Interobserver variability in intraductal papillary mucinous neoplasm subtypes and application of their mucin immunoprofiles

Abstract

Intraductal papillary mucinous neoplasm is considered a precursor lesion to pancreatic adenocarcinoma. These are further classified into four histologic subtypes: gastric, intestinal, pancreatobiliary, and oncocytic. The first aim of this study was to assess the interobserver variability among five gastrointestinal pathologists in diagnosing intraductal papillary mucinous neoplasm subtypes by morphology alone. The second aim of the study was to compare intraductal papillary mucinous neoplasm subtypes, which received consensus diagnoses (≥80% agreement) with their respective mucin immunoprofiles (MUC1, MUC2, MUC5AC, MUC6, and CDX2). A consensus histologic subtype was reached in 58% of cases (29/50) among the five gastrointestinal pathologists. Overall there was moderate agreement (κ = 0.41, Po 0.01) in subtyping intraductal papillary mucinous neoplasms without the use of immunohistochemistry. The histologic subtype with the best interobserver agreement was intestinal type (κ = 0.56, Po 0.01) followed by pancreatobiliary, gastric, mixed, and oncocytic types (κ = 0.43, Po 0.01; κ = 0.38, Po 0.01; κ = 0.17, Po 0.01; κ = 0.08, Po 0.04, respectively). Both kappa values for mixed and oncocytic subtypes were likely artificially low due to the underrepresentation of these subtypes in this study and not a true indication of poor interobserver agreement. Following an intradepartmental consensus meeting between two gastrointestinal pathologists, 68% of cases (34/50) received a consensus intraductal papillary mucinous neoplasm subtype. Sixty-nine percent of cases (11/16) that did not receive a consensus intraductal papillary mucinous neoplasm subtype could be classified based on their respective immunoprofiles. Standardizing the use of immunohistochemistry with a mucin immunopanel (MUC1, MUC2, MUC5AC, and MUC6) may improve the agreement of diagnosing intraductal papillary mucinous neoplasm histologic subtypes.

Endoscopic Ultrasound Through-The-Needle Biopsy of Pancreatic Cysts: Toward Procedure Standardization

“…TTNB specimens allow for the determination of the grade of dysplasia of mucinous lesions that can impact decision-making. However, our results (3 nonconcordant cases) add discussion to the known problem of the interobserver agreement of grade of dysplasia and invasion in IPMN already evaluated in previous studies [30]. We must also consider that the degree of epithelial atypia may vary within the cyst [31].…”

Diagnosis of Pancreatic Cystic Lesions by Virtual Slicing: Comparison of Diagnostic Potential of Needle-Based Confocal Laser Endomicroscopy versus Endoscopic Ultrasound-Guided Fine-Needle Aspiration

“…The diagnosis and grading of PCN can be challenging. [17181920] The nCLE provides images similar to a low-magnification tissue section with intact architectural pattern. This information could be applied in conjunction with morphologic features to improve the diagnostic accuracy of cytology samples.…”

Response

“…9 Therefore, the risk of missing more "aggressive" subtypes on TTNB specimens must be considered. Second, the interobserver agreement among pathologists for IPMN subtyping on surgical specimens was found to be moderate, 10 and it would likely be even worse when assessed on small TTNB specimens. Consequently, an agreement study on TTNB samples for IPMN subtyping would be strongly warranted.…”