Caris AI Tool Aims to Personalize Pancreatic Cancer Therapy

IRVING, TX – March 09, 2026 – By Deborah Cooper

Caris Life Sciences today announced the launch of a novel artificial intelligence tool designed to guide therapy for pancreatic cancer, one of the most challenging and lethal malignancies. The new Caris AI Insights™ signature aims to provide oncologists with molecular-level guidance for selecting the most effective first-line chemotherapy, a decision that has historically relied more on patient fitness than on the specific biology of their tumor.

The AI-driven signature, available through the company's Caris Molecular Tumor Board Report, analyzes comprehensive genomic and transcriptomic data to predict which patients are more likely to benefit from one of two standard-of-care regimens: FOLFIRINOX or gemcitabine/nab-paclitaxel. By personalizing this critical first step, the company hopes to improve outcomes and spare patients from the significant toxicity of treatments unlikely to be effective.

A New Weapon in a Difficult Fight

Pancreatic ductal adenocarcinoma (PDAC) carries a grim prognosis, with a five-year survival rate lingering in the single digits. For decades, clinicians have faced a difficult choice between two primary first-line chemotherapy cocktails for advanced disease. FOLFIRINOX is a more aggressive regimen often associated with better response rates but also with substantial side effects, while the combination of gemcitabine and nab-paclitaxel (gem/nab-p) is generally better tolerated.

Without reliable biomarkers to predict which regimen will work best for an individual, the choice often defaults to the patient's overall health and age, leaving tumor biology out of the equation. This can lead to situations where a patient endures the harsh side effects of FOLFIRINOX for minimal benefit, or conversely, receives a less intensive therapy when their tumor might have been highly susceptible to the more aggressive approach.

Caris's new tool seeks to fill this crucial information gap. It is built upon the company's MI Cancer Seek® assay, the first platform to receive FDA approval for simultaneous Whole Exome Sequencing (WES) and Whole Transcriptome Sequencing (WTS) for solid tumors. By analyzing the complete DNA and RNA landscape of a patient's tumor, the AI signature identifies complex molecular patterns that correlate with differential outcomes between the two standard therapies.

"Caris AI Insights for PDAC represents a meaningful step forward in bringing molecular intelligence to a disease where clinicians have historically had to make difficult treatment decisions with limited biological guidance," said David Spetzler, President of Caris Life Sciences, in a statement. "By harnessing the power of WES and WTS, this Caris AI Insights signature identifies complex molecular patterns that may predict differential benefit between standard first-line regimens."

The resulting report provides clinicians with a risk categorization and a clear treatment recommendation, supplemented by Kaplan-Meier survival curves showing how patients with similar molecular profiles have fared on each regimen. The tool also aims to identify a subset of patients who may be candidates for treatment de-escalation, potentially achieving similar results with a less toxic therapy.

Navigating the Competitive and Regulatory Landscape

Caris is not the only company deploying AI to tackle the complexities of pancreatic cancer treatment. The move places it in direct competition with other major players in the precision oncology space, most notably Tempus AI. Tempus has already validated its own RNA-based algorithm, PurIST®, which classifies PDAC tumors into "classical" or "basal" subtypes to guide the same first-line therapy decision. The growing competition highlights a significant market shift towards using advanced diagnostics to refine treatment strategies in historically difficult-to-treat cancers.

While the underlying MI Cancer Seek assay secured landmark FDA approval in late 2024, the regulatory path for the AI Insights signature itself is more nuanced. The signature is delivered as part of the Caris Molecular Tumor Board Report, which the company has previously described in other contexts as a "research-use-only" (RUO) product. If this designation applies to the new pancreatic cancer signature, it would mean the tool is intended to support research and clinical decisions but does not carry the formal diagnostic claims of an FDA-approved device.

This distinction is critical for clinical adoption and reimbursement. For the AI signature to become a routine part of care, Caris will need to convince both regulators and insurance payers of its clinical utility. The company has stated that a peer-reviewed publication detailing the signature's performance and its ability to identify candidates for treatment de-escalation is expected later this year. This forthcoming data will be essential in making the case for broader clinical integration and securing coverage from Medicare and commercial insurers, who require robust evidence before covering novel diagnostic technologies.

The Promise and Scrutiny of AI in Oncology

The launch represents a broader trend of integrating artificial intelligence into the fabric of cancer care. AI's ability to analyze vast, multimodal datasets—combining genomic, transcriptomic, and clinical information from hundreds of thousands of patients—offers the potential to uncover biological insights that are beyond human capacity to discern. Caris reports that its models were developed using a proprietary database of over 550,000 patients.

However, the increasing reliance on AI in medicine brings both promise and scrutiny. A primary concern within the oncology community is the "black box" phenomenon, where an algorithm provides a recommendation without a clear, interpretable explanation of its reasoning. For clinicians to trust and act on these insights, particularly when making high-stakes decisions, transparency is paramount. The format of Caris's report, which includes survival plots and is designed for discussion in molecular tumor boards, appears aimed at making the AI's output more actionable and understandable for clinical teams.

Furthermore, the ethical implications of AI are a subject of intense discussion. The quality and diversity of the training data are critical; models trained on homogenous patient populations can perpetuate or even worsen existing health disparities. Ensuring that these sophisticated tools are validated across diverse ethnic and demographic groups is essential for their equitable application. As AI-driven diagnostics become more powerful, questions of data privacy, algorithmic bias, and accountability will continue to be at the forefront of the conversation for developers, regulators, and clinicians alike.

The ultimate value of the Caris AI Insights signature will be determined not just by the sophistication of its algorithm, but by its real-world impact on patient care. The forthcoming clinical data will be a critical first step, but the true test will be its adoption by oncologists and its ability to deliver on the promise of more personalized, effective, and less burdensome treatment for patients battling pancreatic cancer.