Six studies (N=579) were included.
Hepatic artery infusion led to a 12.9% improvement in survival at 1 year (95% CI: 4.8, 20.9, p=0.002). The 2-year survival difference was 7.5% (95% CI: 0.9, 14.2, p=0.026). When only those studies that did not allow treatment crossover were included in the analysis, the survival advantage with hepatic regional chemotherapy appeared greater at 1 year, at 19.1% (95% CI: 8, 30.2, p=0.001); after 2 years, the survival difference was 8.6% (95% CI: 0.1, 17.1, p=0.049).
In one study, half of the control group did not actually receive systemic chemotherapy. Re-calculating summary statistics after excluding that study showed that the advantage of hepatic regional chemotherapy was still significant at 1 year (95% CI: 0.4, 19.5, p=0.041), but not at 2 years (95% CI: -1.8, 14.5, p=0.124).