Intraoperative frozen section for identifying the invasion status of lung adenocarcinoma: A systematic review and meta-analysis

Background: For early-stage lung adenocarcinoma, determining the extent of surgical resection and lymphadenectomy according to the invasion status of the tumour may be more reliable. Intraoperative frozen section (FS) is a potentially effective method to identify the invasion status while its accuracy is still unknown. This meta-analysis aimed to evaluate the accuracy of FS for the invasion status of lung adenocarcinoma.

Methods: We conducted a systematic search of PubMed, Embase, Scopus and Cochrane Library databases (from inception to October 26, 2018) to identify studies investigating the accuracy of FS for the invasion status of lung adenocarcinoma. The accuracy of FS was evaluated by calculating the pooled concordance rates (CCR) between FS and final pathology and the pooled sensitivity, specificity, and other parameters of FS for discriminating pre-/minimally invasive adenocarcinoma from invasive adenocarcinoma (IAC). The negative predictive value (NPV) of FS for diagnosing IAC was also calculated to evaluate the chance of underestimation.

Results: Six eligible studies were included. The pooled CCR for differentiating pre-invasive adenocarcinoma, minimally invasive adenocarcinoma and IAC was 88% (95% CI, 84%-93%). When pre-invasive adenocarcinoma and minimally invasive adenocarcinoma were classified as a group, the pooled CCR, sensitivity, specificity of FS for differentiating pre-/minimally invasive adenocarcinoma from IAC were 95% (95% CI, 94%-97%), 95% (95% CI, 92%-97%), 95% (95% CI, 80%-99%), respectively. The pooled NPV of FS for diagnosing IAC was 95% (95% CI, 92%-97%).

Conclusions: Intraoperative FS is reliable for identifying the invasion status of lung adenocarcinoma, with high diagnostic accuracy for differentiating pre-/minimally invasive adenocarcinoma from IAC.