Frozen section pathology at time of hysterectomy accurately predicts endometrial cancer in patients with preoperative diagnosis of atypical endometrial hyperplasia

• Published in: Gynecologic oncology Vol. 125; no. 3; pp. 536 - 540
• Main Authors: Morotti, Matteo, Menada, Mario Valenzano, Moioli, Melita, Sala, Paolo, Maffeo, Ilaria, Abete, Luca, Fulcheri, Ezio, Menoni, Stefania, Venturini, Pierluigi, Papadia, Andrea
• Format: Journal Article
• Language: English
• Published: United States Elsevier Inc 01.06.2012
• Subjects: Adult; Aged; Atypical endometrial hyperplasia; Endometrial cancer; Endometrial Hyperplasia - diagnosis; Endometrial Hyperplasia - pathology; Endometrial Hyperplasia - surgery; Endometrial Neoplasms - diagnosis; Endometrial Neoplasms - pathology; Female; Frozen section; Frozen Sections; Hematology, Oncology and Palliative Medicine; Humans; Hysterectomy; Middle Aged; Monitoring, Intraoperative - methods; Obstetrics and Gynecology; Predictive Value of Tests; Retrospective Studies; Endometrial cancer; Frozen section; Atypical endometrial hyperplasia
• ISSN: 0090-8258; 1095-6859
• DOI: 10.1016/j.ygyno.2012.02.011

Abstract Objectives A significant number of women diagnosed with atypical endometrial hyperplasia (AEH) on endometrial biopsy will be diagnosed with endometrial cancer (EC) on the hysterectomy specimen at permanent section. Surgical treatment for AEH and EC differ substantially. We have assessed the concordance in EC between frozen and permanent sections on patients undergoing hysterectomy for AEH. Materials and methods A retrospective review of 66 frozen sections on patients undergoing hysterectomy for AEH was performed. Frozen and permanent section diagnoses were categorized as negative or positive for malignancy. Permanent section carcinomas were classified as low or high risk based on their histopathology, myometrial invasion and differentiation. Correlation between frozen and permanent section and sensitivity, specificity, PPV, NPV and accuracy of frozen section in predicting EC in permanent section were calculated. Likelihood of diagnosing EC on frozen section was compared based on risk stratification at permanent section. Results Frozen and permanent sections revealed malignancy in 43.9% and 56% of the patients respectively. 94.1% of high risk carcinomas were identified as EC at frozen section as compared to 55% of low risk EC. Concordance was good (κ = 0.75). Sensitivity, specificity, NPV, PPV and accuracy in predicting EC at frozen section were 73%, 93.1%, 73% and 93.1% respectively. Carcinomas were detected at frozen section significantly more often if they were at high risk. Conclusions The substantial agreement between frozen and permanent sections allows minimizing under- and overtreatment of women undergoing hysterectomy for AEH. High risk EC are efficiently identified in frozen section.