Fourteen studies with 1,484 participants were included in the meta-analysis. A further 12 studies were identified, but these were excluded because they did not meet the estimated sample size requirement for meta-analysis. These studies, along with those excluded because they compared two active treatments, were tabulated in the paper.
A meta-analysis of 6 trials with 503 people assessed 2-year survival rates following embolisation/chemoemobolisation. Arterial embolisation improved 2-year survival in comparison with controls (OR 0.53, 95% CI: 0.32, 0.89, p=0.017). There was an objective response in 35% (range: 16 to 61) of the participants.
The sensitivity analysis suggested a significant benefit from chemoemobolisation with doxorubicin or cisplatin (OR 0.42, 95% CI: 0.20, 0.88; 4 studies, 323 participants), but not with embolisation alone (OR 0.59, 95% CI: 0.29, 1.20; 3 studies, 215 participants). In addition, the survival advantage was maintained when looking only at randomised trials with a control arm of conservative management (4 RCTs, 367 participants) and when assessing only 'high quality' studies (5 RCTs, 440 participants). There was no statistically significant heterogeneity for the analysis or any of the sensitivity analyses.
A meta-analysis of 7 trials with 689 people found that tamoxifen had no impact on 1-year survival (OR 0.64, 95% CI: 0.36, 1.13, p=0.13). Sensitivity analyses found no impact in high-quality trials (4 studies, 512 participants), but benefits favouring tamoxifen when assessing the 3 low-quality trials (177 participants). A sensitivity analysis of 3 double-blind, placebo-controlled trials (259 participants) found no survival benefit. An analysis of all studies with tamoxifen as adjuvant therapy found similar negative results (7 trials, 898 participants). There was no statistically significant heterogeneity in the core analysis or in the sensitivity analyses.