Validity of frozen section in cervical conization.

ABSTRACT

Objective: To evaluate the validity of pathological diagnosis of cervical cone specimens prepared by frozen section compared with paraffin section. Study design: Diagnostic test evaluation. Setting: Pathology division, Department of Obstetrics and Gynecology, faculty of Medicine Siriraj Hospital, Mahidol University. Methods: Cervical cone specimens from 78 patients who underwent cold knife conization at Siriraj Hospital from October 1997 to September 1998 were processed by frozen section technique and the pathological diagnoses were made immediately. The remaining cone tissue from each specimen was processed to produce permanent paraffin sections for a final diagnosis. The frozen and permanent pathological diagnoses were compared. Results: The pathological diagnosis from frozen section was in complete agreement with the permanent section in 60.26% of patients. When the subjects were divided into three groups; normal and CIN I, CIN II-III and MIC and invasive cancer, the Kappa analysis for agreement of the pathological diagnoses between the two methods was 0.46 (fair agreement). For the diagnosis of invasive cancer by frozen section, the sensitivity, specificity, positive and negative predictive value, false negative and false positive were 62.5, 97, 71.4, 95.8, 37.5 and 1.9%, respectively. There were three cases of invasive cancer on permanent paraffin section which were diagnosed by frozen section as MIC in tow cases and CIN III in the other. Conclusion: Frozen section evaluation of a cervical cone specimen carries only a moderate degree of agreement with permanent paraffin section. For the diagnosis of invasive cancer in this study, frozen section has a low sensitivity and a high false negative rate. The diagnosis of microinvasive cancer was subject to significant error. The diagnosis of microinvasive cancer by frozen section needs additional careful review of the permanent section.