Four phase III superiority RCTs (n=2,131) were included in the review. Three RCTs were double-blinded double-dummy designs, three used an intention-to-treat analysis and three reported sample-size calculations. Randomisation methods were poorly described.
Fulvestrant versus tamoxifen: One RCT (n=587) found no significant differences between treatments on any outcome. Non-inferiority on the primary endpoint of time to progression was not confirmed (the upper confidence interval of the hazard ratio did not meet the predefined criterion of being 1.25 or lower). Unadjusted analyses showed a statistically significant benefit for tamoxifen in overall survival (HR 1.29, 95% CI 1.01 to 1.64, p=0.04). Results for subgroups of women with oestrogen receptor positive and or progesterone-receptor positive tumours showed no difference in overall survival between treatment groups. There were no significant differences in adverse events between the groups, except that patients on tamoxifen experienced more hot flashes (p=0.0501).
Fulvestrant versus anastrazole: Two RCTs (n=851) were combined. There were no significant differences between the groups on any primary outcome. Following the failure to show superiority of fulvestrant, an unplanned non-inferiority analysis was conducted. This showed non-inferiority for time to progression (5.5 months versus 4.1 months, HR 0.95, 95% CI 0.82 to 1.10) and objective response rate (difference of 2.75%, 95% CI 2.27% to 9.05%) compared to anastrazole. The duration of response was significantly longer in the fulvestrant groups at 22.1 months follow-up (ratio 1.30, 95% CI 1.13 to 1.50, p<0.01) in all patients and in responders only (16.7 versus 13.7 months). The only significant difference in adverse effects was a higher incidence of joint disorders in patients treated with anastrazole (12.8% versus 8.3%, p=0.02).
Fulvestrant versus exemestane: One RCT (n=693) found no statistically significant differences on any primary or secondary efficacy outcomes and no significant differences in adverse effects. Both treatments appeared well tolerated (only 2% of fulvestrant patients and 2.6% of exemestane patients withdrew as a result of adverse events).