Impact on clinical follow-up of the Milan System for salivary gland cytology: A comparison with a traditional diagnostic classification.

INTRODUCTION: No universally accepted classification exists for salivary gland FNA. The proposed Milan System for Reporting Salivary Gland Cytopathology (MSRSGC) offers a uniform classification with management recommendations. We compared FNA diagnoses from a prior study with specific diagnoses with corresponding MSRSGC diagnoses.

METHODS: One-hundred and sixty-four cases from a prior cytological study with histological follow-up were re-reviewed by one of the authors and assigned to one of the MSRSGC categories. The original and MSRSCG diagnoses were compared, as were follow-up recommendations.

RESULTS: The MSRSGC system classified 29 specimens as non-diagnostic (seven histologically shown to be benign salivary gland, two non-mucinous cysts, 14 sialadenitis, one pleomorphic adenoma, one haemangioma, one lymphoma, one adenoid cystic carcinoma, one squamous carcinoma and one benign lymphoid proliferation). The original study diagnosed these lesions as: seven benign cysts, 15 benign salivary gland tissue, one benign neoplasm and two insufficient for diagnosis. In seven cases, MSRSGC disagreed with original diagnoses and surgical resection showed lesions where optimal follow-up was more consistent with original cytological diagnosis. In 10 cases with disagreement, the MSRSGC was associated with a more appropriate follow-up based on the surgical diagnosis. Malignancy risks for the Milan categories were: non-diagnostic (12%), non-neoplastic (5%), atypia of undetermined significance (19%), neoplasm, benign (5%), neoplasm (40%), suspicious for malignancy (60%) and malignant (93%).

CONCLUSION: MSRSGC was comparable with the original reported diagnoses in the majority of cases. Both systems had high accuracy for distinguishing benign from malignant lesions and both were associated with appropriate follow-up in most cases.