FDA Clears AI Tool That Estimates Early Breast Cancer Recurrence Risk From MRI Scans

According to CURE, the company SimBioSys announced on Oct. 15, 2026, that the FDA had cleared TumorSight Risk. The clearance came through the 510(k) pathway, which is used for devices the agency finds substantially equivalent to one already on the market. This is software, not a drug, and it does not treat cancer.

The tool is cleared for adult women with hormone receptor-positive, HER2-negative breast cancer, stage 1 through 3A. That is the most common type of breast cancer. These cancers have not spread beyond the breast and nearby lymph nodes.

The software uses a breast MRI, the pathology report from the tumor and clinical details. According to the company, it needs no extra procedure, no tissue shipped to a lab and no special equipment, and a result can be ready before surgery. It gives a score from 0 to 100, a high, intermediate or low risk group, and an estimate of the chance of staying cancer-free at five and 10 years. The cleared use says the result should be considered alongside other clinical and pathology factors, not on its own.

The evidence comes from a study of 2,129 patients treated at Mayo Clinic, Emory University, Stanford Medicine and Mercy Medical Cedar Rapids. It was designed in advance and analyzed blindly, but it looked back at patients who had already been treated. The results were presented at the 2026 American Society of Clinical Oncology Annual Meeting.

Patients in the high-risk group had 3.3 times the risk of recurrence of those in the low-risk group. The estimated five-year recurrence chance was 2.2% for the lowest scores (0 to 10) and 13.6% for the highest (58 to 100). At 10 years, the estimates ran from 6.2% to 34.2%. Each 10-point rise in score was tied to a 22% higher relative risk. That link held after accounting for stage, grade, age, race and treatment.

Several limits matter. Because the study looked backward, it did not test whether using the score to guide treatment leads to better outcomes. The announcement did not compare the tool with the genomic tests doctors already use for this type of cancer. It did not give a release date, a price or insurance details. Not every patient with early breast cancer has a breast MRI, so the tool will not apply to everyone. According to the company, Mayo Clinic, one of the study sites, has a financial interest in SimBioSys.

For patients, these gaps point to open questions. Is a breast MRI part of my workup? How would this score sit alongside tumor size, lymph node findings, grade and any genomic test? Would a different result change what is recommended, such as whether chemotherapy adds benefit to hormone therapy? Those decisions still rest on the full picture of the cancer and a person's preferences, according to CURE.

What to watch next

Studies that test whether acting on the score improves outcomes, head-to-head comparisons with genomic tests, and word on cost and coverage.