Touch prep (TP) cytology as a tool for determining pulmonary parenchymal resection margin status

Abstract

18095 Background: Accurate determination of parenchymal resection margins is critical in excision of primary and metastatic lung cancers. Complete microscopic examination of the entire margin is difficult and may not give timely results. Staple-line excision may compromise the accuracy of histologic margin exam (HME). TP offers a novel method to obtain pulmonary parenchymal margin status intraoperatively. Methods: Patients undergoing wedge resection for a known malignant lung lesion were studied prospectively. At the time of resection, the specimen stapled margin underwent TP on 3 glass slides for cytological analysis. The presence or absence of malignant cells on TP was correlated to final HME. Local recurrence and overall survival by TP and HME status were compared using logrank test. Results: Thirty specimens from 29 patients were studied between December 2002 and April 2006. Fifteen specimens (50%) were right- sided; 9/29 patients (31%) were male. Median age was 66 years (range 28 - 81). Histologies included non-small cell lung cancer (NSCLC) in 22 (73%), and metastases from colorectal cancer (4/30, 13%), sarcoma (2/20, 7%), and breast (1/30, 3%). All TP negative specimens had negative HME. TP was positive in 10 of 30 (33%) specimens, 6 of which had negative HME. These included 4 NSCLC and metastases from sarcoma (1) and colon (1). Mean margin distance was 4 ± 1.7 mm for TP/HME negative specimens, 0.5 ± 0 mm for TP/HME positive specimens, and 5 ± 1.7 mm for TP positive/HME negative patients. There was no difference in time to local recurrence between patients with positive margins by TP vs HME (18.6 and 18.63 months), or in TP and HME negative patients (30.1 and 30.6 months). Conclusion: TP analysis of lung parenchymal margins is safe and feasible. A negative TP is highly predictive of negative HME. Factors contributing to false-positive TP remain unresolved. Multi-institutional prospective studies are indicated for further characterization of this promising tool. No significant financial relationships to disclose.