Laparoscopic versus open surgery for rectal cancer: individual patient data meta-analysis of the ALaCaRT and Z6051 randomized trials.

BACKGROUND: The multicentre ALaCaRT and ACOSOG Z6051 randomized trials were unable to demonstrate non-inferiority of laparoscopic versus open surgery for rectal cancer with respect to a composite pathology metric indicating successful resection. Neither trial was individually powered to detect differences in long-term recurrence or survival. This planned meta-analysis determined long-term oncological outcomes of laparoscopic versus open proctectomy for rectal adenocarcinoma. METHODS: This prospective meta-analysis included individual patient data from patients with cT1-3 N0-2 M0 rectal adenocarcinoma enrolled in the ALaCaRT and ACOSOG Z6051 trials. Pathologically successful resection was defined as complete or near-complete total mesorectal excision, a clear circumferential resection margin (CRM; > 1 mm), and a clear distal resection margin (> 1 mm). The non-inferiority margin for disease-free survival (DFS) was an absolute difference of 5% at least 3 years after surgery. RESULTS: The combined data set included 935 patients (65.6% men, mean age 60.7 years, mean body mass index 26.7 kg/m2) randomized to open (457 patients) or laparoscopic (478 patients) proctectomy. Pathologically successful resection was lower in the laparoscopic than open group (85.1% versus 89.9%, respectively; pooled estimate 4.6% difference; 95% confidence interval (c.i.) -8.6% to -0.5%). The median follow-up was 60.2 (interquartile range 49.9-61.1) months. Three-year DFS was 75.2% (95% c.i. 71.1% to 79.2%) and 76.5% (95% c.i. 72.5% to 80.6%) for the laparoscopic and open groups, respectively (pooled estimate difference -1.5%; 95% c.i. -7.2% to 4.2%). Non-inferiority of laparoscopic surgery was not demonstrated because the lower one-sided 95% c.i. (-6.3% to 100%) crossed -5%. Three-year locoregional recurrence was higher in laparoscopic than open group (5.4% (95% c.i. 3.3% to 7.5%) versus 2.0% (95% c.i. 0.7% to 3.4%), respectively; pooled estimate 3.1% difference (95% c.i. 0.6% to 5.6%)). A clear CRM was the most significant and only pathological predictor of both DFS (P < 0.0001) and overall survival (P < 0.0001). CONCLUSION: Laparoscopic proctectomy led to a lower rate of pathologically successful resection and a higher rate of locoregional recurrence at 3 years. The possibility of subsequent poorer DFS or overall survival rate requires further evaluation, because this analysis was not specifically powered for these endpoints.