Published online Jan 15, 2020. doi: 10.4251/wjgo.v12.i1.113
Peer-review started: March 22, 2019
First decision: July 31, 2019
Revised: September 20, 2019
Accepted: October 14, 2019
Article in press: October 14, 2019
Published online: January 15, 2020
Processing time: 284 Days and 15.8 Hours
In 2015, Kidane published a Cochrane review and meta-analysis to summarise the impact of preoperative chemotherapy versus surgery alone on survival for resectable thoracic esophageal cancer. The authors concluded that preoperative chemotherapy improved overall survival (OS).
The aim of this article was to assess the validity of the three most powerful studies included in the Cochrane review and the meta-analysis supporting the advantage of preoperative chemotherapy and to investigate the impact of an exclusion of these three studies on the result of the meta-analysis.
OS was selected as the endpoint of interest. Among the ten included papers which analysed this endpoint, we identified the three publications with the highest weights influencing the final result. The validity of these papers was analysed using the CONSORT checklist for randomized controlled trials. We performed a new meta-analysis without the three studies to assess their impact on the general result of the original meta-analysis.
The three analysed studies revealed several inconsistencies. Inappropriate answers were found in up to one third of the items of the CONSORT checklist. Missing information about sample-size calculation and power, unclear or inadequate randomisation, and missing blinded set-up were the most common findings. When the three criticized studies were excluded in the meta-analysis, preoperative chemotherapy showed no benefit in OS.
The three most powerful publications in the Cochrane review show substantial deficits. After the exclusion of these studies from the meta-analysis, preoperative chemotherapy does not seem to result in an advantage in survival. We suggest a more critical appraisal regarding the validity of single studies.
Core tip: The quality of single studies is crucial in order to perform valid meta-analyses that are often used as basis for guideline recommendations. We critically analysed a recent Cochrane meta-analysis that supports the use of preoperative chemotherapy for resectable thoracic esophageal cancer in order to improve overall survival. The most powerful included studies showed several inconsistencies according to the requirements of the Consort checklist for randomized controlled trials. After the exclusion of these studies from the meta-analysis, preoperative chemotherapy does not seem to result in an advantage in survival. We suggest a more critical appraisal regarding the validity of single studies.