Primary surgery followed by selective radiochemotherapy versus conventional preoperative radiochemotherapy for patients with locally advanced rectal cancer with MRI-negative circumferential margin (PSSR): A multicenter, randomized, open-label, noninferiority, phase 3 trial.

3515 Background: Neoadjuvant radiochemotherapy is the standard treatment for locally advanced rectal cancer. However, radiation therapy can lead to bowel, urinary and sexual dysfunction. The study aims to clarify whether locally advanced rectal cancer with a distance of 6 to 12 cm from the anus with negative circumferential margins (CRM) predicted by MRI can be exempted from preoperative radiotherapy. Methods: PSSR (NCT02121405), a multicentre, randomized, open-label, non-inferiority, phase 3 trial, was done across 10 hospitals in China. Patients aged from 18 to 75 years with locally advanced rectal cancer (including all cT3/4 and/or Nany) with negative MRI-predicted CRM and tumor distance of 6 to 12 cm from anus were randomly assigned (1:1), by central randomization, to intervention group (surgery directly, in which positive CRM was supplemented with chemoradiotherapy, negative CRM received adjuvant chemotherapy according to surgical pathologic staging), or control group (neoadjuvant chemoradiotherapy then underwent surgery and adjuvant chemotherapy). The primary endpoint is the rate of 3-year disease-free survival (DFS). A sample size of 350 was set according to the original non-inferiority margin of 15% for the difference of rates for 3-year DFS, which was changed to 5% under the Data Safety Monitoring Board (DSMB) suggestion after an interim analysis. The efficacy analysis followed the per-protocol (PP) principle. Results: From December 2015 to February 2021, 299 patients were recruited. DSMB stopped this trial due to the difference of 3-year DFS’s rate between two groups more than 5% at the interim analysis in July 2021. A total of 275 were included for the final analysis. Based on the PP dataset, 238 patients followed the original protocol (135 patients in the intervention group and 103 patients in the control group). Patients who had positive CRM were 2(1.5%) and 1(1.0%) in two groups, respectively ( P= 1.00). After a median follow-up of 34.6 (IQR: 18.2-45.7) months, a total of 42 patients had positive events (30 patients in the intervention group and 12 patients in the control group). There were 5 (3.7%) patients with local recurrence in the intervention group ( P= 0.062). Patients with pathological complete response (pCR) and near pCR were 19 (18.4%) and 21 (20.4%) in the control group. The rate of 3-year DFS was 81.1% (77.3%-84.9%) (HR = 2.02, 95%CI: 1.01-4.06, P= 0.048) in the intervention group and 86.6% (82.7%-90.5%) in the control group, with difference of 5.4% (5.3%-5.6%), which did not meet the criteria for non-inferiority. Conclusions: In terms of DFS, initial surgery was inferior to conventional preoperative radiochemotherapy for locally advanced middle rectal cancer with MRI negative circumferential margin. Clinical trial information: NCT02121405.