Abstract

BACKGROUND
Core needle biopsy (CNB) is widely used as the initial sampling method for breast cancer. And because frozen section (FS) diagnosis is rapid and reliable, we studied the diagnostic agreement between the diagnosis of FS of CNB and final diagnosis after surgery to evaluate the diagnostic accuracy of the FS of CNB.
METHODS
Of 409 patients who were preoperatively diagnosed by FS of CNB and who underwent final surgery from 1996 through 2000, 24 cases were found to be ductal carcinoma in situ (DCIS) and 385 cases invasive carcinoma (IC). The diagnoses of FS of CNB were compared with final diagnoses.
RESULTS
The diagnostic accuracy of carcinoma is 63.6% for DCIS and 86.9% for invasive carcinoma. Five cases (1.2%) could not be diagnosed because of material insufficiency for diagnosis. Twenty two cases (5.4%) were diagnosed as benign on FS, among which 20 (90.9%) were misdiagnosed by sampling error. Twenty seven cases (6.7%) were deferred on FS, 4 of these cases were DCIS, 5 were invasive lobular carcinoma (ILC), the rest displayed low nuclear grades or marked freezing artifacts.
CONCLUSIONS
The diagnostic accuracy of FS of CNB is very high except for cases of ILC and low grade DCIS. Considering the advantage of rapid evaluation, more definitive diagnosis, familiarity by pathologists and availability of ancillary study, FS of CNB is very useful method as the preoperative evaluation.

• Keywords: Mammary Neoplasms;Biopsy, Needle;Frozen Sections