From detection to ruling out in pancreatic cancer: a new perspective

Pancreatic ductal adenocarcinoma (PDAC) remains one of the most lethal malignancies and is now the third leading cause of cancer-related deaths (1). Its prognosis, however, would markedly improve if the diagnosis could be made at an earlier stage. Achieving this requires the establishment of an effective surveillance strategy for asymptomatic individuals at an elevated risk. Although carbohydrate antigen 19-9 (CA19-9) remains the only blood-based biomarker currently used in clinical practice, its sensitivity and specificity both below 80% limit its utility in early detection of PDAC. CA19-9 can be elevated by inflammation, biliary obstruction, or pancreatic cystic lesions (2). Compounding this challenge is the low prevalence of PDAC in the general population, which undermines the cost-effectiveness of surveillance programs. To address this barrier, the Define-Enrich-Find (DEF) strategy has recently been proposed (3). This framework identifies individuals with an estimated cancer risk of approximately 1%, enriches this group using clinical risk factors or biomarkers, and thereby maximizes the likelihood of detecting PDAC efficiently.

In this context, the METAPAC trial (multimetabolite signatures for patients at risk of or with suspected PDAC) provides pivotal evidence advancing the DEF strategy toward clinical feasibility (4). This prospective, multicenter, masked phase IV diagnostic study targeted patients with pancreatic lesions detected by computed tomography (CT) who required further diagnostic evaluation. Between September 2016 and April 2022, 1,370 patients across 23 German centers were screened, with 1,129 ultimately included in the analysis. Among these, 489 were diagnosed with PDAC and 640 served as controls representing high-risk conditions such as chronic pancreatitis, acute pancreatitis, and pancreatic cystic diseases. Two plasma metabolic signatures were validated: the improved i-Metabolic signature consisting of 12 metabolites plus CA19-9, and the minimalistic m-Metabolic signature consisting of four metabolites plus CA19-9.

While the sensitivity of the i-Metabolic (67.5%) and m-Metabolic (59.9%) signatures was lower than that of CA19-9 alone (81.8%), both signatures achieved notably higher specificity: 90.4% and 93.6%, compared with 79.1% for CA19-9. High specificity is particularly important in PDAC surveillance to minimize false positives and avoid unnecessary invasive procedures such as endoscopic ultrasonography (EUS) and endoscopic retrograde cholangiopancreatography (ERCP). Notably, this high specificity was also preserved among patients with resectable PDAC (stage I–IIB), with specificities of 91.7% and 93.6% for the i-Metabolic and m-Metabolic signatures, respectively, compared with only 45.6% for CA19-9 alone. These results highlight that metabolic signatures may serve as effective exclusion tools in the diagnostic evaluation of PDAC, even at an early stage. For the early detection of pancreatic cancer, it is important to conduct screening using highly sensitive tests. However, a high false-positive rate can lead to the performance of unnecessary invasive procedures. When aiming for “zero missed cases” in the early stages, it is necessary to bear in mind the impact on healthcare economics, cost-effectiveness, and the risk of complications. Therefore, establishing a clinical workflow that involves screening with highly sensitive biomarkers, followed by triaging patients with positive results using non-invasive imaging and a highly specific metabolic signature, should be considered.

Recent years have witnessed an expansion of PDAC biomarker research. Investigations have explored microRNAs, satellite RNAs, and extracellular vesicles (EVs) (5,6). EVs, in particular, are membrane-bound particles carrying proteins and lipids that reflect tumor–microenvironment interactions. Studies have demonstrated that sampling portal blood, pancreatic juice, bile, or duodenal fluid obtained via EUS or ERCP can yield spatially concentrated molecular information not captured in the peripheral blood (7). Such approaches may enhance early cancer detection, real-time monitoring of tumor biology, and risk stratification. Additional advances include novel biosensors capable of highly sensitive CA19-9 detection (8). Unlike these biomarkers, metabolic signatures are systemic biological signals. It is significant, however, that they have been shown to possess a high degree of specificity that is comparable to that of tumor-derived biological signals. Given their potential to enable earlier detection, they are expected to be of benefit in the future management of pancreatic cancer. In addition, while endoscopy is necessary to achieve high accuracy, its invasive nature is a drawback. For this reason, it is likely to be adopted later in the clinical workflow than metabolic signatures. These developments show that PDAC biomarker research has moved beyond simple discovery toward innovations in specimen acquisition and measurement technologies. In addition, DNA, RNA, proteins, and lipids found in blood, urine, and saliva are also emerging as potential new biomarkers (9). When combined with existing markers, these could lead to the early detection of pancreatic cancer with minimal invasiveness. Tajbakhsh et al. are evaluating pancreatic cancer risk using a proprietary algorithm that combines the urinary proteins lymphatic vessel endothelial hyaluronan receptor 1 (LYVE1) and regenerating family member 1β (REG1B) with CA19-9. This algorithm can distinguish between patients with early-stage pancreatic cancer and asymptomatic subjects with an area under the curve (AUC) of 0.93, a sensitivity of 85–90%, and a specificity of 78–87% (10). It is considered that using metabolic signatures in cases that test positive for these highly sensitive biomarkers will help identify pancreatic cancer patients.

However, most biomarker research to date has focused on enhancing sensitivity to “detect” PDAC. High sensitivity often comes at the cost of specificity, leading to high false-positive rates, which can be particularly problematic in a low-prevalence disease such as PDAC. In this context, the METAPAC study’s emphasis on high specificity, enabling “exclusion” of PDAC, represents a strategically distinct and clinically valuable direction.

The methodological rigor of the METAPAC study warrants particular attention. The trial adhered to EDRN (Early Detection Research Network) and STARD/TRIPOD (Standards for Reporting of Diagnostic accuracy/Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis) guidelines and employed an enrichment design combined with inverse probability weighting. Prospective validation of the metabolic signatures with ≥85% specificity and 80% power would ordinarily require enrollment of more than 29,000 chronic pancreatitis cases to obtain 250 PDAC cases—an impractical undertaking. By instead enrolling patients with pancreatic lesions requiring further diagnostic evaluation, the investigators substantially increased PDAC prevalence in the study population while maintaining clinical representativeness. This reflects a realistic clinical scenario in which many patients with pancreatic abnormalities require invasive diagnostic procedures. In addition, the use of quantitative metabolomics on a single liquid chromatograph-mass spectrometry (LC-MS/MS) platform, with clear attention to cost, reproducibility, and standardization, underscores the investigators’ commitment to clinical translation. For a surveillance tool to be cost-effective in a population with 1% PDAC incidence, specificity must exceed 88% and costs must remain below US$400—criteria that these metabolic signatures can potentially meet.

The DEF strategy will be essential for successful surveillance of low-prevalence cancers such as PDAC. Known risk factors, such as family history of PDAC, pancreatic cystic lesions including intraductal papillary mucinous neoplasm (IPMN), chronic pancreatitis, and new-onset diabetes, form the foundation of risk stratification. Surveillance of familial PDAC through the Cancer of the Pancreas Screening (CAPS) program has demonstrated improved survival by detecting PDAC at early stage, yet the prevalence of such high-risk individuals is low, limiting population-level impact (11). New-onset diabetes has emerged as another key to early diagnosis of PDAC. The Enriching New-onset Diabetes for Pancreatic Cancer (ENDPAC) score enriches the prevalence of PDAC among new-onset diabetes cases by incorporating changes in weight, blood glucose, and age at diabetes onset. High-risk individuals identified by this score show a 3.6% prevalence of PDAC, warranting invasive investigation. In a sub-analysis of the METAPAC study, the m-Metabolic signature alone achieved an AUC of 0.84 in newly diagnosed diabetes, suggesting that early metabolic perturbations may distinguish PDAC even at the onset of diabetes. Combining ENDPAC scoring with metabolic signatures is a promising area for further research.

IPMN surveillance represents another major area of interest. Pancreatic cystic lesions, including IPMNs, have a prevalence exceeding 10%, raising concerns about healthcare resource utilization (12,13). While some guidelines allow discontinuation of surveillance for stable, low-risk IPMNs, emerging data suggest a long-term risk of PDAC even after five years (14). CA19-9 offers limited utility for detecting concomitant PDAC in IPMN surveillance (15). It is therefore worth investigating whether metabolic signatures could improve early PDAC detection within this large at-risk population.

Despite its strengths, the METAPAC study has limitations. A registration imbalance resulted in incomplete staging information for resectable PDAC cases, restricting the analysis to a limited subset. Data auditing was also limited by coronavirus disease 2019 (COVID-19)-related constraints, leading to informational gaps. Additionally, given that the study was conducted exclusively in Germany, validation in populations with different racial backgrounds and healthcare systems is needed to establish external generalizability. The study was conducted in Germany and the results of those metabolic signatures should be validated outside Germany, too. The cost analysis should also be considered among populations with different prevalence of pancreatic cancer and various medical systems. While the metabolic signature has high specificity, its sensitivity is lower than CA19-9 and could lead to the missed diagnosis of early-stage cancer. Thus, the metabolic signature might be clinically useful when combined with biomarkers with high sensitivity, including CA19-9. Furthermore, selection of biomarkers needs adaptation, depending on the prevalence of pancreatic cancer (high risk individuals vs. the general population) as well as the aim of pancreatic cancer screening (missing no cancers vs. dismissing non-cancers). Given its high specificity, non-invasiveness with easy repeatability, the metabolic signatures can be the first sieve to select high risk populations who should proceed to invasive but precise procedures such as endoscopic ultrasound. However, a further cost-analysis is needed to incorporate this approach to clinical practice.

Nevertheless, the METAPAC trial stands as the first study to demonstrate a realistic and actionable solution for PDAC surveillance. If an inexpensive, reproducible blood test can accurately rule out PDAC in high-risk individuals, it may profoundly reshape clinical diagnostic workflows. The next steps include integrating metabolic signatures into proactive surveillance programs and evaluating their true impact on patient outcomes and healthcare economics. Furthermore, multiomic integration—including genomics, radiomics, and metabolomics—will likely be essential for advancing the early detection of PDAC. In a meta-analysis reported by Alidina et al., image diagnosis using artificial intelligence and machine learning demonstrated high accuracy, with a sensitivity of 0.88 and a specificity of 0.93. A multifaceted approach is likely to compensate for the low sensitivity that is a weakness of metabolic signatures (16).

In summary, metabolic signatures have opened the door to a paradigm shift—from attempting to “detect every tumor” to confidently “ruling out cancer” in large at-risk populations. This strategy may bring the long-awaited possibility of earlier, life-saving intervention for individuals at high risk of pancreatic cancer.

Acknowledgments

None.

Footnote

Provenance and Peer Review: This article was commissioned by the editorial office, Translational Gastroenterology and Hepatology. The article has undergone external peer review.

Peer Review File: Available at https://tgh.amegroups.com/article/view/10.21037/tgh-2026-0023/prf

Funding: None.

Conflicts of Interest: Both authors have completed the ICMJE uniform disclosure form (available at https://tgh.amegroups.com/article/view/10.21037/tgh-2026-0023/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Siegel RL, Giaquinto AN, Jemal A. Cancer statistics, 2024. CA Cancer J Clin 2024;74:12-49. [Crossref] [PubMed]
2. Haab B, Qian L, Staal B, et al. A rigorous multi-laboratory study of known PDAC biomarkers identifies increased sensitivity and specificity over CA19-9 alone. Cancer Lett 2024;604:217245. [Crossref] [PubMed]
3. Singhi AD, Koay EJ, Chari ST, et al. Early Detection of Pancreatic Cancer: Opportunities and Challenges. Gastroenterology 2019;156:2024-40. [Crossref] [PubMed]
4. Mahajan UM, Oehrle B, Goni E, et al. Validation of two plasma multimetabolite signatures for patients at risk of or with suspected pancreatic ductal adenocarcinoma (METAPAC): a prospective, multicentre, investigator-masked, enrichment design, phase 4 diagnostic study. Lancet Gastroenterol Hepatol 2025;10:634-47. [Crossref] [PubMed]
5. Panda A, Falasca M, Ragunath K. Extracellular vesicles in pancreatic cancer: a new era in precision medicine. Transl Gastroenterol Hepatol 2024;9:29. [Crossref] [PubMed]
6. Seimiya T, Suzuki T, Iwata T, et al. Combination of serum human satellite RNA and miR-21-5p levels as a biomarker for pancreatic cancer. iScience 2023;26:106021. [Crossref] [PubMed]
7. Shimamoto N, Sumiyama K, Fujita Y. Extracellular vesicle biomarkers in pancreatic ductal adenocarcinoma: from bulk detection to single-extracellular vesicle profiling and endoscopic liquid biopsy. J Gastroenterol 2026; Epub ahead of print. [Crossref]
8. Soares GON, Soares AC, Dias R, et al. Supramolecular PDDA/PEDOT:PSS Biosensor for Early Pancreatic Cancer Detection via CA19-9: Clinical Validation on Human Blood Samples. ACS Omega 2026;11:6452-63. [Crossref] [PubMed]
9. Brezgyte G, Shah V, Jach D, et al. Non-Invasive Biomarkers for Earlier Detection of Pancreatic Cancer-A Comprehensive Review. Cancers (Basel) 2021;13:2722. [Crossref] [PubMed]
10. Tajbakhsh J, Debernardi S, Blyuss O, et al. A Clinical Laboratory Improvement Amendments/College of American Pathologists-Compliant Noninvasive Laboratory-Developed Test for Early Detection of Pancreatic Ductal Adenocarcinoma. J Mol Diagn 2025;27:54-61. [Crossref] [PubMed]
11. Blackford AL, Canto MI, Dbouk M, et al. Pancreatic Cancer Surveillance and Survival of High-Risk Individuals. JAMA Oncol 2024;10:1087-96. [Crossref] [PubMed]
12. Mizuno S, Isayama H, Nakai Y, et al. Prevalence of Pancreatic Cystic Lesions Is Associated With Diabetes Mellitus and Obesity: An Analysis of 5296 Individuals Who Underwent a Preventive Medical Examination. Pancreas 2017;46:801-5. [Crossref] [PubMed]
13. Hamada T, Oyama H, Igarashi A, et al. Optimal age to discontinue long-term surveillance of intraductal papillary mucinous neoplasms: comparative cost-effectiveness of surveillance by age. Gut 2024;73:955-65. [Crossref] [PubMed]
14. Ohtsuka T, Fernandez-Del Castillo C, Furukawa T, et al. International evidence-based Kyoto guidelines for the management of intraductal papillary mucinous neoplasm of the pancreas. Pancreatology 2024;24:255-70. [Crossref] [PubMed]
15. Oyama H, Hamada T, Nakai Y, et al. Clinical trajectory of intraductal papillary mucinous neoplasms progressing to pancreatic carcinomas during long-term surveillance: a prospective series of 100 carcinoma cases. J Gastroenterol 2023;58:1068-80. [Crossref] [PubMed]
16. Alidina Z, Hussain AAM, Banani I, et al. Radiomics for early detection of pancreatic cancer: a systematic review and meta-analysis. J Gastrointest Surg 2026;30:102374. [Crossref] [PubMed]