Factors Predicting Inaccuracy Between Frozen Section Analysis and Postoperative Pathology Results: A Tertiary Center Experience

FACTORS PREDICTING INACCURACY BETWEEN FROZEN SECTION ANALYSIS AND POSTOPERATIVE PATHOLOGY RESULTS: A TERTIARY CENTER EXPERIENCE ABSTRACT Objective: To evaluate the diagnostic accuracy and to identify the factors determining the inaccuracy between FS analysis and postoperative pathology results in our hospital. Materials and Methods: This study included 1435 patients with pelvic masses who underwent surgical intervention and FS consultation. Results: On univariate analysis, menopausal state, abnormal uterine bleeding, preoperative serum Ca125 level, preoperative leukocyte value, preoperative neutrophil value, tumor size, presence of cystic component, pathological findings in Doppler USG, ascites and cell type were found to have statistical significance for benign / borderline / malign discrimination. Menopausal state (p lt;0.0001), preoperative serum Ca125 level (p lt;0.0001) and tumor size (p lt;0.0001) were identified as independent predictors for determining inaccuracy between intraoperative and postoperative pathological evaluation. Inconsistency increased 2.5 times with a serum Ca125 gt; 35 IU/ml, 3.8 times with tumor size gt; 79 mm and 5 times in postmenopausal patients. For the discrimination of benign / borderline / malign definitions, FS results and final pathology results were compatible in 1250 (87%) patients while it was not in 185 (13%) patients. Conclusions: Increased preoperative serum Ca125 level was a predictor for inaccuracy between FS examination and postoperative pathology results. Also, tumor size ≥80 mm and menopausal state were related to misdiagnosis in FS results. Maximal effort should be done to minimize preventable errors during intraoperative FS analysis. Keywords: Frozen section, Pelvic mass, Inaccuracy, Borderline tumors