Background: Cyclin dependent kinase (CDK) 4/6 inhibition in addition with endocrine therapy (ET) is first line therapy for patients with hormone receptor-positive (HR+), HER2 non-amplified, metastatic breast cancer (mBC). Recognizing which patients are going to benefit from treatment and identifying which patients acquire resistance are still poorly understood. Molecular alterations in CCNE1, MYC, and RB1 from breast tissue samples have been associated with resistance to CDK4/6 inhibitor therapy. Circulating tumor DNA (ctDNA) offers a unique noninvasive method to capture mBC heterogeneity and treatment response. Studies have used ctDNA to monitor for emergence of mutations and monitor dynamic response to treatment in HR+ mBC. Here, we test the hypothesis that CCNE1, MYC, and RB1 can be detected in ctDNA and utilized as a biomarker to predict response to CDK4/6 inhibitor therapy.

Methods: In this study, we analyzed a subset of patients from the Dallas Metastatic Breast Cancer Study comprised of patients with HR+, HER2 nonamplified, mBC who underwent treatment with a CDK4/6 inhibitor (palbociclib, ribociclib, abemaciclib) and ET, became resistant to therapy, and had ctDNA testing (n=102). Commerical testing results through Tempus xF and FoundationOne Liquid CDx, which detects cancer-relevant genetic alternations, were obtained and analyzed. The data was collected from multiple hospitals within a single academic medical center starting with the initial ctDNA collection from 2019 to 2024. 13 patients had ctDNA collected before the start of therapy, 84 patients had ctDNA collected after progression, and 5 patients had ctDNA collected before therapy and after progression.

Results: 88 of 102 patients did not have CCNE1, MYC, or RB mutations present had a median progression-free survival (mPFS) of 13.5 months. Among the remaining 14 patients, two had CCNE1 alterations with a mPFS of 5.5 months: one had ctDNA collected before treatment and one had ctDNA collected after progression. Three patients had MYC alterations with a mPFS of 19 months: one had ctDNA collected before treatment and after progression, and two had ctDNA collected after progression. Nine patients had RB1 alterations with a mPFS of 7 months: four had ctDNA collected before treatment and after progression, and five had ctDNA after progression. Of the patients that had both timepoints collected, one patient had acquired a MYC alteration at time of progression. Four patients had acquired RB1 alterations at time of progression. Among patients with RB1 mutations (n=9), 3 had loss of function mutations, 1 had a missense mutation, 1 had low coverage region, 5 had multiple single nucleotide variants (SNVs) and among these, 2 had splice site variants.

Conclusion: There is an urgent need to develop predictive biomarkers that capture real-time changes in tumor biology. Tissue analyses from patients resistant to CDK4/6 inhibitors have demonstrated a 14%-28% expression of CCNE1, 16% of MYC mutations, and 9%-10% expression of RB mutations. In our study, we observe a lower rate of detection in ctDNA of each gene. This result suggests that correlation between tissue mutations and detectable ctDNA mutations may not occur in a linear fashion. This analysis further suggests that new research is needed to identify other mutations detected in ctDNA which could predict CDK4/6 inhibitor resistance beyond CCNE1, MYC, and RB. Broadening our understanding of other correlative mutations in ctDNA with therapy resistance could allow for improved clinical implementation. Our study was limited by the small sample size and lack of consecutive ctDNA collection. Further studies are needed to determine the temporal and dynamic changes of acquired tumor mutations detected through ctDNA and the predictive value of specific mutations on outcomes.

VIEW THE PUBLICATION