Root Cause Analysis of Problems in Frozen Section Diagnosis of Bronchioloalveolar Carcinoma and Early Invasive Adenocarcinoma of the Lung.
Alberto M Marchevsky, Robert J McKenna, Ann E Walts. Cedars-Sinai Medical Center, Los Angeles, CA
Background: Pathologists at our hospital are asked to distinguish between bronchioloalveolar carcinoma (BAC), early invasive adenocarcinoma (BAC with <5mm invasion;EIA), and mixed adenocarcinoma (MAC) of the lung at intraoperative frozen section (FS). BAC patients generally undergo wedge/segment resection; most EIA and MAC patients undergo lobectomy. A root cause analysis (RCA) approach was used to review our FS experience.
Design: FS from 224 consecutive single primary resected pulmonary tumors (27 BAC, 46 EIA, and 151 MAC) (median size <2cm) were reviewed and compared with the final diagnoses. Discrepancies were classified as errors or deferrals and as benign vs. malignant, EIA vs. BAC, and BAC or EIA vs. MAC. Reasons for the FS errors and deferrals were tabulated.
Results: FS diagnoses were correct in 183 (8l.7%) of the cases. The 41 (18.3%) discrepancies included 27 (12%) errors and 14 (6.3%) deferrals. These cases had been interpreted by experienced pulmonary pathologists and 15 of the 41 had been reviewed by ≥2 pathologists at the time of FS. One piece of tissue had been frozen in 25 (61%) and ≥2 were frozen in 16 (39%) cases with ≥2 levels examined intraoperatively in 37 (90.2%) cases. There were no false positive diagnoses of malignancy. FS errors involved 21 EIA (13 read as MAC, 5 as benign, 3 as BAC) and 6 BAC (2 read as benign, 2 as EIA, 2 as MAC). Deferrals included 12 benign vs. malignant and 2 EIA vs. BAC. Errors and deferrals were retrospectively attributed to FS misinterpretation in 23 cases as a result of scar with inflammation (9), inflammation with reactive atypia (4), obscuring inflammation (4), scar with inflammation and difficulty estimating extent of invasion (3), minimal cytological atypia (1), scar with inflammation and extensive reactive atypia (1), difficulty estimating extent of invasion (1) and to technical problems in 15 cases resulting from tissue sampling in 11 cases (9 with tumor or invasion only in blocks not frozen and 2 with tumor or invasion seen only at deeper levels of the FS block) and poor fixation in 4 cases. FS slides were not available for review in 3 cases.
Conclusions: -RCA suggests that 19 (46.3%) of the 41 errors and deferrals in FS diagnosis of BAC, EIA, and MAC can be attributed to potentially reducible problems that involve sampling, fixation, and estimating extent of invasion. -Confounding scar and/or severe inlammation remain as more difficult diagnostic problems in 8.5% of the 224 cases.
Tuesday, March 1, 2011 1:00 PM
Poster Session IV # 276, Tuesday Afternoon