A New Frontier in Early Diagnostics
Researchers in Israel have developed a novel blood-based diagnostic test that achieves 95% accuracy in detecting early-stage breast cancer, according to findings published on November 11, 2026, in the scientific journal Nature. The breakthrough, led by a collaborative team from the Weizmann Institute of Science, the Hebrew University of Jerusalem, and Kaplan Medical Center, offers a significant advancement over traditional mammography and invasive biopsies, particularly in cases involving dense breast tissue where standard imaging often struggles.
The study, which incorporates data from clinical monitoring at the Mayo Clinic, marks a pivot toward epigenetic mapping as a primary tool for cancer screening. By identifying specific cellular instructions—known as epigenetic tags—that regulate gene activity, the research team was able to detect cancer indicators that traditional DNA sequencing might miss due to low concentration levels in the bloodstream during early disease stages. Traditional early detection methods often struggle with low concentrations of circulating DNA, which can lead to high rates of false alarms. By shifting the focus to epigenetic signatures, the researchers have managed to improve both specificity and sensitivity.
Decoding Epigenetic Signatures
The research, directed by Prof. Efrat Shema at the Weizmann Institute and Prof. Guy Ron at the Hebrew University, utilized advanced single-molecule microscopy to map epigenetic changes across millions of nucleosomes. Key members of the research team included Dr. Noa Pert and Dr. Kristina Reinschneider from Prof. Shema’s laboratory, and Yishai Arieli from Prof. Guy Ron’s team. By analyzing surgical tissue samples, the team identified unique signatures that differentiate malignant cells from healthy ones. These signatures were processed by artificial intelligence models trained on blood samples from 77 breast cancer patients monitored for at least five years at the Mayo Clinic in Florida, alongside 30 healthy individuals. The resulting test demonstrated 95% accuracy in identifying localized tumors smaller than two centimeters, a critical threshold for improving patient survival rates.
Predicting Disease Recurrence
Beyond initial detection, the research team addressed the clinical challenge of cancer recurrence, which affects roughly 20% to 40% of patients following successful primary treatment. The investigation revealed a rare population of aggressive, invasive cells hiding within seemingly non-aggressive tumors. By tracking the epigenetic footprint of these aggressive cells in patient blood samples, the scientists achieved 81% accuracy in predicting the risk of future recurrence. This capability allows clinicians to identify high-risk patients who might otherwise be categorized as low-risk under standard clinical metrics, potentially enabling earlier intervention. Prof. Shema noted that the team is currently exploring therapeutic strategies aimed at reducing the plasticity of these aggressive cancer cells, effectively stripping them of the flexibility required to survive and spread. This collaborative effort also involved contributions from Sheba Medical Center and various research teams across the United States and Europe, signaling a potentially global shift in how clinicians monitor post-treatment recovery.
Context of AI in Screening
The innovation arrives as the medical community continues to explore the integration of artificial intelligence into breast screening workflows. While mammography remains the standard, recent feasibility studies, such as those published in Nature in March 2026 regarding AI systems in breast cancer screening, highlight that AI can match or exceed specialist radiologist performance, particularly in identifying interval cancers. The Israeli team’s blood-based approach complements these advancements by providing a non-invasive, high-accuracy alternative that does not rely on radiological imaging, thereby offering a multi-faceted approach to oncology. As the field moves toward more personalized screening, the ability to combine epigenetic liquid biopsies with AI-driven imaging analysis could fundamentally alter early detection and long-term prognosis for millions of patients worldwide. Future steps involve further clinical validation and the potential integration of these epigenetic markers into standard diagnostic panels.