Background: Resistance to hormonal therapy may be associated with activation of the PI3K/AKT pathway. Preclinically, everolimus (RAD), an oral inhibitor of mTOR, has been shown to reverse resistance to tamoxifen (TAM). In a prior randomized phase II trial in estrogen-receptor positive operable breast cancer pts, RAD significantly increased neoadjuvant AI (letrozole) efficacy when given in combination. The objective of this randomized phase II study was to estimate the efficacy of the RAD+TAM combination in AI pretreated hormone-receptor positive/HER2 negative MBC pts based on the assumption that prior exposure to an AI might potentially enrich the proportion of pts whose tumor may be driven by an activation of the PI3K/AKT/mTOR pathway. Methods: Eligible patients were stratified by time to progression after prior AI treatment and randomized 1:1 to receive either TAM (20mg/day) alone or RAD+TAM (RAD: 10 mg/d; TAM: 20mg/d). The primary objective was to estimate clinical benefit rate (CB) defined as the absence of progression at 6 months in the RAD+TAM arm. Using a Simon two-stage Minimax design, with alpha=5% and power=90%, considering a gain in CB of 20% as the minimum needed to warrant further study for the combination and assuming a CB of 50% in the TAM arm, 53 evaluable patients were needed in both arms. Secondary endpoints included safety and time to progression (TTP). Results: In total, 111 pts (TAM: 57, RAD+TAM: 54) were randomized. Baseline characteristics were well balanced between the two treatment arms; median age was 64 years (range, 41-86); most pts were PS 0 (55 pts, 51%) or PS 1 (46 pts, 43%) Prior AI treatment had been given to 34 pts (31%) in the adjuvant setting; to 67 pts (60%) in the metastatic setting and 10 pts (9%) in both the adjuvant and metastatic setting. This population was poorly hormone sensitive as all but 10 pts (9%) had progressed either during AI or within 6 months after adjuvant AI. Furthermore, 57 pts (51%) and 28 pts (25%) had received prior chemotherapy in the adjuvant and/or metastatic setting, respectively. Efficacy: In an intent-to-treat analysis with a median follow-up of 13 months, CB was 42.1% (95% CI, 29.1-55.9) in the TAM arm and 61.1% (95% CI, 46.9-74.1) in the RAD+TAM arm. Median TTP was 4.5 months (95% CI, 3.7-8.7) with TAM and 8.5 months (95% CI, 6.01-13.9) with RAD+TAM (log-rank test: p=0.008, exploratory analysis). At the time of analysis, 17 pts had died in the TAM arm and 5 patients had died in the RAD+TAM arm (there was no toxic death). Safety: Safety data showed that toxicity was manageable in both groups. RAD had to be decreased to 5 mg/day for 15 pts (28%); 3 and 2 pts had to stop the treatment due to toxicities in the TAM and RAD+TAM arms respectively. Severe adverse events (G3-4) >10% were stomatitis (0/11%, TAM/RAD+TAM) and pain (19%/7%). Conclusions: RAD combined with tamoxifen provides significant improvement in the 6 months clinical benefit rate compared to tamoxifen alone. Based on these promising results, this combination warrants further study in hormone-receptor positive/HER2 negative MBC after progression on AI. Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr S1-6.