Zhaohui Su, VP of Biostatistics at Ontada, shared a post on LinkedIn:
“This new exciting study published in JAMA Surgery introduces e19‑9, an AI-derived electronic biomarker that offers treatment insights for approximately 30% of pancreatic cancer patients who do not produce CA19‑9.
By using routine electronic health record (EHR) lab data and a machine-learning (ML) model validated across 58 health systems, Dr. Thalji and colleagues have developed a scalable, noninvasive alternative to a potential missing biomarker.
In patients who are CA19‑9 nonproducers, dynamic changes in e19‑9-particularly a ≥50% decline or a post‑treatment value of less than 100-strongly predicted treatment completion, metastatic progression, and overall survival, while CA19‑9 itself provided no significant signal.
The significance of this study lies in demonstrating how AI-enabled electronic biomarkers can transform everyday clinical data into actionable insights in oncology. This advancement enables adaptive care, expands trial eligibility, and brings precision medicine to patients who previously had limited options.”
Title: AI–Derived Electronic Tumor Marker For Cancer Antigen 19-9 Nonproducers With Pancreatic Ductal Adenocarcinoma
Authors: Sam Z. Thalji, Mohammed Aldakkak, Adhitya Ramamurthi, Taylor J. Jaraczewski, Mouloud Belbahri, Gopika SenthilKumar, Tahseen Shaik, Jennifer R. Merrill, Anjishnu Banerjee, Bradley W. Taylor, Mandana Kamgar, Ben George, Beth Erickson, William A. Hall, Nikki K. Lytle, Y. David Seo, Kathleen K. Christians, Callisia N. Clarke, Douglas B. Evans, Susan Tsai, Anai N. Kothari
Read The Full Article
Other articles about Pancreatic Cancer on OncoDaily.