There are great expectations for artificial intelligence (AI) in medicine. We aimed to develop an AI prognostic model for surgically resected non-small cell lung cancer (NSCLC). This study enrolled 1049 patients with pathological stage I–IIIA surgically resected NSCLC at Kyushu University. We set 17 clinicopathological factors and 30 preoperative and 22 postoperative blood test results as explanatory variables. Disease-free survival (DFS), overall survival (OS), and cancer-specific survival (CSS) were set as objective variables. The eXtreme Gradient Boosting (XGBoost) was used as the machine learning algorithm. The median age was 69 (23–89) years, and 605 patients (57.7%) were male. The numbers of patients with pathological stage IA, IB, IIA, IIB, and IIIA were 553 (52.7%), 223 (21.4%), 100 (9.5%), 55 (5.3%), and 118 (11.2%), respectively. The 5-year DFS, OS, and CSS rates were 71.0%, 82.8%, and 88.7%, respectively. Our AI prognostic model showed that the areas under the curve of the receiver operating characteristic curves of DFS, OS, and CSS at 5 years were 0.890, 0.926, and 0.960, respectively. The AI prognostic model using XGBoost showed good prediction accuracy and provided accurate predictive probability of postoperative prognosis of NSCLC.
Background: Immunotherapy has become a standard‐of‐care for patients with non‐small‐cell lung cancer (NSCLC). Although several biomarkers, such as programmed cell death‐1, have been shown to be useful in selecting patients likely to benefit from immune checkpoint inhibitors (ICIs), more useful and reliable ones should be investigated. The prognostic nutritional index (PNI) is a marker of the immune and nutritional status of the host, and is derived from serum albumin level and peripheral lymphocyte count. Although several groups reported its prognostic role in patients with NSCLC receiving a single ICI, there exist no reports which have demonstrated its role in the first‐line ICI combined with or without chemotherapy.Materials and Methods: Two‐hundred and eighteen patients with NSCLC were included in the current study and received pembrolizumab alone or chemoimmunotherapy as the first‐line therapy. Cutoff value of the pretreatment PNI was set as 42.17.Results: Among 218 patients, 123 (56.4%) had a high PNI (≥42.17), while 95 (43.6%) had a low PNI (<42.17). A significant association was observed between the PNI and both the progression‐free survival (PFS; hazard ratio [HR] = 0.67, 95% confidence interval [CI]: 0.51–0.88, p = 0.0021) and overall survival (OS; HR = 0.46, 95% CI: 0.32–0.67, p < 0.0001) in the entire population, respectively. The multivariate analysis identified the pretreatment PNI as an independent prognosticator for the PFS (p = 0.0011) and OS (p < 0.0001), and in patients receiving either pembrolizumab alone or chemoimmunotherapy, the pretreatment PNI remained an independent prognostic factor for the OS (p = 0.0270 and 0.0006, respectively).Conclusion: The PNI might help clinicians appropriately identifying patients with better treatment outcomes when receiving first‐line ICI therapy.
OBJECTIVES: Spread through air spaces (STAS) is considered a poor prognostic factor in patients with resected non-small lung cell cancer; however, the clinical significance of STAS extent remains unclear. We hypothesized that the further the tumour cells spread from the tumour edge, the worse the prognosis becomes. METHODS: This study retrospectively reviewed the data of 642 patients with completely resected pathological stage I–III non-small lung cell cancer between 2008 and 2018. The maximum spread distance (MSD) from the tumour edge to the farthest STAS was quantitatively evaluated, and STAS was categorized as limited (MSD ≤1000 μm) or extended (MSD >1000 μm), based on the median MSD. Recurrence-free survival (RFS) and overall survival (OS) were compared among patients stratified by STAS classification. RESULTS: Patients were classified into STAS-negative (n = 382, 59.6%), limited STAS (n = 130, 20.2%) and extended STAS (n = 130, 20.2%) groups. Extended STAS was associated with a high maximum standardized uptake value, advanced pathological stage and vascular invasion compared with limited STAS. The extended STAS group demonstrated significantly shorter RFS and OS than both the limited STAS and STAS-negative groups (both P < 0.001 for RFS; P = 0.007 and P < 0.001 for OS, respectively). Multivariable analysis showed that extended STAS was an independent prognostic factor for both RFS and OS (P < 0.001, P < 0.001, respectively). CONCLUSIONS: The distance from tumour edge to STAS affects prognosis in patients with completely resected non-small lung cell cancer. Clinical registration number: IRB approval number: 2019-232