All randomised controlled trials (closed to recruitment before 2004) with individual patient data (IPD) comparing chemotherapy after resectable gastric cancer with surgery alone for overall survival and disease-free survival were eligible for inclusion. Overall survival was defined as the time from randomisation to mortality from any cause or to last follow up. Disease-free survival was calculated from time to relapse, second cancer, or mortality (whichever came first). Relevant chemotherapy regimens were classified as: mono-chemotherapy; fluorouracil and mitomycin C with anthracyclines; fluorouracil and mitomycin C and others without anthracyclines; and other poly-chemotherapy. Trials investigating immunotherapy or neoadjuvant or perioperative chemotherapy, radiotherapy or intraperitoneal chemotherapy were excluded.
Male patients predominated (over 60%), and most were aged between 50 and 59 years (29%). Dosages and schedules were variable in the different trials, as was cancer stage measured using the International Union Against Cancer staging tool. Patient characteristics were listed in separate supplementary online tables (see URL for Additional Data).
The authors did not state how many reviewers selected the studies.