Grant: Knowledge Synthesis Grant

Laparoscopic Rectal Cancer Surgery; Meta-Analysis; Meta-Regression; Minimally Important Difference; Non-Inferiority Randomized Clinical Trials; Open Rectal Cancer Surgery; Rectal Cancer; Surgical Management; Systematic Review

<html> <body> <p>Surgical resection is the primary curative treatment modality for rectal cancer (RC).¹ Although better short-term outcomes are achieved with a laparoscopic approach,² it is not clear that laparoscopic surgery achieves the excellent tumour resection and oncologic outcomes achieved with modern open surgery. Laparoscopic rectal dissection is very challenging and it is possible that this approach compromises oncologic resection, resulting in increased local recurrence. A number of randomized controlled trials (RCTs) have been conducted to assess whether laparoscopy is non-inferior to open resection for RC. Early studies from South Korea (COREAN trial)³ and Europe (COLOR II trial)⁴ demonstrated comparable outcomes, however two recently published RCTs, found the quality of surgical dissection tended to be lower in the minimally invasive group.</p> <p>Several systematic reviews have previously compared outcomes of laparoscopic and open surgery for RC. These syntheses need to be updated incorporating the long-term results of COLOR II trial⁵ and the short-term oncologic outcomes of two large RCTs published in 2015.⁶^,⁷ Additionally, methodological limitations of the previous reviews, particularly evaluation of non-inferiority in meta-analysis, need to be addressed. Finally, the effect of patient, surgeon, tumour and treatment factors known to affect the difficulty of RC dissection should be evaluated.</p> <p>Our team has expertise in RC surgery, knowledge synthesis and meta-analytic techniques. We aim (1) To determine if adequacy of surgery and oncologic outcomes of laparoscopic surgery are non-inferior to those of open surgery for rectal cancer (RC); and (2) to evaluate if differences in study&nbsp;populations associated with difficult pelvic dissection result in the observed differences in outcomes between trials. We will systematically review available RCTs in accordance with recommendations for non-inferiority meta-analysis.⁸ For short-term oncologic outcomes, we will meta-analyze data on circumferential resection margins, distal resection margin, and total mesorectal excision using random effects models. We will interpret the findings using a minimally important difference defined <em>a priori</em> in order to conclude non-inferiority. In addition to the risk of bias assessment using the Cochrane tool for RCT, we will evaluate specific factors associated with bias for studies of equivalence or non-inferiority and will perform meta-analysis both per intention-to-treat and per-protocol. For long-term outcomes (overall and disease free survival) hazard ratios and their variance will be used (extracted directly when available and if not by indirect methods⁹). We will perform stratified analyses by factors recognized with difficult RC dissection if subgroup data are available. We will also measure the effect of these factors on the observed effects by fitting meta-regression models.¹⁰^,¹¹ Sensitivity analysis of the effect of individual studies on the pooled meta-analysis results for each outcome (leave-one-out meta-analysis) will be performed.</p> <p>Our knowledge synthesis will provide evidence that will enable development of recommendations for the use of laparoscopic surgery for RC. There have been no previous meta-regression analyses on this subject, and this study provides an opportunity to add an informative dimension, enabling surgeons to better individualize care for their patients with RC when considering surgical approach.</p> </body> </html>