Association of post-radiotherapy MRI response pattern with long-term survival outcome in rectal cancer patients...

Pathologic complete response (pCR) after neoadjuvant radiotherapy has been used as a prognostic criterionin locally advanced rectal cancer. However, while MRI abnormalities may persist after treatment, this may be attributable to treatment-related changes rather than residual malignancy. In this study, we examined the correlation between post-radiotherapy MRI response pattern and pathologic response, and their association with long-term oncologic outcome.

This retrospective observational cohort study included 225 patients with stage II-III rectal adenocarcinoma who received neoadjuvant radiotherapy prior to total mesorectal excision (TME) with curative intent between January 1, 2017 and December 31, 2023. Patients were classified according to pathologic response as pathologic complete response (pCR; ypT0N0, n = 48) or non-pathologic complete response (non-pCR, n = 177). Evaluated findings included baseline clinical characteristics, pretreatment MRI features, MRI response pattern after radiotherapy, pathologic results, recurrence, progression-free survival (PFS), and overall survival (OS). Data were analyzed with Kaplan-Meier method and Cox proportional hazards regression.

Overall, 21.3% of patients achieved pCR. These patients had significantly smaller baseline maximum tumor diameter (45.0 mm vs. 50.0 mm, P < 0.001), smaller gross tumor volume (46.5 cm³ vs. 60.8 cm³, P < 0.001), and a lower prevalence of MRI-defined node-positive disease (41.7% vs. 64.4%, P = 0.008) compared with the non-pCR group. Patients who achieved pCR had lower rates of recurrence or distant metastasis than those who did not (10.4% vs. 23.7%, P = 0.047). Death occurred in 4.2% and 10.7% of patients, respectively (P = 0.262). Median PFS was 51.0 months in the pCR group and 41.0 months in the non-pCR group; median OS was 62.5 months and 50.0 months, respectively. In univariable Cox regression analysis, pCR was associated with a lower risk of progression or recurrence (HR = 0.40, 95% CI 0.16-1.00; P = 0.050); however, this association was attenuated after multivariable adjustment (adjusted HR = 0.46, 95% CI 0.20-1.08; P = 0.073). Residual DWI hyperintensity (83.3%), tumor-bed fibrosis (97.9%), fibrotic scar formation (68.8%), and submucosal edema (68.8%) were frequently observed in pCR patients. In the pCR group, none of these post-treatment MRI response patterns was significantly associated with progression-free or overall survival.

Patients with locally advanced rectal cancer who achieved pathologic complete response after neoadjuvant therapy had a lower recurrence rate, while the observed difference in overall survival did not reach statistical significance. Despite complete pathologic remission, abnormal MRI findings often persisted after treatment and were not significantly associated with unfavorable prognosis within the pCR cohort. This study suggests that post-treatment MRI abnormalities should be interpreted in conjunction with pathologic and clinical findings, rather than being automatically attributed to residual disease.