Objectives Multi-institutional studies are required for the validation of the Milan System for Reporting Salivary Gland Cytopathology (MSRSGC). Methods A total of 1,560 fine-needle aspirations of the salivary glands were retrieved from two institutions for a 12-year period. The diagnoses were reclassified based on the MSRSGC. Risk of malignancy (ROM) for each category was calculated based on 694 histologic follow-up cases. Results The ROM for each category was: 18.3% for nondiagnostic, 8.9% for nonneoplastic, 37.5% for atypia of undetermined significance (AUS), 2.9% for benign neoplasm, 40.7% for salivary gland neoplasm of uncertain malignant potential (SUMP), 100% for suspicious for malignancy, and 98.3% for malignant. The sensitivity, specificity, positive predictive rate, and negative predictive rates were 89%, 99%, 98%, and 96%, respectively. Conclusions The results of the current study are in keeping with the MSRSGC. The indeterminate categories of AUS and SUMP showed intermediate ROMs at 37.5% and 40.7%, respectively.

BACKGROUND The newly unveiled Milan System for Reporting Salivary Gland Cytopathology (MSRSGC) has proposed salivary gland neoplasm of uncertain malignant potential (SUMP) as an indeterminate category. The category is reserved for fine‐needle aspiration (FNA) cases that are diagnostic of a salivary gland neoplasm but cannot be further designated as a specific tumor type. The objective of the current study was to evaluate the clinical utility of subtyping SUMP cases based on different cell types. METHODS A retrospective search of cytology databases at 2 institutions for salivary gland FNAs from 2006 through 2017 was conducted. The cytologic diagnosis of each case was reclassified according to the MSRSGC. Histologic follow‐up was retrieved for correlation. Cases reclassified as SUMP that had a follow‐up pathologic diagnosis were subject to cytology review and subtyping into oncocytic/squamoid, basaloid, or myoepithelial subtypes based on cytomorphology. The risk of malignancy (ROM) for each subtype was analyzed. RESULTS There were 92 SUMP cases, which comprised 5.9% of 1560 consecutive salivary gland FNAs within the 12‐year study period. Histologic follow‐up was available for 59 patients. After cytology review, there were 18 cases (30.5%) of oncocytic/squamoid subtype, 25 (42.4%) of basaloid subtype, and 16 (27.1%) of myoepithelial subtype. Pathologic correlation revealed an ROM of 61.1% (11 of 18 cases) for the oncocytic/squamoid subtype, 40.0% (10 of 25 cases) for the basaloid subtype, and 18.8% (3 of 16 cases) for the myoepithelial subtype. The differences in ROM among the 3 subtypes were statistically significant (P = .0476). CONCLUSIONS Subtyping SUMP cases into categories based on cell type demonstrated differential ROMs for better clinical stratification. Future prospective studies are mandatory to confirm this finding.

No abstract