Summary of the four analysed studies
| Study (year) | Neri et al 14 | Sakuramoto et al 15 | Fujimoto et al 16 | Douglass et al 17 |
| Number of included patients (intervention vs control) | 69 vs 68 | 529 vs 530 | 129 vs 120 | 71 vs 71 |
| Inclusion criteria | Resected, node-positive gastric cancer, no distant metastases | R0 resected, node positive (D2 or more extensive lymph node dissection), no distant metastases, no neoadjuvant CTx | Resected gastric cancer, no neoadjuvant and intraoperative CTx | Resected stomach cancer or cancer at the gastro-oesophageal junction, complete recovery from surgery, ingestion of a solid diet, maintenance of weight and absence of infection, no distant metastasis (directed extension of the neoplasm was accepted) |
| Intervention group | Postsurgical CTx with epidoxorubicin, leucovorin and 5-FU | Postsurgical CTx with S-1, an oral fluoropyrimidine | Postsurgical CTx with 5-FU and FT-207, a derivate of 5-FU | Postsurgical CTx using 5-FU and methyl-CCNU |
| Control group | No treatment | No treatment | No treatment | No treatment |
| Outcome (intervention vs control) | Median survival time (31 months (range 7 to 60+) vs 18 months (range 2 to 60+), p<0.01) | HR for death 0.68 (95% CI 0.52 to 0.87, p=0.003). 3-year overall survival rate 80.1% (95% CI 76.1 to 84.0) vs 70.1% (95% CI 65.5 to 74.6) | Higher survival rates for intervention (χ2 at 24 and 36 months, p<0.05 and p<0.1, respectively) | 29 deaths vs 40 deaths. Log-rank testing of the two survival patterns revealed a p value of 0.06. Covariate analysis increased the value of this test to a level of significance, p<0.03. 50th percentile at 56 months vs median survival at 33 months |
| Weight assigned in the Cochrane review (%) | 3.8 | 4.4 | 2.6 | 2.9 |
5-FU, 5-fluorouracil; CTx, chemotherapy.