Performance and Safety of a Multi-Cancer Early Detection Test: The PATHFINDER 2 Study

Authors

• Nima Nabavizadeh, Oregon Health & Science University
• Charles McDonnell, Sutter Health Research Institute
• Dax Kurbegov, HCA Healthcare Sarah Cannon Cancer Network
• Marc Matrana, Ochsner Clinic Foundation
• Shirish Gadgeel, Emory University
• Raymond H. Kim, Princess Margaret Cancer Center
• Gretchen Stipec, Long Beach Memorial Medical Center
• Kevin C. Oeffinger, Duke University Health System
• Michael J. Demeure, Hoag Memorial Hospital Presbyterian
• James W. Lillard Jr., Morehouse School of Medicine
• Rebecca Kaltman, Inova Fairfax Hospital
• Jennifer Tran, MedStar Washington Hospital Center/Georgetown University School of Medicine
• Alan Roth, Jamaica Hospital Medical Center
• Samia Tahhan, Macon & Joan Brock Virginia Health Sciences at Old Dominion UniversityFollow
• Andrews Poklepovic, VCU Health Stony Point
• Donald Richards, Texas Oncology
• Margarita Lopatin, GRAIL, Inc.
• Celine Marquez, GRAIL, Inc.
• Margaret McCusker, GRAIL, Inc.
• Karthik V. Giridhar, Mayo Clinic

Document Type

Article

Publication Date

2026

DOI

10.1038/s41591-026-04618-w

Publication Title

Nature Medicine

Volume

32

Issue

10

Pages

3760-3769

Abstract

Early cancer detection improves survival; however, most cancer types do not have guideline-recommended screening. We previously reported the first PATHFINDER study that established feasibility of cancer screening with a blood-based multi-cancer early detection (MCED) test that predicts cancer signal origin (CSO) to guide diagnostic workup. The prospective, interventional PATHFINDER 2 study expands clinical evidence for the MCED test by evaluating co-primary endpoints of MCED performance metrics and safety in participants aged 50 years or older without clinical suspicion of cancer. Performance measures included cancer detection rate, positive and negative predictive values, episode sensitivity, specificity and CSO prediction accuracy. Safety was evaluated in terms of number and type of invasive procedures performed and adverse events due to diagnostic testing, after 12 months of follow-up. Endpoints were analyzed descriptively; no hypothesis testing was performed. Between December 2021 and July 2024, 35,878 participants were enrolled. Among 32,007 performance-analyzable participants, cancer detection rate was 0.54% (95% confidence interval (CI): 0.47−0.63%), positive predictive value was 60.3% (95% CI: 54.5−65.8%), negative predictive value was 99.2% (95% CI: 99.1−99.3%) and specificity was 99.64% (95% CI: 99.57−99.70%). Episode sensitivity was 39.3% (95% CI: 34.9−44.0%; all cancers) and 69.8% (95% CI: 62.8−76.0%; prespecified 12-cancer subgroup). Positive and negative likelihood ratios were 108.9 (95% CI: 87.6−135.2) and 0.61 (95% CI: 0.56−0.66), respectively, and the number needed to screen to detect one cancer was 185 (95% CI: 159−215). CSO prediction accuracy was 91.3%. Among 35,335 safety-analyzable participants, 213 (0.6%) had invasive procedure(s) after a positive MCED test; 90.5% were nonsurgical. No serious study-related adverse events were reported by the time of this analysis. These results provide insights into performance and safety of the MCED test in an intended-use population. Randomized trials and longer follow-up are needed to assess the clinical utility of MCED tests. ClinicalTrials.gov identifier: NCT05155605.

Rights

© 2026 The Authors.

This article is licensed under a Creative Commons Attribution 4.0 International (CC BY 4.0) License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original authors and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

Data Availability

Article states: "Anonymized participant data will be available after the completion of the 3-year follow-up and database lock (anticipated December 2027). Requests for access to anonymized data may be emailed with a proposal to [email protected]. Investigators will assess proposals for scientific merit. Proposals should have a clearly specified research question supported by a methodologically sound analysis proposal and a prespecified statistical analysis plan, together with documentation of the requesting investigators’ qualifications and any applicable ethics and data protection approvals. Data will be provided after approval of the proposal and completion of a data transfer agreement. Decisions regarding the proposals will be communicated within 4−6 weeks."

Original Publication Citation

Nabavizadeh, N., McDonnell, C., Kurbegov, D., Matrana, M., Gadgeel, S., Kim, R. H., Stipec, G., Oeffinger, K. C., Demeure, M. J., Lillard, J. W., Jr., Kaltman, R., Tran, J., Roth, A., Tahhan, S., Poklepovic, A., Richards, D., Lopatin, M., Marquez, C., McCusker, M., & Giridhar, K. V. (2026). Performance and safety of a multi-cancer early detection test: The PATHFINDER 2 study. Nature Medicine, 32(10), 3760-3769. https://doi.org/10.1038/s41591-026-04618-w

Repository Citation

Nabavizadeh, N., McDonnell, C., Kurbegov, D., Matrana, M., Gadgeel, S., Kim, R. H., Stipec, G., Oeffinger, K. C., Demeure, M. J., Lillard, J. W., Jr., Kaltman, R., Tran, J., Roth, A., Tahhan, S., Poklepovic, A., Richards, D., Lopatin, M., Marquez, C., McCusker, M., & Giridhar, K. V. (2026). Performance and safety of a multi-cancer early detection test: The PATHFINDER 2 study. Nature Medicine, 32(10), 3760-3769. https://doi.org/10.1038/s41591-026-04618-w