Accuracy of Frozen Section in the Intraoperative Diagnosis of Ophthalmic Diseases
Jignesh Parikh, Remedios Huerto, Yo-Chen Chang, Steven McCormick, Codrin Iacob, Tatyana Milman. University of Tennessee, Memphis, TN; The New York Eye and Ear Infirmary, New York, NY
Background: Frozen section procedure is often employed during surgery to diagnose obscure lesions, to differentiate between malignant and benign lesions, to determine negative margins in oncological surgeries, and to confirm the presence of lesional tissue. In ophthalmic surgery this procedure is not widely used. This may be attributed to the small quantity of tissue obtained during surgery and lack of proper training in frozen section preparation and diagnosis among ophthalmic pathologists. In this study, we aim to determine the accuracy of intraoperative frozen section diagnosis in ophthalmic pathology at a major tertiary care ophthalmic institute.
Design: Data collection was done by searching the computer data bank for intra-operative and final diagnoses of ophthalmic specimens submitted to the Pathology Department between January 2005 and January 2010. Pathology reports were reviewed and classified as: accurate, inaccurate, and deferred.
Results: Between 2005 and 2010, 1501 ophthalmic specimens were submitted for frozen section diagnosis. Thirty cases (1.9%) were deferred for permanent section diagnosis. Out of the remaining 1471 cases, 1359 (92.4%) frozen section diagnoses were consistent with final diagnosis based on permanent sections, while 112 (7.6%) cases showed discrepancies between frozen section and final diagnosis. By location, highest diagnostic accuracy was achieved for anterior chamber (100%), optic nerve (100%), temporal artery (100%), and eyelid (96.5%) biopsies. Diagnostic accuracy was slightly lower for biopsies of conjunctiva (89.4%), orbit (88.1%), and eyebrow (87.5%). Clinically significant negative discrepancies were highest in the diagnosis of conjunctival/corneal lesions (5.02%). Clinically significant positive discrepancies were highest in orbital biopsies (6.5%). The specificity and sensitivity for frozen section diagnosis are 92.4%; with a positive predictive value of 93.1% and a negative predictive value of 91.6%.
Conclusions: Our data confirm that frozen section diagnosis is a reliable method for ophthalmic surgeries. Diagnostic errors may be reduced by adequate tissue sampling, complete clinical information, and good communication between surgeon and pathologist.
Wednesday, March 21, 2012 1:00 PM
Poster Session VI # 314, Wednesday Afternoon