Caris launches 3-day AI breast cancer test to predict long-term recurrence

(P1) Caris Life Sciences (NASDAQ: CAI) on Tuesday launched an AI-powered test that predicts both early and late-stage breast cancer recurrence risk from a single tissue sample at the time of diagnosis. The test, Caris MI Clarity, provides results in three business days, aiming to replace a slower, fragmented system and give clinicians a more complete picture for personalizing patient treatment.

(P2) "Breast cancer clinicians have long been forced to make some of the most consequential treatment decisions with an incomplete picture of how recurrence risk changes over time," George W. Sledge, Jr., MD, Chief Medical Officer at Caris, said in a statement. "Caris MI Clarity is designed to bring early and late distant recurrence risk together in a single test at diagnosis, when treatment decisions matter most."

(P3) The test is designed for postmenopausal patients with HR-positive/HER2-negative, node-negative early-stage breast cancer. It analyzes digitized pathology slides to provide distinct risk stratification for distant recurrence in both an early (0–5 years) and late (5–15 years) window. The AI model was validated using patient samples from the major TAILORx and NSABP B-42 clinical trials, according to the company.

(P4) By assessing long-term risk from the outset, the test could help thousands of patients avoid years of unnecessary hormone therapy or, conversely, ensure high-risk patients receive more aggressive treatment. For Caris, it marks a significant push into the competitive oncology diagnostics space, leveraging its AI platform against established genomic tests. The company's stock, which closed at $18.69 on Monday, has traded below its 200-day moving average of $26.47, and this launch could be a key factor for investors evaluating its growth prospects.

A New Approach Without Sequencing

Unlike many competing prognostic tools that rely on genomic sequencing, Caris MI Clarity uses AI to analyze routine hematoxylin and eosin (H&E) stained tissue slides. This approach means the test can be performed on standard pathology specimens already collected during diagnosis, with no need for additional, more complex tissue handling or sequencing.

This key difference could allow for faster, more scalable adoption in clinical settings worldwide. The company states the model was trained on its proprietary database of tens of thousands of breast cancer cases. While Caris is expanding its AI-driven oncology portfolio, the announcement did not include specific performance metrics such as hazard ratios or sensitivity, which will be critical for clinical adoption.

Clinical and Competitive Implications

The launch positions Caris to compete with established players in the breast cancer recurrence market, such as Veracyte, Inc. (NASDAQ: VCYT), which markets the Prosigna Breast Cancer Prognostic Gene Signature Assay, and Agendia NV. Both offer genomic tests that provide prognostic information, but Caris's new tool is the first to specifically delineate between early and late recurrence risk from a single test at diagnosis.

Accurately distinguishing between these risk periods is crucial. The biological drivers of recurrence within the first five years often differ from those that cause cancer to return 5 to 15 years later. Providing clinicians with this dual-risk perspective from the start could refine decisions about the duration and intensity of adjuvant endocrine therapy, potentially sparing some women from extended treatment and its associated side effects while identifying others who may benefit from it.

This article is for informational purposes only and does not constitute investment advice.