medwireNews: The improved-Metabolic (i-Met) and minimalistic-Metabolic (m-Met) biomarker panels outperform carbohydrate antigen (CA) 19-9 alone in ruling out pancreatic ductal adenocarcinoma (PDAC) in high-risk patients, suggests the METAPAC study.
Relative to CA 19-9 alone, the study saw better specificity, negative predictive value (NPV), and diagnostic accuracy with the m-Met signature, which included CA 19-9, ceramide (d18:1, C24:0), lysophosphatidylethanolamine (C18:0), phosphatidylethanolamine (C18:0, C22:6), and sphingomyelin (d17:1, C16:0). The i-Met signature, which included the same parameters as the m-Met plus seven others including histidine and proline, also had better outcomes than CA 19-9 alone.
“These findings could offer a surveillance tool in patients with an annual risk of pancreatic ductal adenocarcinoma of 1% to reduce unnecessary invasive procedures and facilitate earlier detection of resectable disease,” write Julia Mayerle (LMU University Hospital, Munich, Germany) and colleagues in The Lancet Gastroenterology & Hepatology.
The prospective phase 4 trial enrolled 1129 patients with pancreatic lesions identified on computed tomography (CT) imaging that needed further diagnostic assessment. The investigators took blood samples from the patients and profiled specific metabolic markers using liquid chromatography–tandem mass spectrometry before following up for 24 months.
The participants were a median of 67 years old, had a median BMI of 25.2 kg/m², and 51% were men. A total of 27% had diabetes or glucose intolerance, and 49% had lost weight in the last 6 months.
A total of 43% of the participants had PDAC confirmed by histopathology and the remaining 57% had other conditions including chronic pancreatitis (10%), intraductal papillary mucinous neoplasms (21%), and other types of cystic lesions (24%).
The majority (72%) of PDAC cases were planned for surgical resection, and among the patients with staging information, 73% were staged between I and IIB as primary resectable PDAC.
Mayerle and colleagues found that i-Met and m-Met achieved the primary endpoint of the study, which was the exclusion of PDAC with at least 85% specificity and maximum diagnostic accuracy. The tests outperformed CA 19-9 alone at all stages of PDAC in terms of specificity with corresponding rates of 90.4% and 93.6% versus 79.1%, and in overall diagnostic accuracy with areas under the curve (AUC) of 0.846 each versus 0.799, respectively. The corresponding sensitivity values for the tests were 67.5% and 59.9% versus 81.8%.
The improvement in specificity with i-Met and m-Met versus CA 19-9 also applied to patients with resectable tumors (91.7 and 93.6 vs 45.6% , respectively), with greater specificity for detectable CA 19-9 at a cutoff of more than 10 U/mL (82.0 and 86.1 vs 79.2%), and patients with non-detectable CA 19-9 at a cutoff of less than 37 U/mL (98.1 and 98.8%).
The results were consistent when comparing between sexes and centers, indicating they are generalizable approaches, the researchers write.
The i-MET and m-Met tests also predicted PDAC significantly more accurately than CA 19-9 alone among patients with resectable tumors (AUCs of 0.859 and 0.861 vs 0.815, respectively), and among patients with detectable CA 19-9 (0.846 and 0.840 vs 0.736).
The i-Met and m-Met tests were also similar to CA 19-9 alone in terms of the NPV in the full cohort (78.4, 75.3, and 75.3%, respectively), and better among patients with resectable tumors (95.4 and 93.7 vs 80.6%), but worse among patients with detectable CA 19-9 (70.9 and 64.5 vs 86.1%).
To identify the diagnostic performance of m-Met as a low-cost test in people with new-onset diabetes, the investigators took data from the previous SHIP-TREND-1 study, where 4420 healthy people were followed up for 5 years, including 242 who developed new-onset diabetes and three who developed PDAC. Using plasma samples taken at least 12 months before PDAC diagnosis, the m-Met test excluding CA 19-9 significantly discriminated between people with new-onset diabetes with and without PDAC with an AUC of 0.840.
In an editorial related to the study, Julie Earl at the Ramón y Cajal Health Research Institute in Madrid, Spain, writes that the findings “are important because there is an urgent need for sensitive biomarkers to complement the currently used imaging-based screening methods in high-risk individuals, especially to exclude a diagnosis of pancreatic ductal adenocarcinoma when suspicious lesions are detected.”
Earl concludes: “The study also provides an excellent example for the biomarker field in general, as a robust, well organised, and well executed study that complies with the current guidelines for biomarker validation and reporting.”
medwireNews is an independent medical news service provided by Springer Healthcare Ltd. © 2025 Springer Healthcare Ltd, part of Springer Nature
Lancet Gastroenterol Hepatol 2025; doi:10.1016/S2468-1253(25)00056-1
Lancet Gastroenterol Hepatol 2025; doi:10.1016/S2468-1253(25)00089-5