Using causal artificial intelligence and digital pathology, the test moves beyond population-level averages to predict patient-specific chemotherapy outcomes.

Ataraxis AI has launched Ataraxis Breast CTX, a predictive artificial intelligence (AI) test designed to estimate individualized chemotherapy benefit in breast cancer patients. The company says it is the first tool of its kind to generate patient-level predictions of adjuvant chemotherapy benefit, distinguishing it from existing genomic assays that estimate recurrence risk based on population averages.

The test operates within the Ataraxis Breast platform, which combines digitized hematoxylin and eosin (H&E) pathology slides with clinical data to generate prognostic and predictive outputs. Ataraxis Breast CTX specifically predicts patient-level outcomes under two treatment strategies—with and without adjuvant chemotherapy—enabling a side-by-side comparison of expected benefit for each individual patient.

How the Technology Works

Central to the test is Ataraxis Tau, the company’s causal inference layer. Ataraxis Tau models how therapies influence outcomes at the individual patient level while adjusting for underlying risk factors. By separating prognosis from treatment effect, the platform aims to account for treatment selection bias and confounding that are commonly present in retrospective and real-world datasets—a methodological challenge that has historically complicated the interpretation of observational oncology data.

The test was developed using multimodal data from more than 10,000 breast cancer patients worldwide and independently validated across observational real-world studies and randomized controlled trials, according to the company. It is offered as a laboratory-developed test through a Clinical Laboratory Improvement Amendments-certified laboratory, meaning it integrates into existing pathology workflows without requiring additional tissue or specialized assays beyond the standard H&E slide.

“For too long, chemotherapy decisions in breast cancer have been based on how a treatment works on average, not whether it will actually benefit the specific person sitting across from you,” says Jan Witowski, MD, PhD, chief executive officer and co-founder of Ataraxis AI, in a release. “Ataraxis Breast CTX was built to change that. It is a first-in-class test, not just best-in-class, because no tool has estimated individualized chemotherapy benefit this way before.”

Addressing Gaps Left by Existing Tools

In early-stage breast cancer, treatment decisions are typically guided by recurrence risk stratification. Current tools—including standard clinicopathologic factors and genomic assays—provide prognostic information but are generally not designed to predict whether a specific patient will benefit from a given treatment. Therapy selection has largely been extrapolated from how treatments perform on average in similar patient populations.

Ataraxis Breast CTX is positioned as the predictive complement to Ataraxis Breast RISK, an AI-based prognostic test already within the platform. RISK is validated for use across all molecular subtypes, including HER2-positive and triple-negative breast cancer patients—populations the company notes are often not addressed by legacy genomic tools. Together, RISK and CTX are designed to support a more individualized approach to treatment planning by separating baseline prognosis from estimated treatment effect.

Implications for Overtreatment

One clinical application highlighted by Ataraxis AI is the potential to identify patients who are high-risk for recurrence but unlikely to derive meaningful benefit from adjuvant chemotherapy—a subgroup that, under current average-based frameworks, may receive treatment without corresponding benefit. The company says CTX can support more informed decision-making in these cases and help clinicians consider alternative strategies when appropriate.

Since its launch, the test has been adopted by National Cancer Institute-designated cancer centers and community clinics across the US, according to the company.

“CTX is the first in what we’re building as a new generation of predictive tools, each one bringing us closer to our vision of fully personalized cancer care,” says Witowski, in a release.

The broader Ataraxis Breast platform was developed using more than 20,000 multimodal patient cases. Ataraxis AI says the platform is designed to fit into existing pathology workflows, a consideration of particular relevance to clinical laboratories and digital pathology operations evaluating new AI-enabled testing capabilities.

ID 1465986 | Lab © Wa Li | Dreamstime.com

Related Reading: