Torakochirurgia
Evaluation of frozen section usefulness for the diagnosis of pulmonary lesions, pleural and mediastinal diseases, for the detection of metastasis of lung cancer in mediastinal lymph nodes, and for the assessment of surgical margins in lung cancer resection specimens

Background: The main reasons for performing intraoperative consultation (also called frozen section – FS) are: to establish the presence and nature of a lesion, to determine the adequacy of surgical resection margins, and to establish whether the tissue obtained contains enough diagnosable material for microscopic examination.
Aim: The aim of our investigation was to determine the accuracy of intraoperative consultation in the diagnosis of pulmonary, pleural and mediastinal lesions, for the detection of metastasis of lung cancer in mediastinal lymph nodes, and for the assessment of the adequacy of the resection margins in lung cancer specimens.
Material and methods: The retrospective study was based on a series of 539 cases, of which 905 frozen section slides were performed. Among them, 894 FS were connected with lung, pleural and mediastinal pathology, while 11 included different specimens. Discordant diagnoses between FS and the definitive diagnoses were identified and their consequences were analyzed.
Results: FS of lung lesions included 308 cases. The examination accuracy was determined at 96.8%, with 2.6% postponed for a permanent section, while those classified as false negative and false positive which had serious consequences for patients were 0.3% each. FS evaluation of surgical margins was performed in 59 cases, of which 98.3% were correct and 1.7% was put off until final paraffin section. There were no false negative or false positive results. FS of pleural samples was estimated in 15 cases. Correct diagnoses were established in 86% of FS, and false negatives without surgical consequence accounted for 13.3% of the sample. Evaluation of mediastinal lymph nodes was consistent with the final pathology diagnosis in 99.8% of cases, whereas 0.1% of cases were false negative and 0.1% false positive, without significant influence for surgical decision. FS examinations were performed in 19 cases of mediastinal disorders. The majority of estimated samples were shown to be useful for the final pathologic diagnosis (57.9%), while 0.5% of cases required a subsequent paraffin section. There were no false positive or negative results. The overall rate of FS accuracy was 98.2%, whereas of the remaining 1.8% of surgical pathology reports that needed revision, 1.2% required subsequent paraffin sections, 0.4% of cases were false negative, and 0.2% false positive. False negative results that differed from final pathological diagnoses were 0.2%, but only one (0.1%) represented a clinically significant error that may have affected patient management.
Conclusions: Intraoperative consultation is a crucial histopathological examination, above all in oncological proceedings, highly sensitive and specific especially for the diagnosis of malignancy.