Nouveau test de dépistage du cancer du sein: un schéma caché lié aux chances de survie

A groundbreaking method for diagnosing and grading breast cancer has emerged from research conducted at Columbia University, offering potential improvements in predicting patient outcomes and tailoring treatment strategies. Scientists have developed a technique that transforms visual patterns in tissue samples into measurable data, aiming to enhance the precision of breast cancer assessments. This innovation could significantly influence how doctors evaluate the severity of the disease and select appropriate therapies. The current diagnostic process involves examining tissue samples under a microscope to assess cellular changes. Cancer cells receive grades based on how closely they resemble healthy cells, which helps determine prognosis and suitable treatments. A lower grade typically indicates slower growth and reduced likelihood of spreading, whereas a higher grade suggests faster progression or existing metastasis. However, traditional methods often exhibit variability in accuracy across different ethnic groups, potentially affecting treatment efficacy. Researchers have employed mathematical tools called topology to create biomarkers that quantify the organizational structure of breast cancer tissues. According to a press release, these numerical scores demonstrated superior prediction capabilities regarding patient survival and treatment responses compared to conventional biomarkers. Moreover, the new approach showed consistent performance across diverse racial and ethnic populations. Dr. Kevin Gardner, pathologist-in-chief at NewYork-Presbyterian/Columbia University Irving Medical Center, emphasized the integration of digital pathology, artificial intelligence, and machine learning in understanding breast cancer. By merging detailed tumor images with genetic and protein data, the team aims to construct a comprehensive view of each patient’s condition. Their goal is to refine diagnostic accuracy, forecast disease progression, and facilitate personalized treatment choices. The study analyzed over 550 breast cancer cases from North Carolina, mapping the spatial relationships between tumor cells and immune cells. Models measured organizational patterns within the tissue, revealing that higher topology scores correlated with improved survival rates and better outcomes. These findings surpassed conventional grading methods in effectiveness. The research highlighted that topology-based biomarkers maintained predictive power across non-Hispanic Black and non-Hispanic white patient groups. Additionally, the methodology successfully anticipated therapeutic responses in separate clinical trials. Jasmine McDonald, Ph.D., a study co-author and associate professor of epidemiology at Columbia, noted that the spatial arrangement of tumors holds critical biological insights that traditional markers might overlook. Further analysis linked low topology scores to specific biological pathways related to metabolism, immune suppression, and epithelial-to-mesenchymal transition, a process connected to cancer invasion and metastasis. This suggests that structural alterations in tumors may reflect underlying metabolic and immune dynamics within the tumor environment. Researchers envision integrating these novel techniques into existing pathology practices to inform treatment decisions. They are currently exploring applications of similar topological methods on standard pathology slides used globally in clinical settings. The ultimate aim is to digitize tissue samples for broader accessibility and application in medical practice worldwide. The implications of this research extend beyond immediate diagnostic improvements. By addressing disparities in biomarker accuracy among different populations, the new method could contribute to more equitable healthcare outcomes. As the technology develops, its adoption in routine clinical practice may lead to more effective and targeted cancer treatments tailored to individual patient profiles.