Pancreatic cystic lesions: the challenge of risk stratification for malignancy

We read with interest the study by Al-Haddad et al. [1] on the accuracy of integrated molecular pathology (IMP) testing in assessing the malignant potential of pancreatic cysts, and the comparison with the 2012 Sendai guideline management criteria model. The authors found that IMP testing can differentiate between the need for surveillance and the need for surgery more accurately than the Sendai criteria. We were surprised at the positive predictive value (PPV) calculated for the 2012 Sendai criteria for malignancy in this article. The authors found a PPV of 20.8 % for the Sendai criteria versus 57.9 % for IMP testing. This value is significantly lower than the PPV (88 %) for the 2012 Sendai criteria reported by Goh et al. [2].

This remarkable difference is likely due to the surgical criteria chosen by Al-Haddad et al. In their study, the surgical category required the presence of at least one of the following features: obstructive jaundice in a patient with a cystic lesion of the pancreatic head, cyst size larger than 3 cm, enhancing solid component or definite mural nodule, severe cytological findings, main duct involvement, main duct dilatation (≥ 1 cm), abrupt changes in duct caliber, or a presumptive diagnosis of mucinous cystic neoplasm based on cytologic indications of mucin. In the 2012 Sendai guidelines [3], a size larger than 3 cm is not considered to be a definitive surgical criterion, but rather an indication for endoscopic ultrasonography, as are non-enhancing mural nodule and main duct involvement. The identification of mucin on cytologic examination or carcinoembryonic antigen (CEA) in cystic fluid (≥ 192 ng/mL) is not a surgical indication, but only indirect evidence of the mucinous nature of a lesion.

In our opinion, the interpretation of the 2012 Sendai guidelines on surgical indications reported by Al-Haddad et al. may explain why their PPV was less statistically significant than that reported by Goh et al. The authors repeated the calculation, completely excluding cyst size and omitting all presumptive cytologic and CEA criteria for mucinous cysts, with minimal impact on the performance of the 2012 Sendai model (only [sic] 8 times and 11.6 times higher risks for malignant outcome, respectively, in patients meeting surgical criteria). However, there is no discussion of other surgical criteria, such as non-enhancing mural nodule, main duct involvement, and cytologic indications of mucin, which at best are considered “worrisome features” by the 2012 Sendai guidelines, not definitive surgical indications [3]. Furthermore it is not clear why the authors recalculated the relative risk, omitting the doubtful surgical criteria (cyst size and the presence of mucus or cystic CEA > 192 ng/mL), but not the PPV, which can offer a more useful comparative value.

The authors chose four molecular criteria for malignancy (listed in Table 1 of their article), referring to a multicenter study by Khalid et al. [4], but the chosen criteria correspond only partially to the criteria cited in the study of Khalid et al. Furthermore, one of the molecular criteria for malignancy chosen by Al-Haddad et al. (a single low clonality oncogene mutation) seems to correlate more with the mucinous nature of a cyst than with malignancy [4]. We believe that further prospective studies are needed before conclusions can be drawn about the clinical application of IMP testing, and molecular analysis in general, to risk stratification for malignancy in pancreatic cystic lesions.

• References

• 1 Al-Haddad MA, Kowalski T. Siddiqui A et al. Integrated molecular pathology accurately determines the malignant potential of pancreatic cysts. Endoscopy 14.10.2014; DOI: DOI: 10.1055/s-0034-1390742. [Epub ahead of print].
• 2 Goh BK, Tan DM, Thng CH et al. Are the Sendai and Fukuoka consensus guidelines for cystic mucinous neoplasm of the pancreas useful in the initial triage of all suspected pancreatic cystic neoplasms? A single-institution experience with 317 surgically-treated patients. Ann Surg Oncol 2014; 21: 1919-1926
• 3 Tanaka M, Fernandez-del Castillo CC, Adsay V et al. International consensus guidelines 2012 for the management of IPMN and MCN of the pancreas. Pancreatology 2012; 12: 183-197
• 4 Khalid A, Zahid M, Finkelstein SD et al. Pancreatic cyst fluid DNA analysis in evaluating pancreatic cysts: a report of the PANDA study. Gastrointest Endosc 2009; 69: 1095-1102