Diagnostic role of DOG-1, GFAP and B-catenin in Basal cell Adenoma and Cellular Pleomorphic Adenoma of the Salivary Gland

Abstract

Background: Pleomorphic Adenoma (PA) and Basal cell adenoma (BCA) are benign salivary gland tumors that may pose a diagnostic challenge if typical features are not present. Due to the increased relapse and malignant transformation rate of the former, a correct diagnosis carries relevant prognostic information. Even though immunohistochemistry (IHC) plays a limited role in the diagnosis of these tumors, the use of IHC panels could increase diagnostic accuracy. In the present work, we aimed to demonstrate that the use of an IHC panel consisting of Glial Fibrillary Acid Protein (GFAP), B-Catenin and Discovered On GIST 1 (DOG-1) can aid in the differential diagnosis between PA and BCA.

Methods: We analyzed 18 cases of benign salivary gland tumors (Pleomorphic adenomas and Basal cell adenomas) with overlapping histologic features. First, a head and neck pathologist diagnosed the cases relying on morphology alone. Afterwards, cases were re-evaluated considering the IHC panel results. Inter-observer IHC scoring concordance was evaluated with pre-defined marker cut-off points using Cohen's Kappa scores.

Results: Based on morphology alone, 9 cases were classified as PA while the remaining tumors were considered to be BCA. Five out of nine BCA cases showed GFAP staining and absent nuclear B-catenin and DOG-1 positivity. Conversely, 2 PA cases showed absent GFAP and positive nuclear B-catenin with concurrent DOG-1 expression. Therefore, after IHC evaluation, up to 40% of morphologic diagnoses were reconsidered. Overall, the inter-observer concordance for IHC evaluation was good (resulting Kappa Scores between 0.78 and 1).

Conclusion: Our work supports the use of a concise IHC panel to improve the diagnostic accuracy of benign salivary gland tumors with overlapping histologic features.

Keywords: Basal-cell adenoma; Immunohistochemistry; Pleomorphic Adenoma.

Fig. 1

Bar plot showing concordance Kappa scores and 95% CI for each immunohistochemical marker. Since DOG1 showed perfect concordance confidence intervals could not be calculated

Fig. 2

Representative images from cases 18 and 9. Case 18 was initially diagnosed as Pleomorphic Adenoma and reclassified as Basal cell Adenoma. Case 9 was originally diagnosed as basal cell adenoma and reclassified as pleomorphic adenoma

Fig. 3

Parallel set plot denoting diagnostic flux through immunohistochemical evaluation. Blue connector lines represent cases initially diagnosed as basal cell adenomas and red connector lines represent cases initially diagnosed as pleomorphic adenomas