| Abstrakt: |
Introduction: There is no clear consensus on using the response MRI as opposed to the pretreatment MRI for surgical planning in cT4 low rectal cancer. The objective of this study is to determine the safety of using response MRI in surgical planning for T4 rectal cancer. Methods: This study is a retrospective review of a prospectively maintained database of abdominoperineal resections conducted at a single tertiary cancer center. Patients undergoing an abdominoperineal resection were divided into 2 groups: group A (clinical T3, mesorectal fascia positive) and group B (clinical T4), and propensity matching was used to account for uneven distribution of baseline characteristics. Primary outcome was the rate of pathological circumferential resection margin positivity. Secondary outcomes were survival outcomes and recurrence patterns. Results: There were 237 patients in group A and 127 in group B, in the unmatched cohort, with a significantly higher number of females (43.3% vs. 28.7%, p = 0.005) and anterior circumferential resection margin positivity (68.5% vs. 49%, p < 0.001), with a lower number of patients receiving neoadjuvant chemotherapy in group B (38.6% vs. 49.8%, p = 0.041). After propensity matching baseline characters were comparable. There was a higher percentage of extended-total mesorectal excisions in group B (58.5% vs. 40.5%, p = 0.004). The rate of pathological circumferential positivity was comparable in both groups (20/168 in group A {11.9%} vs. 13/107 in group B {12.1%}, p = 0.951) with no impact of group on circumferential resection margin positivity on univariate (OR 1.023, p = 0.951) or multivariate regression (OR 0.993, p = 0.987). Both the DFS (median DFS 39 months vs. 54 months, p = 0.970) and OS (3-year OS 72% vs. 67%, p = 0.798) were comparable between both groups. Conclusion: For T4 low rectal cancers, post-treatment MRI can be used for surgical planning without any detriment in pathological or long-term oncological outcomes. [ABSTRACT FROM AUTHOR] |
Full text from SpringerLink