Discrepancies in frozen section mucosal margin tissue in laryngeal squamous cell carcinoma

Background. Head and neck surgeons commonly request frozen sections. Practice patterns vary from laboratory to laboratory on how the tissue is used in performing the frozen section. Some pathologists wish to see all the material submitted by consuming it completely during frozen section, whereas others reserve some for permanent section. We wished to determine whether knowledge of margin status was initially inaccurate because of reserving tissue for permanent section. Methods. Sixty-five laryngectomies (total and partial) with margin assessment enhanced by frozen section evaluation were studied. Forty-five laryngectomy specimens, generating 249 frozen sections in which a permanent section was prepared from tissue remaining from frozen section examination, were studied. Results. Five of the 249 frozen sections contained a discrepancy between the frozen section and permanent section because of insufficient leveling of the frozen section block. These five discrepancies were called negative on frozen section, but permanent section revealed dysplasia (two cases of mild dysplasia, one case with moderate dysplasia, and one case with severe dysplasia) or carcinoma in situ (one case). Twenty laryngectomies in which the frozen section tissue was consumed at the time of frozen section generated 103 frozen sections. In eight of the frozen sections involving six cases, the diagnostic tissue was not present on one or two of the frozen section levels examined. Conclusions. We conclude that in examining margins for laryngeal squamous cell carcinoma the frozen section tissue should be completely sampled by examining several levels at the time of frozen section. This requires consuming or exhausting the frozen section tissue rather than reserving any remaining frozen tissue for a paraffin-embedded permanent section. (C) 2002 Wiley Periodicals, Inc.