% CI = 0.759-0.961). The validation showed robust predictive ability (CR AUC = 0.796, 95% CI = 0.5974-0.9952; pCR AUC = 0.917, 95% CI = 0.7921-1.0000). Irinotecan sensitivity in PDTOs was a predictive and prognostic factor in LARC.