Pathologic Discordance In Sarcomas: Prospective Comparison Of External And Sarcoma Center Pathologic Diagnosis.

11020 Background: With more than 80 different histologic subtypes, sarcomas are a unique pathologic challenge. As therapeutic decisions have become histology-specific, obtaining an accurate pathologic diagnosis is critical in guiding treatment decisions. The aim of this study is to determine the discordance between the diagnosis rendered by an external non-specialized pathologist and pathologic re-review by a specialized sarcoma pathologist at a high-volume sarcoma center. Methods: Patients who presented at the UCLA Multidisciplinary Sarcoma Conference (MSC) in 2017 that had a pathologic diagnosis from an outside facility were included in this study. All specimens underwent pathologic re-review at UCLA by an experienced sarcoma pathologist. The pathology was classified as concordant (identical diagnoses), minor discordance (difference with minor impact on prognosis/therapy) and major discordance (difference with significant impact on prognosis/therapy). Results: 1350 patients were presented at the UCLA MSC in 2017. Of the 635 new patients, 196 presented with an outside pathologic diagnosis and underwent pathologic re-review at UCLA. 44% (n = 87) were concordant, 22% (n = 43) had minor discordance, and 34% (n = 66) had major discordance. Major discordance included substantial discrepancies in histologic subtype (n = 24, 36%), benign/malignant mismatch (n = 23, 35%), diagnostic from non-diagnostic (n = 12, 18%) and major grading discrepancy (n = 7, 11%). Major discordance was most often seen in biopsies [needle (n = 27, 32%), incisional (n = 30, 44%)] as compared to resection (n = 9, 21%). Conclusions: 56% of external non-specialized sarcoma pathologic diagnoses were discordant from specialized sarcoma pathologist review, 34% of which were major discordances. Pathologic re-review of a presumed sarcoma by a specialized sarcoma pathologist is critical for both patient care and investigational studies. [Table: see text]