Dynamic Monitoring and Identification of Reliable Biomarkers to Predict Efficacy of Bireociclib and Fulvestrant: An Exploratory ctDNA Analysis of the BRIGHT-2 Study

Background Bireociclib, a novel cyclin-dependent kinase 4 and 6 (CDK4/6) inhibitor, has demonstrated efficacy in hormone receptor (HR)-positive, human epidermal growth factor receptor 2 (HER2)-negative advanced breast cancer. This exploratory analysis aimed to identify biomarkers of response and resistance to bireociclib through dynamic circulating tumor DNA (ctDNA) profiling. Methods In this exploratory analysis of the phase Ⅲ BRIGHT-2 trial, plasma samples were collected at baseline, on-treatment, and end-of-treatment from patients randomized 2:1 to receive bireociclib plus fulvestrant or placebo plus fulvestrant. ctDNA was extracted and profiled using a 1,021-gene targeted sequencing panel. Somatic single-nucleotide variants, insertions/deletions, and copy number alterations were identified, and the molecular tumor burden index (mTBI) was calculated. Associations between baseline and longitudinal ctDNA features and clinical outcomes were assessed using Kaplan–Meier estimates, Cox regression, stratified log-rank tests, and treatment–biomarker interaction analyses with false discovery rate adjustment. Results A total of 596 plasma samples were successfully sequenced across three time points. The most frequently altered genes were PIK3CA (48%), TP53 (37%), and ESR1 (23%), defining the mutational landscape of East Asian HR+/HER2 − advanced breast cancer. Baseline ctDNA negativity or low mTBI, together with on-treatment ctDNA clearance or mTBI reduction, identified patients most likely to benefit from bireociclib. Within the bireociclib-treated cohort, PIK3CA–ESR1 co-mutation was associated with significantly prolonged PFS (p = 0.003) and OS (p = 0.024), and longitudinal clearance of these mutations conferred outcomes exceeding those of persistently negative patients. Mechanisms of acquired resistance to bireociclib were heterogeneous, involving activation of the PI3K/AKT pathway, amplifications of FGFR1 , MDM2 , and MYC , as well as upregulation of MAPK signaling. Conclusions Both baseline ctDNA features and on-treatment ctDNA dynamics demonstrated clinically relevant predictive value for treatment response and survival outcomes of bireociclib. These findings provide a molecular framework for patient stratification and highlight dynamic ctDNA monitoring as a tool to guide precision use of CDK4/6 inhibitors.