Scholarlink

Background <p>Patients of non-small cell lung cancer (NSCLC) face a high risk of recurrence postoperatively, yet there is a lack of comprehensive predictive models that integrate genetic and other multifaceted information.</p> Methods <p>This retrospective cohort study analyzed 911 patients with stage ⅠA1-ⅢA NSCLC in West China Hospital between November 2013 and August 2020, aimed to develop a prediction model incorporating demographic, clinical, pathological, radiological, and genetic data to enhance postoperative risk stratification and inform personalized follow-up and treatment strategies. After Lasso regression and multivariate Cox proportional hazards regression, mutations in <i>JAK1</i> and <i>STK11</i>, disease stage, visceral pleural invasion (VPI), lymphovascular invasion (LVI), tumor spread through air spaces (STAS), radiological density, cavitary sign, and smoking index (SI), were identified as significant risk factors.</p> Results <p>These variables were integrated into a nomogram model to classify patients into three risk categories for recurrence: low (total score ≤ 100), moderate (100 &lt; total score ≤ 175.16), and high (total score &gt; 175.16). The performance of the nomogram was rigorously assessed through calibration curves, and decision curve analysis (DCA) and receiver operating characteristic (ROC) curve analysis both in training set [AUC:0.88, 95% CI: 0.83–0.93 1year; AUC: 0.85, 95% CI: 0.81-0.89r 3year, AUC: 0.85, 95% CI: 0.81-0.895year] and validation set (AUC: 0.85, 95% CI: 0.79–0.92 1year; AUC: 0.83, 95% CI: 0.76–0.89 3year, AUC: 0.84, 95% CI: 0.78–0.91 5year).</p> Conclusions <p>Our multi-source imaging–genomic–clinical model accurately predicts the risk of recurrence after surgery in stage ⅠA1-ⅢA NSCLC patients and can be used for risk stratification to guide clinical follow-up management and postoperative treatment strategies.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Development and validation of a predictive model for recurrence in postoperative patients with stage ⅠA1-ⅢA non-small cell lung cancer

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年11月27日

查看全文
  • Yi Li,
  • Renjie Xu,
  • Jinghong Xian,
  • Zhoufeng Wang,
  • Wang Chen,
  • Weimin Li

摘要

Background

Patients of non-small cell lung cancer (NSCLC) face a high risk of recurrence postoperatively, yet there is a lack of comprehensive predictive models that integrate genetic and other multifaceted information.

Methods

This retrospective cohort study analyzed 911 patients with stage ⅠA1-ⅢA NSCLC in West China Hospital between November 2013 and August 2020, aimed to develop a prediction model incorporating demographic, clinical, pathological, radiological, and genetic data to enhance postoperative risk stratification and inform personalized follow-up and treatment strategies. After Lasso regression and multivariate Cox proportional hazards regression, mutations in JAK1 and STK11, disease stage, visceral pleural invasion (VPI), lymphovascular invasion (LVI), tumor spread through air spaces (STAS), radiological density, cavitary sign, and smoking index (SI), were identified as significant risk factors.

Results

These variables were integrated into a nomogram model to classify patients into three risk categories for recurrence: low (total score ≤ 100), moderate (100 < total score ≤ 175.16), and high (total score > 175.16). The performance of the nomogram was rigorously assessed through calibration curves, and decision curve analysis (DCA) and receiver operating characteristic (ROC) curve analysis both in training set [AUC:0.88, 95% CI: 0.83–0.93 1year; AUC: 0.85, 95% CI: 0.81-0.89r 3year, AUC: 0.85, 95% CI: 0.81-0.895year] and validation set (AUC: 0.85, 95% CI: 0.79–0.92 1year; AUC: 0.83, 95% CI: 0.76–0.89 3year, AUC: 0.84, 95% CI: 0.78–0.91 5year).

Conclusions

Our multi-source imaging–genomic–clinical model accurately predicts the risk of recurrence after surgery in stage ⅠA1-ⅢA NSCLC patients and can be used for risk stratification to guide clinical follow-up management and postoperative treatment strategies.