Caris Launches AI Tool To Show When Chemo Helps—or Hurts—Lung Cancer Patients
Caris Life Sciences (NASDAQ: CAI) has introduced a multimodal, AI‑powered “immunotherapy treatment selection” signature for a subset of non‑small cell lung cancer (NSCLC) patients who are candidates for single‑agent immunotherapy. The tool is designed to identify which patients are likely to benefit from adding chemotherapy to immunotherapy—and which patients may actually be harmed by that extra chemo.
How the multimodal AI works
The new signature is built on Caris’ CodeAI platform, which combines Whole Exome Sequencing (WES), Whole Transcriptome Sequencing (WTS), and real‑world clinical data to generate proprietary “Caris AI Insights.” Using machine learning on large clinico‑genomic datasets, the model first stratifies NSCLC patients into higher‑ versus standard‑risk groups, then predicts whether adding chemotherapy to immunotherapy is likely to improve outcomes or reduce overall survival.
Why it matters for NSCLC care
For PD‑L1 high, driver‑negative NSCLC, physicians often choose between immunotherapy alone or chemo‑immunotherapy based largely on clinical judgment rather than deep tumor biology. Caris’ AI signature aims to make that decision more precise, helping avoid unnecessary chemotherapy—and its toxicity—for patients unlikely to benefit, while guiding higher‑risk patients toward more intensive combinations when they’re likely to help.
Part of a broader AI oncology strategy
This NSCLC tool joins a growing portfolio of Caris AI Insights, which already includes signatures like FOLFIRSTai for sequencing chemotherapy in metastatic colorectal cancer and risk tools for brain metastases in breast and lung cancer. Caris positions these WES/WTS‑based models as a step beyond small panel biomarker tests, arguing that large‑scale, multimodal data are needed to capture complex response patterns and deliver truly personalized oncology.
