Data from >75,000 patients underscore advances in AI-driven virtual genomics, Signatera™-guided treatment monitoring, and continued expansion of MRD and other capabilities
Natera, Inc. (NASDAQ: NTRA) a global leader in cell-free DNA and precision medicine, today announced that 20 abstracts, including two oral presentations, featuring data generated using its technologies will be presented at the 2026 American Association for Cancer Research® (AACR) Annual Meeting 2026, taking place April 17–22 in San Diego, CA.
These presentations, developed in collaboration with leading academic and clinical partners, span multiple tumor types and applications, including artificial intelligence, digital pathology, molecular residual disease (MRD) and real-world evidence (RWE).
“These presentations reflect both the scale of our platform and the pace of innovation we’re driving across oncology,” said Alexey Aleshin, M.D., corporate chief medical officer and general manager, oncology and early cancer detection at Natera. “From AI-enabled genomic insights to real-world evidence supporting ctDNA-guided treatment decisions, this body of work demonstrates how we are continuing to innovate within and beyond MRD to deliver more precise, data-driven approaches across cancer care.”
Key highlights
Artificial Intelligence
- Natera’s oral presentation, A large-scale, multi-target deep learning model for virtual genomic profiling in colorectal cancer,describes a digital pathology deep learning model trained on molecular and histopathology data from more than 45,000 colorectal cancer patients in a first-of-its-kind analysis. The model predicts risk of recurrence, hundreds of genomic alterations and key guideline biomarkers directly from routine H&E images, with an industry-leading area under the receiver operating curve (AUROC) of 0.98 for MSI-status and 0.93 for BRAF V600E status. This highlights the potential to transform routine pathology slides into a scalable, AI-enabled genomic profiling platform, expanding access to high-value molecular data, reducing tissue constraints and accelerating time to treatment decisions.
- An additional poster in breast cancer, A machine learning approach to classify breast cancer receptor subtype using genomic features, analyzed data from nearly 20,000 patients. The study concluded that a machine learning model using tumor DNA can accurately classify breast cancer into its major subtypes, including high precision of 93.5% and recall of 91.1% for identifying HR+/HER2- disease.
ctDNA Decision-Making
- A second oral presentation, Circulating tumor DNA (ctDNA) clearance dynamics in microsatellite instability-high metastatic (MSI-H) colorectal cancer (CRC) treated with immune checkpoint inhibitors (ICI), includes a real-world analysis of 465 patients with MSI-H metastatic CRC treated with ICI. The data shows that ~40% of patients had early ctDNA clearance on ICI. Signatera negativity at the first post-ICI timepoint was predictive of overall survival (OS), with 3-year OS of 96%. Moreover, MRD dynamics were strongly associated with improved OS with 65% of Signatera-positive patients achieving clearance anytime on or post-ICI.
Platform Expansion
- A poster on Natera’s methylation-based, tissue-free MRD platform entitled, Quantification of circulating tumor DNA (ctDNA) in patients using cancer-specific, methylation-based, tissue-free tests for the detection of molecular residual disease (MRD), shows that quantitative ctDNA levels strongly correlate with Signatera quantitative levels across colorectal, breast, lung and bladder cancers. LatitudeTM, Natera’s tissue-free MRD assay, launched in CRC in 2025 and is expected to launch commercially in additional histologies with quantitative features later this year.
- A poster in early cancer detection, Performance of a blood-based screening test for the early detection of advanced precancerous lesions, summarizes previously announced PROCEED trial data that showed strong performance of Natera’s blood-based screening assay for detecting advanced precancerous lesions. The study demonstrated sensitivity of 22.5% (CI: 15.4%-32.4%) and specificity of 91.5% (CI: 88.2%-93.9%) and encompassed all histologic subtypes, including serrated polyps.
A full list of abstracts is included here.
