Prediction of Spread Through Air Spaces By Intraoperative Frozen Section for Patients With cT1N0M0 Invasive Lung Adenocarcinoma

Objective: To investigate the value of intraoperative assessment of spread through air spaces (STAS) on frozen sections (FS) in peripheral small-sized lung adenocarcinoma. Background: Surgical decision-making based on FS diagnosis of STAS may be useful to prevent local control failure after sublobar resection. Methods: We conducted a multicenter prospective observational study of consecutive patients with cT1N0M0 invasive lung adenocarcinoma to evaluate the accuracy of FS for the intraoperative detection of STAS. The final pathology (FP) diagnosis of STAS was based on corresponding permanent paraffin sections. Results: This study included 878 patients with cT1N0M0 invasive lung adenocarcinoma. A total of 833 cases (95%) were assessable for STAS on FS. In total, 26.4% of the cases evaluated positive for STAS on FP, whereas 18.2% on FS. The accuracy, sensitivity, and specificity of FS diagnosis of STAS were 85.1%, 56.4%, and 95.4%, respectively, with moderate agreement (κ=0.575). Interobserver agreement was substantial (κ=0.756) among the 3 pathologists. Subgroup analysis based on tumor size or consolidation-to-tumor ratio all showed moderate agreement for concordance. After rigorous reassessment of false-positive cases, the presence of artifacts may be the main cause of interpretation errors. In addition, true positive cases showed more high-grade histologic patterns and more advanced p-TNM stages than false negative cases. Conclusions: This is the largest prospective observational study to evaluate STAS on FS in patients with cT1N0M0 invasive lung adenocarcinoma. FS is highly specific with moderate agreement, but is not sensitive for STAS detection. While appropriately reporting STAS on FS may provide surgeons with valuable information for intraoperative decision-making, better approaches are needed.