SABCS 2025

Integrative modeling of multimodal real-world data for improved risk stratification of first-line CDK4/6 inhibitor outcomes in patients with estrogen receptor (ER)-positive/human epidermal growth factor receptor 2 (HER2)-negative metastatic breast cancer

We developed a machine learning model that used clinical, genomic, and transcriptomic features to stratify patients with ER-positive/HER2-negative metastatic breast cancer, based on response to first-line CDK4/6 inhibitor plus endocrine treatment and identify predictors of response. This study demonstrated that multimodal real-world data collected during routine care can provide valuable insights into the biology of response to CDK4/6 inhibitors in patients with metastatic breast cancer and help improve patient stratification.

Clinical and Molecular Landscape of ESR1 and PIK3CA Co-Mutated Hormone Receptor–Positive Metastatic Breast Cancer (HR+ MBC): Insights from 8,626 Patients Including Polyclonality and TP53 Alterations

Distinct mutation groups—no-MUT, PIK3CA-MUT, ESR1-MUT, and co-MUT (PIK3CA + ESR1)—exhibited unique clinical and genomic features; co-MUT was linked to more bone metastases, elevated tumor mutational burden, increased polyclonality, and shorter real-world overall survival. Notably, the frequency of ESR1 and co-mutations was higher in patients who underwent late molecular testing, indicating these alterations often emerge after extended endocrine therapy and highlighting the importance of timing in molecular assessment for guiding treatment strategies.

Distinct Transcriptional and Immunosuppressive Microenvironment Signatures in PIK3CA- and ESR1-Mutant Hormone Receptor Positive (HR+)/HER2- Metastatic Breast Cancer (MBC)

This study compared transcriptomic and immune profiles in HR+/HER2- metastatic breast cancer across wild-type, PIK3CA-mutant, ESR1-mutant, and co-mutant groups. SFRP2 downregulation was specific to ESR1-mutant tumors, while SCGB2A2 was robustly upregulated in PIK3CA-mutant and co-mutant tumors, suggesting its potential as a diagnostic and therapeutic target. Immune analysis revealed increased M2 macrophages and regulatory T cells in PIK3CA-mutant and co-mutant tumors, with the most pronounced immunosuppressive microenvironment in PIK3CA-mutant cases.

Treatment Sequencing in HR+ HER2- Metastatic Breast Cancer and Associated Real World Outcomes

This study investigates the impact of first-line and second-line therapy selections on real-world progression-free survival (rwPFS) and real-world overall survival (rwOS) in patients with HR+ HER2- mBC. CDK4/6i significantly outperforms taxane in first-line. Patients who received second-line CDK4/6i following first-line taxane had longer second-line rwPFS than those who received CDK4/6i first-line treatment. The findings suggest that taxane may cause sensitization to CDK4/6i, potentially conferring better survival outcomes in a subset of patients where first-line taxane is advised, or a potential advantage to chemo before CDK4/6i re-challenge.

Real-world second-line treatment use and clinical outcomes in patients with HR-positive/HER2-negative metastatic breast cancer and an ESR1 mutation

The study objective was to characterize real-world second-line treatment use and clinical outcomes (rwPFS and rwOS) in a metastatic breast cancer patient population with positive ESR1m test after first-line treatment with AI+CDK4/6i. Findings show that this cohort describes a complex subgroup of patients to treat, with rwPFS outcomes highlighting the need for more effective strategies that address resistance to AI+CDK4/6i and improve patient outcomes. Results from Flatiron Health EHR and Tempus both indicated consistent findings.

Clinicogenomic and Immune Profiles of Male Breast Cancer by Race: Insights from a Large Real-World Cohort

Real-world analysis of male breast cancer revealed broadly similar clinicogenomic profiles between Black and White patients, with HR+/HER2-negative and luminal subtypes predominating. Notable immune microenvironment differences—lower M2 macrophage infiltration and higher PD-L1 negativity in Black patients—may impact immunotherapy response and warrant further research into racial variation.

Cathepsin Protease Expression and Therapeutic Outcomes to Trastuzumab Deruxtecan (T-DXd) in Metastatic Breast Cancer

This study characterized pre-treatment expression of cathepsin proteases prior to T-DXd (cleavable linker) vs. T-DM1 (non-cleavable linker) treatment to understand their impact on clinical outcomes in metastatic breast cancer using the Tempus real-world database. This study demonstrated that among T-DXd treated patients, high cathepsin B and L protease expression was associated with improved outcomes in patients with HR+/HER2- disease, whereas in HER2+ patients, high expression of these proteases was associated with worse survival. In contrast, in the T-DM1 cohort there were no significant associations between protease expression and OS in the overall cohort as well as in the HER2+ cohort. This highlights the potential role of proteases as biomarkers of response to T-DXd in HER2 low/ultralow, but not HER2+ metastatic breast cancer.

Comparative analysis of the tumor immune microenvironment (TIME) and primary and metastatic tissue in HR+/HER2- and triple-negative breast cancer (TNBC)

Tempus Lens was leveraged to assess the Tumor Immune Microenvironment (TIME) across metastatic sites of disease in HR+/HER2- breast cancer compared to TNBC, finding significant differences by site in both subtypes, with HR+/HER2- cancer being less immunogenic overall. Lung, pleura, and peritoneum metastases, along with breast and lymph nodes, showed the highest CD8+ T cell proportions, suggesting a subset of HR+/HER2- patients with these sites may potentially respond to Immune Checkpoint Inhibitor (ICI) therapy. This high variability of TIME profiles across metastatic sites warrants further study in prospective trials to guide patient selection for ICI.

The Molecular and Immune Landscape of HER2 Positive Breast Cancer

Tempus Lens was employed to evaluate the tumor immune microenvironment of HER2+ breast cancer, assessing immune cell infiltration, TMB, PD-L1 status, and HLA allele prevalence, to uncover biomarkers for treatment guidance. A notable percentage of patients with localized and de novo metastatic disease displayed TMB high status and/or PD-L1 positivity. Additionally, TMB-high and PD-L1 positive patients with de novo metastatic disease treated with first-line chemotherapy or anti-HER2 therapy had significantly worse real-world overall survival (rwOS), suggesting a potential therapeutic benefit of incorporating immunotherapy into the treatment paradigm, in both localized and metastatic disease settings. Furthermore, observed ethnic HLA polymorphisms in the cohort may contribute to differences in outcomes and could potentially guide the development of population-specific immunotherapeutic strategies.

Real world data (RWD) outcome analysis of ESR1 mutation emergence in HR+/HER2- metastatic breast cancer through the continuum of standard of care hormonal therapy

This large multimodal RWD outcome analysis from longitudinal molecular surveillance testing (xF) in HR+/HER2- mBC pts treated with AI+CDK4/6i sheds light on the continuum of ESR1m emergence and patient outcomes from first-line and beyond outside of clinical trial data. Analyses show higher ESR1m incidence is associated with reduced survival regardless of line of therapy.

Experience live demos of our newest technologies for providers, biopharma, and research partners.

Stop by our booth to discover how we are leveraging AI-enabled technologies to advance precision oncology.

Tempus Hub

Streamline ordering and get quick access to patient information through our secure online platform. Place orders tailored to your patient’s diagnosis with one click, view comprehensive smart reports, and review patient medical history—all in one place.

Tempus One

Get real-time answers and uncover patient insights with our generative AI assistant built for healthcare providers and researchers. Just ask a question to access data-driven responses instantly.

Tempus Lens

Quickly define detailed patient cohorts using natural language processing and transform complex criteria into actionable datasets. Explore unstructured clinical data and extract nuanced insights. With Lens, the power of Tempus data is at your fingertips.

Tempus Next

Identify patients who may have deviated from care pathways and help providers close gaps in care, accelerating adoption of precision medicine.