[Effects of sampling methods on evaluating post-treatment pathological response in resected non-small cell lung...

Objective: To investigate the effects of sampling methods on pathological assessment of resected non-small cell lung cancer (NSCLC) specimen with tumor maximum diameter >3 cm after neoadjuvant therapy. Methods: NSCLC patients with a large tumor (diameter >3 cm) that were resected after neoadjuvant therapy from June 2020 to July 2023 were retrospectively collected in the Department of Pathology, Shanghai Pulmonary Hospital, Shanghai, China. Sampling methods of the tumor bed were performed in accordance with the international and Chinese experts recommendations for resection specimens following neoadjuvant therapy (recommended sampling method, RSM), and all remaining tumor bed lesions were completely sampled after recommended sampling (complete sampling method, CSM). The difference of pathological response assessment of residual viable tumor (RVT) between RSM and CSM was examined. Results: A total of 90 cases were identified and analyzed, including 39 cases of squamous cell carcinoma and 51 cases of adenocarcinoma, treated with neoadjuvant therapy including chemotherapy in 22 cases (24.4%), targeted therapy in 14 cases (15.6%), and chemoimmunotherapy in 54 cases (60.0%). There were 62 males and 28 females with an average age of (62.7±17.9) years. The average tumor maximum diameter was 4.3 cm (range, 3.1-8.0 cm). The average number of sampled blocks was 8 blocks (range, 5 to 16) and 15 blocks (range, 8 to 36) per case by RSM and CSM, respectively. According to the definition of major pathological response (MPR) in which RVT is ≤10%, the numbers of patients with MPR were 34 cases by RSM and 30 cases by CSM, respectively. Four cases showed inconsistent RVT between the two methods, including one case of squamous cell carcinoma and three cases of adenocarcinoma. The RVT of the four inconsistent cases was 7%, 7%, 5% and 9% (MPR by RSM), and 15%, 15%, 15% and 20% (non-MPR by CSM), respectively. The kappa values of MPR consistency evaluated by the two sampling methods were 0.893 for all cases, 0.906 for squamous cell carcinoma cases and 0.751 for adenocarcinoma cases. According to MPR cut-off of 65% for invasive primary adenocarcinoma, 24 cases and 20 cases achieved MPR by RSM and CSM, respectively. Of the four inconsistent cases, the RVT by RSM was 60% in three cases and 65% in one case (MPR), whereas the RVT by CSM was 70% in three cases and 75% in one case (non-MPR). The kappa value of the two sampling methods was 0.741. Conclusions: There is high consistency between RSM and CSM in the pathological assessment of post-treatment responses in resected NSCLC specimens with tumor maximum diameter larger than 3 cm. When the percentage of RVT cells is close to MPR, re-evaluation of the specimen is required and CSM may be necessary to accurately evaluate the degree of pathological remission, assist in clinical postoperative treatment, and predict patient prognosis.