Serum CA19.9 for Detecting High-Grade Dysplasia and Invasive Carcinoma in Patients With Intraductal Papillary Mucinous Neoplasm: A Systematic Review and Meta-Analysis
A partir d'une revue systématique de la littérature publiée jusqu'en mars 2025 (24 études incluant au total 5 281 patients), cette méta-analyse évalue la performance, du point de vue notamment de la sensibilité, de la spécificité et des rapports de vraisemblance, du niveau sérique de l'antigène CA19.9 au diagnostic pour détecter une dysplasie de haut grade ou un carcinome invasif chez les patients présentant une tumeur mucineuse papillaire intracanalaire du pancréas
Importance : Current guidelines recommend routine assessment of serum carbohydrate antigen 19.9 (CA19.9) at diagnosis and during follow-up of patients with intraductal papillary mucinous neoplasm (IPMN) to detect high-grade dysplasia (HGD) and invasive carcinoma (IC). Guidelines consider an elevated serum CA19.9 level (≥37 U/mL) a worrisome feature and a relative indication for surgery.
Objective : To evaluate the diagnostic accuracy of CA19.9 for identifying patients with IPMN at high risk of HGD and IC.
Data Sources : PubMed, Embase (Ovid), and Web of Science databases (inception to March 1, 2025).
Study Selection : Studies reporting on diagnostic accuracy of CA19.9 (cut-off 37 U/mL) in patients with resected IPMN.
Data Extraction and Synthesis : Data extraction was completed by 3 reviewers (March to December 2025). Risk of bias was assessed using the QUADAS-2 tool and evidence certainty with Grading of Recommendations Assessment, Development and Evaluation. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses of Diagnostic Test Accuracy Studies reporting guidelines and the Cochrane Diagnostic Test Accuracy protocol were applied for overall conduct. A random-effects meta-analysis was conducted in December 2025 to obtain a pooled estimate of diagnostic accuracy. Main pancreatic duct involvement and rate of HGD or IC were assessed using meta-regression. Summary receiver operating characteristics curves were generated. The area under the curve (AUC) was calculated as overall test performance.
Main Outcomes and Measures : Diagnostic accuracy of serum CA19.9 (≥37 U/mL), reported as pooled sensitivity, specificity, positive likelihood ratio (PLR) and negative likelihood ratios (NLR), diagnostic odds ratio (DOR), and AUC and their 95% CIs. Pathology was considered as reference standard.
Results : Overall, 24 studies assessing 5281 patients with IPMN were included. Mean age was 64.7 (95% CI, 62.24-67.10) years, and 2919 patients (55%) were male. Serum CA19.9 data were available for 4653 patients (88.1%), with 966 patients (20.8%) showing elevated levels. Pooled estimates of diagnostic accuracy for HGD or IC (present in 1939 patients [41.7%]) demonstrated sensitivity of 0.35 (95% CI, 0.29-0.43), specificity of 0.90 (95% CI, 0.87-0.92), PLR of 3.37 (95% CI, 2.56-4.44), NLR of 0.72 (95% CI, 0.65-0.80), DOR of 4.67 (95% CI, 3.26-6.70), and AUC of 0.78 (95% CI, 0.74-0.82).
Conclusions And Relevance : In this systematic review and meta-analysis, CA19.9 with the current cutoff (37 U/mL) demonstrated poor performance in excluding HGD or IC and only modest value for ruling in IC. These findings suggest CA19.9 should not be used as a standalone or screening test, while its limited rule-in value may aid decision-making in patients with moderate pretest probability. Future studies should investigate whether different cutoffs and dynamic trends could improve diagnostic accuracy.
JAMA Network Open , article en libre accès, 2026