Objectives: This study aimed to evaluate the accuracy of the risk factors proposed by Fukuoka guidelines in detecting malignancy of branch-duct intraductal papillary mucinous neoplasms.
Method: Diagnostic meta-analysis of cohort studies. A systematic literature search was conducted using MEDLINE, the Cochrane Library, Scopus, and the ISI-Web of Science databases to identify all studies published up to 2014.
Results: Twenty-five studies (2025 patients) were suitable for the meta-analysis. The "high risk stigmata" showed the highest pooled diagnostic odds ratio (jaundice, 6.3; positive citology, 5.5; mural nodules, 4.8) together with 2 "worrisome features" (thickened/enhancing walls, 4.2; duct dilatation, 4.0) and 1 "other parameters" (carbohydrate antigen 19-9 serum levels, 4.6).
Conclusions: An "ideal risk factor" capable of recognizing all malignant branch-duct intraductal papillary mucinous neoplasms was not identified and some "dismal areas" remain. However, "high risk stigmata" were strongly related to malignancy, mainly enhancing mural nodules. Among the "worrisome features," duct dilatation and thickened/enhancing walls were underestimated, and their diagnostic performance was similar to those of "high risk stigmata." The carbohydrate antigen 19-9 serum level should be added to the Fukuoka algorithm because this value could help in carrying out correct management.