建立早期非小细胞肺癌（NSCLC）治疗方式选择的预测模型。立体定向全身放射治疗（SBRT）和微创手术（MIS）用于早期非小细胞肺癌的局部治疗。然而，由于影响决策过程的因素众多，选择接受 SBRT 或 MIS 的患者仍然具有挑战性。我们分析了 2020 年 1 月至 2023 年 7 月期间接受预期 MIS 或 SBRT 治疗的 1291 名临床 I 期 NSCLC 患者。 SBRT 选择模型是基于多变量逻辑回归分析创建的。受试者工作特征曲线分析将队列分为 3 个与治疗相关的风险类别。研究了术后结果、复发和总生存期 (OS)，以评估模型的性能。总共有 1116 名患者接受了 MIS 治疗，175 名患者接受了 SBRT 治疗。预测模型包括年龄、体力状态、既往肺切除术、MSK-Frailty 评分、FEV1 和 DLCO，并显示曲线下面积为 0.908（95%CI，0.876-0.938）。根据概率评分 (n=1197)，将患者分为低风险（MIS，n=970 和 SBRT，n=28）、中风险（MIS，n=96 和 SBRT，n=53）和高风险类别（MIS，n=10 和 SBRT，n=40）。治疗方式与 OS 无关（SBRT 的 HR，1.67 [95%CI：0.80-3.48]；P=0.20）。临床专业知识可以转化为稳健的预测模型，指导 I 期 NSCLC 患者选择 MIS 与SBRT 并有效地将其分为三个不同的风险组。中级患者可从多学科评估中获益最多。版权所有 © 2024 Wolters Kluwer Health, Inc. 保留所有权利。

To generate a prediction model for selection of treatment modality for early-stage non-small cell lung cancer (NSCLC).Stereotactic body radiotherapy (SBRT) and minimally invasive surgery (MIS) are used in the local treatment of early-stage NSCLC. However, selection of patients for either SBRT or MIS remains challenging, due to the multitude of factors influencing the decision-making process.We analyzed 1291 patients with clinical stage I NSCLC treated with intended MIS or SBRT from January 2020 to July 2023. A prediction model for selection for SBRT was created based on multivariable logistic regression analysis. The receiver operating characteristic curve analysis stratified the cohort into 3 treatment-related risk categories. Post-procedural outcomes, recurrence and overall survival (OS) were investigated to assess the performance of the model.In total, 1116 patients underwent MIS and 175 SBRT. The prediction model included age, performance status, previous pulmonary resection, MSK-Frailty score, FEV1 and DLCO, and demonstrated an area-under-the-curve of 0.908 (95%CI, 0.876-0.938). Based on the probability scores (n=1197), patients were stratified into a low-risk (MIS, n=970 and SBRT, n=28), intermediate-risk (MIS, n=96 and SBRT, n=53) and high-risk category (MIS, n=10 and SBRT, n=40). Treatment modality was not associated with OS (HR of SBRT, 1.67 [95%CI: 0.80-3.48]; P=0.20).Clinical expertise can be translated into a robust predictive model, guiding the selection of stage I NSCLC patients for MIS versus SBRT and effectively categorizing them into three distinct risk groups. Patients in the intermediate category could benefit most from multidisciplinary evaluation.Copyright © 2024 Wolters Kluwer Health, Inc. All rights reserved.