Poster Display session 2 Poster Display session

214P - Pre-specified interim analysis of the SAFE trial (NCT2236806): A 4-arm randomized, double-blind, controlled study evaluating the efficacy and safety of cardiotoxicity prevention in non-metastatic breast cancer patients treated with anthracyclines with or without trastuzumab (ID 3613)

Presentation Number
Lecture Time
12:00 - 12:00
  • Lorenzo Livi (Firenze, Italy)
Session Name
Poster Display session 2
Poster Area (Hall 4), Fira Gran Via, Barcelona, Spain
12:00 - 13:00



SAFE trial (NCT2236806) is a phase 3 study comparing the effect on subclinical heart damage of bisoprolol (B), ramipril (R), or both drugs (R+B), as compared to placebo (P), in breast cancer treated with (neo)adjuvant anthracyclines +/- trastuzumab.


Primary endpoint is subclinical cardiotoxicity measured with echocardiography and global linear strain (GLS). This interim analysis was pre-specified on the first 120 patients who had completed cardiological assessments at 12-mos. Stopping rules per arm were: dose reduction >15%, study withdraw rate >5%, and no significant impact on 3D-left ventricular ejection fraction (3D-LVEF) as compared to T0 at 12-mos assessment.


A total of 191 out of 480 patients have been enrolled; overall 123 patients were available for the analysis (P = 34; R = 28; B = 31; R+B=30). 3D-LVEF decreased at 3-mos (-3.3%; p < 0.001), at 6-mos (-5.2%; p < 0.001) and at 12-mos (-3.7%; p = 0.004) respect to T0 in P; at 3-mos (-2.4%; p = 0.001), at 6-mos (-1.9%; p = 0.010), at 12-mos (-2.2%; p = 0.045) in R. In B and R+B patients no significant changes were observed at 3- and 12-mos, with a decrease at 6-mos (-2.5% and 3.0%, respectively; p = 0.002). Arm differences were significant (p = 0.038). GLS increased at 3-mos (5.7%; p < 0.001), at 6-mos (7.8%; p < 0.001) and at 12-mos (7.1%; p < 0.001) respect to T0 in P; at 3-mos (2.7%; p = 0.002), at 6-mos (3.2%; p = 0.014), but not at 12-mos in R; no significant changes at 3-mos, a significant increase at 6-mos (2.7%; p = 0.035), at 12-mos (3.2%; p = 0.008) in B; no significant changes at 3-, 6-, and 12-mos in R+B. Arm differences were significant (p = 0.002). Both R+B and the R arms showed a withdraw rate of 7%, with a dose reduction rates of 21% and 17%, respectively. R arm showed a significant decrease on 3D-LVEF at 12-mos as compared to T0 and will be evaluate for closure.


Following the stopping rules, the closure of the R arm is required and the study will continue with 3 arms. At the interim analysis, a cardioprevention strategy significantly impact on subclinical heart damage.

Clinical trial identification


Legal entity responsible for the study

University of Florence.


Has not received any funding.


All authors have declared no conflicts of interest.