LAUSR

BACKGROUND Randomized controlled trials (RCTs) have shown improvements in breast cancer outcomes from extending treatment with aromatase inhibitors (AIs) beyond the initial 5years after diagnosis. Consistency of this effect in common clinicopathologically defined subgroups was not been reported systematically. METHODS We identified RCTs comparing extended AIs to placebo or no treatment using a systematic search of MEDLINE and a review of abstracts from key conferences between 2013 and 2016. Hazard ratios (HRs) and 95% confidence intervals (CI) for disease-free survival (DFS) were included in a meta-analysis using generic inverse variance and random effects modelling. Pre-specified subgroups included: age (<60±5years versus ≥60±5 years), tumor size (>2cm versus ≤2cm), nodal status (positive versus negative), hormone receptor status (double versus single receptor expression) and receipt of adjuvant chemotherapy (yes versus no). RESULTS Seven trials comprising 16,349 patients were analyzed. Overall, the effect of extended AIs was similar in all subgroups. Non-significantly greater effect sizes were seen in patients with larger tumors (HR for DFS 0.77 versus 0.88, p for difference=0.44), nodal involvement (HR=0.72 versus 0.83, p for difference=0.22), double hormone receptor expression (HR=0.68 versus 1.01, p for difference=0.31) and receipt of adjuvant chemotherapy (HR=0.71 versus 0.80, p for difference=0.51). CONCLUSIONS Extended treatment with AIs is associated with similar relative improvements in DFS in all subgroups analyzed. The combination of greater effect size and higher absolute risk of recurrence in node positive and larger tumors will likely translate to higher absolute benefits from extended AIs in these groups.