Prediction of dose reductions and impact of reduced starting doses of CDK4/6i on effectiveness in HR-positive/HER2-negative metastatic breast cancer: Real-world data from the OPAL registry.

Cancer patients prone to toxicities might benefit from dose reduction over fixed-dose recommendations. We develop a predictive index to identify patients with increased probability of dose reduction, intolerable toxicities, or therapy discontinuation (hereafter: dose reduction) in metastatic breast cancer (MBC) and compare real-world effectiveness of reduced (RSD) versus full starting dose (FSD) using this index. This analysis included 618 patients with HR-positive, HER2-negative MBC from the prospective, observational, multicenter registry OPAL (NCT03417115), receiving first-line palbociclib (n = 386) or ribociclib (n = 232) plus endocrine therapy. A logistic regression model was employed to derive the predictive index. Inverse probability of treatment weighting was used to emulate a head-to-head comparison of RSD and FSD by analyzing progression-free (PFS) and overall survival (OS). Within 6 months, 215 patients (35%) underwent dose reduction, including 109 (51%) with RSD. Predictors for dose reduction were age ≥65 years and Charlson comorbidity index (CCI) ≥1. Among patients with increased probability of dose reduction (index ≥1: ≥65 years and/or CCI ≥ 1), median PFS and OS were 30.1 [21.7, 54.0] and 57.6 [40.0, NA] months with RSD vs. 29.3 [24.9, 32.0] and 43.1 [38.8, 50.3] months with FSD. For low-probability patients (index = 0: <65 years and CCI = 0), median PFS and OS were 17.6 [9.1, 29.8] and 32.9 [23.1, 40.9] months with RSD vs. 24.5 [19.8, 32.0] and 54.2 [48.8, NA] months with FSD. In this real-world MBC setting, patients ≥65 years and/or with CCI ≥ 1 had an increased probability of dose reduction and may benefit from RSD, as this yielded outcomes comparable to FSD. Younger, fitter patients may require full dosing. Future studies, ideally randomized controlled trials, should aim to confirm these findings.