In Reply We appreciate the thoughtful review and statistical analysis of our results by Hasegawa and colleagues. The primary end point of efficacy in OPTIMIZE-2 (measured by the skeletal-related event [SRE]) was met using a predefined noninferiority margin of 10%, albeit the upper limit of the 1-sided 97.5% CI (9.8%) was very close to the noninferiority boundary of 10%. This predefined noninferiority margin was chosen on the basis of SRE data from the first year of treatment due to the absence of SRE data from the second year of treatment when the trial was designed,1 and it went through several levels of review and endorsement by clinical experts. From a practical perspective, as the predefined margins are narrowed, the sample size required increases, thus making the completion of the trial difficult. In addition, it is practically impossible to demonstrate complete equivalence, and all noninferiority trials have the possibility of demonstrating some reduction in efficacy between the 2 groups. The predefined margins, determined at the start of the trial, indicate that these differences were considered acceptable. For the purpose of interpretation of trial results from OPTIMIZE-2, we urge readers to use their own judgment to decide whether the potential differences in efficacy are acceptable or not.
Gabriel N. Hortobagyi, Ming Zheng. Zoledronic Acid Dosing in Patients With Metastatic Breast Cancer—Reply. JAMA Oncol. Published online November 30, 2017. doi:10.1001/jamaoncol.2017.3868