Ataraxis AI Laboratory Receives CAP Accreditation and a California Clinical Laboratory License, Opening Ataraxis Breast to the State

An inspection by practicing pathologists, and deeming authority from CMS

The release describes CAP accreditation as the outcome of an on-site inspection by practicing pathologists that covers clinical protocols, personnel qualifications, instrumentation, safety programs and laboratory management. Because the CAP program holds deeming authority from the Centers for Medicare and Medicaid Services, that inspection stands in for a CMS inspection under CLIA. The California license is a separate matter, since the state requires out-of-state laboratories to hold its own license before receiving specimens from California patients, and Ataraxis says clinicians in the state can now order the tests. The company's boilerplate states that all Ataraxis Breast tests are offered as laboratory-developed tests through a CLIA-certified, CAP-accredited laboratory, and the release makes no claim of FDA clearance or approval for any of them.

"With this approval, our first-in-class AI tools are now available for clinical use by the overwhelming majority of U.S. providers, including in California, where Ataraxis already has a large base of collaborators," said Jan Witowski, MD, PhD, co-founder and CEO of Ataraxis AI. "We are excited to partner with physicians across the state who are leaders in adopting new technologies, and to work with their patients to provide the best possible care."

Three outputs from one H&E specimen

Ataraxis Breast is a multimodal platform that combines digitized H&E pathology slides with clinical data and returns three outputs from one specimen, according to the release: an individualized recurrence risk, a predicted benefit from adjuvant chemotherapy, and a forecast of response to neoadjuvant therapy. The company's product pages add the operating details a pathology department would ask about. Tests can be ordered on the core needle biopsy, so before surgery; no additional tissue, stain or procedure is required; the AI output is signed out by board-certified pathologists; and the recurrence-risk and neoadjuvant predictions are offered across subtypes, including triple-negative and HER2-positive disease, while the chemotherapy-benefit test is intended for hormone receptor-positive patients. The company also states that no tissue is exhausted, since the H&E slide is scanned and returned to downstream testing.

The release says the platform was developed on more than 30,000 multimodal patient cases. The company's own product page gives a lower figure, 23,800-plus patients represented in its development and validation datasets, split roughly among North America (7,200-plus), Europe (5,700-plus), Asia Pacific (4,900-plus) and Latin America (6,000-plus). Ataraxis did not reconcile the two numbers in the material reviewed.

The evidence behind the recurrence test

The prognostic component has the most published support. The company's summary of its Nature Communications paper, published May 20, 2026, describes a validation in 8,161 patients across 15 cohorts with more than 50 academic collaborators. In a head-to-head comparison against the 21-gene assay in 858 patients, Ataraxis reports a C-index of 0.67 against 0.61 and independent prognostic value in multivariable analysis (HR 2.95, p below 0.001), with performance carried into triple-negative (C-index 0.71) and HER2-positive (C-index 0.67) disease, two settings the company describes as having no NCCN-recommended prognostic tool. At ASCO 2026 the company presented a secondary analysis of the phase III UNIRAD trial in which the test stratified 689 node-positive HR-positive, HER2-negative patients (HR 2.83, p below 0.01) and identified a clinically high-risk subgroup with 91% five-year disease-free survival on chemoendocrine therapy alone. An analytical validation of the RISK test appeared in Diagnostics on March 29, 2026, and a January 6, 2026 agreement with Unicancer covers further validation on phase III trial data. All of these figures are the company's own reporting of its studies.

Why This Matters to the APO|APE Reader

Artera walked the same route in October 2024, when it announced a California license for its ArteraAI Prostate test, so Ataraxis is the second H&E-based AI laboratory-developed test to clear the state's separate bar. For a pathology practice the operational fact is that the ordering pathway starts with a glass slide or a scan leaving the building and coming back with a signed report inside a day, which puts slide logistics on the critical path. The open evidence question is on the predictive side: for the chemotherapy-benefit and neoadjuvant-response tests launched this spring, the company's evidence page lists SABCS 2025 and ASCO 2026 abstracts, and no journal paper on the scale of the May Nature Communications study.