Please select Print from the file menu or right mouse click and choose print from the menu to print this page.

[1560] Lessons Learnt from Mistakes and Deferrals in Frozen Section Diagnosis of Pulmonary Lesions: An Evidence Based Pathology Approach

R Gupta, AS Khuranna, RJ McKenna, AM Marchevsky. Cedars Sinai Medical Center, LA, CA

Background: The diagnosis of lung lesions by frozen section (FS) diagnosis can be difficult as inflammatory atypia and artifacts introduced by the procedure may closely simulate a malignancy. Deferred diagnoses or mistakes may result in second thoracotomies or unnecessary lobectomies.
Design: We reviewed our experience with 2,405 patients that underwent FS from 2002-2007. In 143 of these cases, including 65 benign conditions, 35 adenocarcinomas, 22 carcinoids, 15 lymphomas and 6 other neoplasms the diagnosis was deferred or erroneous. Deferral (DR) and error rates (ER) for the FS diagnoses of reactive atypia (RA) and adenocarcinoma (AC) were calculated. We reviewed all FS slides and corresponding permanent sections and compared the presence of 23 pathological parameters in AC and RA using descriptive statistics. Each pathological parameter found to be significant was assessed for its diagnostic applicability using odds ratio (OR).
Results: Overall DR and ER were 4.36% and 1.58% respectively. DR and ER for the diagnoses of lymphoma and carcinoid tumor were considerably higher at 10.38% and 18.18%. Thirteen of 23 pathological parameters, including grossly visible nodule, glandular/papillary growth patterns, granulomas, atypical mitoses, and others were significant at a p<0.05 level. However, the diagnostic applicability of individual parameters was found to be variable, somewhat limited and independent of their chi-square p-value, as shown by OR ranging from 0.16 to 13.
Conclusions: Pathological parameters that can be helpful to distinguish RA, AC and other conditions on pulmonary FS are discussed. Recognition that lymphomas and carcinoids can be particularly difficult to diagnose by FS may reduce ER and DR.An evidence-based approach that distinguishes between statistically significant and really useful diagnostic parameters in daily pathology practice is emphasized.
Category: Pulmonary

Monday, March 3, 2008

Poster # 215, Monday Morning