The Milan system for reporting salivary gland cytopathology - Assessment of utility and the risk of malignancy.

Abstract

Cytological diagnosis of salivary gland lesions furnishes several challenges to pathologists and clinicians owing to their diverse morphology and varying reporting terminology. The Milan system for reporting salivary gland cytopathology was proposed to overcome these defects. The aim of this study was to grade histopathologically diagnosed cases of salivary gland pathology using Milan system and to correlate Milan system with histopathological diagnosis. The study aims to assess the utility of the Milan system and the risk of malignancy in each category. This is a retrospective study of 4 years duration. All the cases of salivary gland pathology with fine-needle aspiration cytology findings from January 2015 to December 2018 were retrieved and classified according to the Milan system. Histopathology follow-up of these cases were reviewed from the archives, and risk of malignancy was calculated. Sensitivity, specificity, positive predictive value, and negative predictive value were calculated using histopathological diagnosis as the gold standard. Cytology slides of 151 cases were reviewed, and the Milan system was applied. Histopathology follow-up was obtained in 94 cases (attrition rate = 37.74%). The percentage of cases in each category were as follows: nondiagnostic = 9.27%, nonneoplastic = 4.64%, atypia of undetermined significance = 3.97%, benign neoplasm = 23.84%, neoplasm of uncertain malignant potential = 8.61%, suspicious for malignancy = 11.92%, and malignant = 37.75%. Sensitivity, specificity, positive predictive value, and negative predictive value of the Milan system were 94.20%, 96.00%, 98.48%, and 85.71%, respectively. The Milan system has high sensitivity and specificity in comparison with histopathology. This system proves helpful to the pathologist as well as the clinician in achieving uniformity of reporting terminology by classifying salivary gland lesions into subcategories with prognostic relevance.