Intraoperative Evaluation of Pediatric Bone and Soft Tissue Lesions: Retrospective Analysis of 595 Frozen Sections With Emphasis on Discrepancy and Diagnostic Pitfalls.

Abstract

Context.—: Frozen section (FS) evaluation of pediatric bone and soft tissue (BST) lesions is infrequently encountered and may pose considerable diagnostic challenges. Limited data exist about the accuracy and related diagnostic difficulties.

Objective.—: To identify and analyze discrepancy between the FS diagnosis and final diagnosis in order to increase the awareness of common diagnostic pitfalls in FS evaluation of pediatric BST lesions.

Design.—: We retrospectively reviewed 595 consecutive FSs of pediatric BST lesions from 373 patients and analyzed the accuracy and causes for interpretation errors.

Results.—: Discrepant diagnoses were found in 23 of 595 FSs (3.9%). Discrepancy rates were slightly higher in benign, soft tissue lesions and FSs requested for diagnosis/adequacy, although no statistically significant difference was observed. Pathologist misinterpretation contributed to discrepancy in 17 of 23 FSs (73.9%), which were classified into 6 patterns of error. For margin, 3 patterns were found: normal hematopoietic elements versus malignant cells in Ewing sarcoma bone marrow margin (n = 3), prominent sinonasal vasculature and stroma versus sinonasal tract angiofibroma (n = 3), and atrophic skeletal muscles versus malignant cells in rhabdomyosarcoma and Ewing sarcoma (n = 2). For diagnosis, another 3 patterns were identified: misclassification of benign bone tumors (n = 5), misclassification of benign spindle neoplasms (n = 2), and vascular malformation versus normal tissue (n = 2).

Conclusions.—: FS is a valuable tool for guiding surgical management of pediatric BST lesions, which can be challenging entities and represent significant diagnostic pitfalls. Awareness of these FS pitfalls may help in further increasing diagnostic accuracy.