|
Abstract INTRODUCTION
Breast cancer remains a leading global health concern, with 2.26 million new cases and nearly 685,000 deaths reported in 2020 (Globocan). Despite advances in therapeutic strategies, disease relapse remains a critical challenge, affecting about 30% of patients. This high recurrence rate highlights the need for a deeper understanding of breast cancer’s dynamic progression and treatment responses. At the heart of this challenge lies tumor heterogeneity—specifically, the presence of genetically related but functionally distinct subpopulations within tumors. These heterogeneous tumor clusters exhibit unique molecular features, resistance mechanisms, and metabolic behaviors that can alter disease trajectory and therapeutic effectiveness.
Breast tumor heterogeneity manifests as a dynamic system where subpopulations interact within the tumor microenvironment (TME), influencing disease evolution and drug resistance. These clusters may compete, cooperate, or coexist, with inter-cluster communication enabling resistant clones to survive and support others. This complexity can result in partial treatment success and residual disease, contributing to relapse and metastasis.
To conceptualize this behavior, we apply the notion of “biological psychohistory,” inspired by Isaac Asimov’s Foundation series. This model proposes that breast cancer follows an inherent, semi-predictable evolutionary path shaped by molecular origins and environmental interactions, rather than being dictated solely by subtype classification. Understanding this adaptive trajectory is key to disrupting disease progression through more personalized and dynamic treatment strategies.
METHODS
This study aimed to investigate breast cancer heterogeneity longitudinally, focusing on luminal, triple-negative, and HER2 subtypes. Using matrix-assisted laser desorption/ionization mass spectrometry imaging (MALDI MSI), we conducted a prospective and retrospective analysis to examine the spatial and temporal evolution of tumor heterogeneity.
MALDI MSI enabled high-resolution, spatially resolved imaging of tumor samples without the need for labeling, preserving histological context. An unsupervised segmentation algorithm was applied to detect distinct molecular clusters. Spatial proteomics, combined with nano-LC-MS/MS, allowed the profiling of subpopulations directly from tissue, identifying clone-specific reference proteins. These proteomic profiles were integrated with MSI cluster data through machine learning to construct a database of heterogeneous tumor clones. This database serves as a reference for evaluating tumor evolution, treatment efficacy, and potential druggable targets, while also predicting recurrence risk based on heterogeneity profiles.
RESULTS
The cohort, including luminal, triple-negative, and HER2-low cases, provided multiple biopsies over time. Three representative cases highlighted the dynamic nature of tumor heterogeneity over time. Patient 2 (HR+/HER2-low) showed increasing aggressiveness and treatment resistance through evolving proteomic profiles and pathway shifts. Patient 11 transitioned from HER2+ to HR+/HER2-low, reflecting a positive therapeutic response marked by reduced activity in aggressive pathways. Patient 18 (HR-/HER2-low) revealed significant heterogeneity across four metastatic sites, with emerging molecular clusters and subtype shifts suggesting common mechanisms between primary and metastatic tumors. These findings underscore the individualized and adaptive behavior of breast cancer.Importantly, the study underscored the critical role of the TME in shaping tumor evolution. Interactions between clusters fostered adaptive behaviors and resistance, forming a “breast cancer heterogeneity community” that influenced disease progression and treatment outcomes.
A co-segmentation analysis across all tumor tissues identified 11 unique clusters using the Silhouette criterion. Notably, these clusters were not directly tied to a specific breast cancer subtype or therapy. However, consistent heterogeneity patterns and evolution were observed over time, especially in Patients 2, 11, and 18. This indicates that while treatment and subtype contribute to tumor behavior, the trajectory of heterogeneity is more closely aligned with initial molecular characteristics and tumor adaptability.
CONCLUSION
This study demonstrates the pivotal role of tumor heterogeneity and the TME in breast cancer progression, therapeutic resistance, and recurrence. The findings support the concept of “tumor psychohistory,” wherein the tumor's evolution is guided by its foundational molecular identity and shaped by ongoing interactions within the TME.
Recognizing and characterizing this adaptive progression is essential for developing more effective, personalized therapies. A one-size-fits-all approach, based solely on subtype classification, overlooks the nuanced and evolving nature of tumor biology. By leveraging advanced spatial proteomics techniques like MALDI MSI and integrating them with machine learning, this study offers a powerful framework for tracking tumor evolution and identifying specific therapeutic vulnerabilities.
Ultimately, treating breast cancer effectively requires strategies that are as dynamic and adaptable as the disease itself. Personalized approaches that consider individual heterogeneity, spatial context, and molecular evolution are vital for improving therapeutic success and long-term patient outcomes.
|
= Discovery stage. (53.14%, 2025)
= Translation stage. (22.33%, 2025)
= Clinically available. (24.53%, 2025)
