Systematic review and meta-analysis:Diagnostic performance of DNA alterations in pancreatic juice for the detection of pancreatic cancer

Background and aims: Pancreatic cancer has a dismal prognosis. So far, imaging has been proven incapable of establishing an early enough diagnosis. Thus, biomarkers are urgently needed for early detection and improved survival. Our aim was to evaluate the pooled diagnostic performance of DNA alterations in pancreatic juice. Methods: A systematic literature search was performed in EMBASE, MEDLINE Ovid, Cochrane CENTRAL and Web of Science for studies concerning the diagnostic performance of DNA alterations in pancreatic juice to differentiate patients with high-grade dysplasia or pancreatic cancer from controls. Study quality was assessed using QUADAS-2. The pooled prevalence, sensitivity, specificity and diagnostic odds ratio were calculated. Results: Studies mostly concerned cell-free DNA mutations (32 studies: 939 cases, 1678 controls) and methylation patterns (14 studies: 579 cases, 467 controls). KRAS, TP53, CDKN2A, GNAS and SMAD4 mutations were evaluated most. Of these, TP53 had the highest diagnostic performance with a pooled sensitivity of 42% (95% CI: 31–54%), specificity of 98% (95%-CI: 92%–100%) and diagnostic odds ratio of 36 (95% CI: 9–133). Of DNA methylation patterns, hypermethylation of CDKN2A, NPTX2 and ppENK were studied most. Hypermethylation of NPTX2 performed best with a sensitivity of 39–70% and specificity of 94–100% for distinguishing pancreatic cancer from controls. Conclusions: This meta-analysis shows that, in pancreatic juice, the presence of distinct DNA mutations (TP53, SMAD4 or CDKN2A) and NPTX2 hypermethylation have a high specificity (close to 100%) for the presence of high-grade dysplasia or pancreatic cancer. However, the sensitivity of these DNA alterations is poor to moderate, yet may increase if they are combined in a panel.