Pre-operative endoscopic assessment and MRI as predictors of pathological complete response and long-term survival in locally advanced rectal cancer after neoadjuvant therapy

Abstract Background The optimal modality for predicting pathological complete response (pCR) after neoadjuvant therapy in rectal cancer remains uncertain. This study evaluated the ability of pre‑operative endoscopy and magnetic resonance imaging (MRI) to predict pCR and long‑term survival in stage II/III rectal cancer following neoadjuvant treatment. Methods Retrospective cohort study of a prospectively maintained database of patients with stage II/III rectal cancer undergoing neoadjuvant therapy followed by surgery between 2016 and 2024 at a single institution. Patients underwent pre-operative endoscopic assessment and MRI restaging 4–8 weeks after completion of their neoadjuvant therapy, followed by surgery at 8–12 weeks. Endoscopic response was classified according to the Memorial Sloan Kettering schema. Strictures were classified as a separate phenotype. MRI tumour response was classified according to the tumour regression grade (TRG) score. Primary outcome was the pCR rate. Results Of 203 patients, the overall pCR rate was 19.7%. pCR rates were 76.4% for cCR, 47.6% for nCR, 50% for strictures and 0% for iCR. Endoscopic response showed a greater discrimination for pCR compared to MRI TRG, (AUC 0.92 vs. 0.65, z = 5.74, p < 0.001). When cCR, nCR and strictures were grouped as “good endoscopic response” and iCR as “poor response”, 8-year OS was 80.1% versus 62.3% ( χ 2 = 4.9, p = 0.03) and 8-year DFS was 80.3% versus 60.4% ( χ 2 = 6.2, p = 0.01), with poor response associated with higher hazards of death and recurrence in Cox models. MRI TRG also stratified prognosis, and each one‑point increase in TRG was associated with a 2.31‑fold higher hazard of death and a 1.73‑fold higher hazard of a DFS event. Conclusions Endoscopic assessment provides a stronger predictor of pCR compared to MRI alone. Survival differences by response category suggest that intrinsic tumour biology, rather than delays in surgery, drives outcomes and supports careful WW surveillance for selected patients.