Abstract Background: A CDK4/6 inhibitor (CDK4/6i) in combination with endocrine therapy (ET) is standard first-line therapy for advanced, hormone receptor (HR)-positive, HER2-negative breast cancer (BC). However, not all patients respond and responders eventually develop drug resistance and disease progression. Exosomes are small extracellular vesicles that are secreted by both normal and tumor cells as a mechanism for intercellular communication. The protein cargo of exosomes reflects biological processes activated in cancer cells and may serve as predictive biomarkers to select patients most likely to benefit from treatment and to identify mechanisms of resistance. Methods: Whole blood was collected in Streck tubes at baseline and at time points during treatment from patients with advanced, HR+/HER2- BC enrolled in a single arm, phase 2 trial of first line therapy with palbociclib plus tamoxifen that was conducted by The Big Ten Cancer Research Consortium (NCT02668666). Plasma was separated and stored at -80C within 48 hours of collection. Different exosome protein isolation methods were evaluated and optimized to maximize protein recovery. Exosome and plasma proteins were extracted, purified, and digested with trypsin. Tryptic peptides were isotopically labeled with Tandem Mass Tag (TMT) 10plex for protein expression level quantitation. Triplicate samples from each patient were analyzed by LC-MS/MS with QExactive HF Orbitrap mass spectrometer. An unsupervised clustering method was used to classify patients based on exosomal proteomic profiles. Results: We developed a sensitive and efficient exosome extraction method to obtain exosome protein from minimal volumes of patient plasma. The optimized exosome isolation method quantitatively identified 800 proteins from a 100 µl plasma sample. Significant enrichment of exosome-specific markers was observed when comparing patient samples with healthy donor samples. A network model was developed to differentiate responders/stable disease patients from non-responders using exosome proteomics data generated from pretreatment plasma samples. Preliminary data from the first 22 patients analyzed (responders, n= 6; stable disease, n=12, and non-responder, n=4) identified a network of 45 proteins that predicted response/stable disease vs progressive disease with high specificity (95%) and sensitivity (89%). We also noted significant differences in the exosome proteomic profiles of patients with de novo vs. recurrent metastatic disease. A network of 22 proteins differentiated de novo vs recurrent metastatic disease with > 85% sensitivity and 78% specificity, providing molecular evidence differentiating the two disease states. This finding is relevant in light of the higher response rate and improved PFS in patients with de novo metastatic disease in this trial, and confirms that this approach may provide molecular insight into mechanisms of primary resistance to CDK4/6i. Results for the entire trial cohort of 46 patients will be presented, along with analysis of serial samples collected at various time points during treatment. Conclusion: This proof-of-concept study demonstrates that an ultrasensitive exosome proteomics platform combined with deep learning methods is ideally suited for developing predictive protein biomarkers and for exploring molecular mechanisms of drug resistance. If results are confirmed, this novel approach holds great promise for identifying protein biomarkers that could be used to select patients unlikely to respond to ET and CDK4/6i in order to spare them ineffective treatment and for selecting participants for clinical trials of novel agents. Additionally, exosome proteomics data generated from serially collected specimens can be used to identify mechanisms of resistance that emerge during therapy. This approach can be widely applied to other treatment regimens and disease sites. This study was funded by Pfizer. Citation Format: Ziwei Zhang, Xiuyuan Ma, Julia Ekiert, Gayatry Mohapatra, Louis Coleman, Cristina I. Truica, Anne Blaes, Jatin Rana, Tandra Pavankumar, Lauren Green, Menggang Yu, Deborah Toppmeyer, Ruth O’Regan, Kari B. Wisinski, Oana C. Danciu, Kent Hoskins, Yu Gao. Quantitative proteomic analysis of plasma exosomes from patients with advanced hormone receptor-positive/HER2-negative breast cancer receiving palbociclib and tamoxifen [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 P5-02-47.
Read full abstract