Core Needle and Incisional Biopsy of Pleomorphic Adenoma Family of Tumors: Assessment Terminology Employed, Risk Stratification and Performance Characteristics

Summary: Background: Pleomorphic adenoma (PA) remains the most common specific diagnosis rendered, either directly or with various other related terms. We evaluate the distribution of the ‘PA-centric’ terminology in core needle (CNB), punch (PBx), and incisional bx (IB) at major salivary gland (MSG) and palatopharyngeal (P/P) sites and its performance on resection correlation. Design: CNB, PBx and IB from MSG and P/P sites were queried (1994-2024) using ‘pleomorphic adenoma,’ or ‘carcinoma ex pleomorphic adenoma’ in the bx and/or resection diagnosis and comment. Of 312 records, 209 were retained. Diagnoses were stratified using the UPMC risk classification system (PMID:38952287) and grouped by terminology. Risk of malignancy (ROM) and probability of a ‘PA family’ final diagnosis (including atypical PA and carcinoma ex PA) were evaluated. Results: Biopsy sites were as follows: major salivary (parotid and submandibular) glands (n=135, 64.6%) and palate/pharynx (n=74, 35.4%). CNB were essentially exclusive to MSG, comprising the main bx type at this site (n=99, 73.3%). Of the P/P cases, 73 (98.6%) underwent IB or PBx. Risk stratification was similar across sites. Biopsy of ‘PA family’ tumors were rarely inadequate or nonneoplastic (n=5, 2.5%). Most were in the benign neoplasm (R1) category (n=125, 59.8%). Suspicious/malignant (R4a&b) prevalence was 12.4% (26/209). Specific term distribution was even between sites except for ‘myoepithelial rich neoplasm/myoepithelioma’ which was more frequent in P/P sites (n=15, 20.3%). 96 biopsies (no frozen section) had subsequent resections. The overall prevalence of malignancy was similar across site groups (MSG:16/57, 28.1%; P/P:13/39, 33.3%). Risk classifier showed an expected ascending ROM. Using a PA-centric diagnostic grouping, ROM when PA was in the line diagnosis or comment was 8.9%. Atypical PA was not frequent but had a high ROM (4/6, 66.7%). PA family diagnosis probability was high throughout these groupings (overall: 87.8%, range: 75-100%). Subset assessment of nonmalignant biopsies with malignant resections shows that while many nonclassical features had a higher ROM, only cellular, myoepithelial-rich, and mitotically active lesions were significantly different (Bonferroni adjusted p<0.05) Conclusion: PA family of tumors can be readily recognized on tissue biopsy. The ROM for a PA diagnosis is not inconsequential; myoepithelial-rich, cellular and mitotically active features should up-classify PA to at least uncertain malignant potential (R2). Atypical PA, though rare, may qualify as suspicious for malignancy (R4a). Rayan Rammal, N. Paul Ohori, and Raja Seethala