Abstract P5-13-03: Real-world clinico-genomic data reveal differences in genomic landscape associated with CDK4/6 inhibitors in HR+/HER2- breast cancer

Abstract

Abstract Purpose: Studies based on data from routine clinical practice (real-world data, RWD) benefit from larger patient numbers and are more representative of patient diversity than clinical trial studies. When combined with comprehensive genomic profiling (CGP), RWD may uncover the impact of therapies and patient characteristics on tumor genomic landscape. Here we aimed at assessing the feasibility of using RWD to identify changes in the prevalence of genomic alterations upon treatment and proposed a methodology to address RWD inherent caveats. Experimental Design: To explore associations between tumor genomics and treatment chosen by physicians, we evaluated data from 5323 patients with metastatic hormone receptor-positive HER2-negative (HR+/HER2-) breast cancers from a nationwide real-world clinico-genomic database, originating from approximately 280 US cancer clinics (~800 sites of care). To perform our comparisons, we defined groups based on the therapy administered in the metastatic setting and the timing of the CGP relative to treatment. We used bootstrapping to estimate the significance of the effect and stratified analyses to assess the impact of potential confounders such as the site of the collected samples or disease history. Results: ESR1 alteration prevalence increased from 5.6% (CI: 2.8-8.9) pre-treatment to 21.4% (CI: 13.3-29.6) following aromatase inhibitor. Yet, it was significantly less than the prevalence following treatments including CDK4/6 inhibitors (CDK4/6i; 37.1% [CI: 27.8-46.4]; P=0.006). Further, exposure to CDK4/6i led to an increase in FGFR1 and TP53 alterations as well as genes of the cell cycle (FDR< 0.2). Overall, we found that more pathways were likely to be altered in a given tumor following AI+CDK4/6i than after AI alone (P=0.02). In particular, alterations of the MAPK pathway were not exclusive to ESR1 alterations in the post-AI+CDK4/6i group compared to AI only, suggesting that MAPK pathway alterations alone may not overcome CDK4/6i-based treatments. Differences following exposure to CDK4/6i were retained in samples taken after the second-line treatment. Stratified analyses confirmed that these results are independent of exposure to adjuvant therapy or treatment duration and showed that ESR1 mutations occurred in both primary and metastatic samples. Conclusions: Analysis of EHR-derived clinical data linked to CGP results from routine care can replicate associations previously observed in clinical trials and uncover unknown changes in tumor genomic landscape. Bootstrapping and stratified analysis reinforced our confidence in the results and thus allowed us to identify that CDK4/6i exposure led to a different – more altered – genomic landscape in HR+/HER2- breast cancer patients. This finding can inform design of clinical trials post-CDK4/6i and may help guide treatment choice for late stage patients. Thus, our work demonstrates the feasibility of leveraging real-world clinico-genomic data for translational research in oncology and leads the way for analyses including a more diverse patient population. Citation Format: Nayan Chaudhary, Ciara Mecalfe, Marc Hafner. Real-world clinico-genomic data reveal differences in genomic landscape associated with CDK4/6 inhibitors in HR+/HER2- breast cancer [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr P5-13-03.