Protein Patterns Reveal Hidden Aggressive Breast Cancers
A new study led by researchers at CCBIO shows that protein patterns in the breast tumor microenvironment can improve how patients with hormone receptor鈥損ositive breast cancer are stratified by risk. The findings may help explain why some patients experience aggressive disease despite being classified as low-risk using today鈥檚 standard methods.
Published:
The paper, in press in npj Breast Cancer, is a collaboration with researchers at Karolinska Institutet and Karolinska University Hospital in Stockholm.
A protein signature that adds new clinical information
Using proteomics, Finne et al. identified a 35-protein tumor microenvironment signature that separates patients with breast cancer into distinct prognostic groups, independent of current molecular subtypes, which are largely based on tumor cell characteristics.
鈥淲e found that patients who are classified as low-risk using standard tumor markers may have a high-risk tumor microenvironment and experience aggressive disease,鈥 says Kenneth Finne, first author of the study and researcher at CCBIO. 鈥淭his adds a layer of information that is not captured by existing classification systems.鈥
Identifying high-risk disease in luminal A breast cancer
The strongest clinical impact of the protein signature was seen in luminal A breast cancer, the most common subtype and one that is usually associated with a favorable prognosis.
鈥淎cross several large cohorts, patients with luminal A breast cancer and high levels of the signature proteins had significantly worse survival,鈥 Finne explains. 鈥淭his shows that not all luminal A tumors are biologically alike, even though they are treated as a single group in clinical practice.鈥
In these patients, the protein signature was linked to biological features typically seen in more aggressive, high-risk tumors, despite otherwise favorable tumor characteristics.
Robust validation and potential clinical relevance
The researchers validated the protein signature across multiple independent datasets, including large international cohorts and a randomized clinical trial with more than 1,000 hormone receptor-positive patients followed for up to 20 years. In all settings, high signature scores were associated with poorer survival, even after adjustment for established prognostic factors such as tumor size, grade, lymph node involvement, and molecular subtype.
The researchers also showed that the full 35-protein signature could potentially be reduced to smaller protein panels, down to as few as five proteins, while retaining strong prognostic power. This opens the door to future clinical implementation.
While further validation is needed before clinical use, the study demonstrates that protein-level information can meaningfully improve breast cancer stratification.
鈥淭his study shows that we can extract clinically relevant protein information from the tumor microenvironment,鈥 Finne concludes. 鈥淲e hope this work adds new insight and represents a step toward more precise and biologically informed breast cancer classification.鈥
See .