A meta-analysis of safety and efficacy on first-line S-1 therapy in cancer patients

Background: The incidence of grade 3/4 adverse effects due to S-1 therapy and the efficacy of S-1-based therapy vs. S-1 monotherapy have not been well described. We conducted an updated meta-analysis to evaluate this problem. Methods: We searched the electronic databases, including PubMed, Embase, and Cochrane database to investigate the effects of phase 2 and 3 prospective clinical trials on first-line S-1 therapy in cancer patients. Data from included studies were pooled using Stata version 12.0. Results: Twenty eight studies were included. First-line S-1 monotherapy showed low incidence of grade 3/4 adverse effects. And the highest rate grade 3/4 hematological event was neutropenia [7%, 95% confidence interval (CI): 5-8%]; the highest rate grade 3/4 non-hematological event was anorexia (7%, 95% CI: 6-9%). Longer overall survival (OS) time and progression-free survival (PFS) time was exhibited in S-1-based therapy, compared with S-1 monotherapy [hazard ratio (HR) 0.836, 95% CI: 0.761-0.911, P=0.000, and HR 0.650, 95% CI: 0.540-0.759, P=0.000, respectively]. However, the incidence of grade 3/4 adverse effects was also higher in S-1-based therapy than S-1 monotherapy in cancer patients, with relative risk (RR) of neutropenia and anorexia were respectively 4.62 (95% CI: 2.92-7.30) and 1.46 (95% CI: 0.84-2.55). Conclusions: S-1 monotherapy was demonstrated with low incidence of high grade adverse effects, therefore it is well tolerated for majority cancer patients; S-1-based therapy significantly improved OS and PFS compared with S-1 monotherapy, with an increased risk of high grade adverse effects.