Scholarlink

Background <p>Checkpoint inhibitor pneumonitis (CIP) is an uncommon but clinically severe adverse event that can seriously compromise the quality of life and can be potentially life-threatening in lung cancer patients receiving immune checkpoint inhibitors (ICI). However, there is still a lack of effective predictive models to predict the occurrence of CIP. The aim of this study was to develop a novel scoring system for predicting the risk of CIP based on a nomogram model.</p> Methods <p>We retrospectively screened patients with lung cancer who received ICI treatment at our hospital. The independent risk factors of CIP were identified by the univariable and multivariable analysis of the COX hazard regression model and were integrated to develop a nomogram predictive model. The receiver operating characteristic (ROC) curve, the concordance index (C- index), and the calibration curve were used to evaluate the discrimination and prediction accuracy of the model. The clinical utility of the model was evaluated by decision curve analysis (DCA).</p> Results <p>A total of 2082 cancer patients were included in the analysis. In the final multivariate Cox regression analysis identified that sex, body mass index (BMI), chemotherapy, radiotherapy, C-reactive protein (CRP), CD4/CD8, white blood cell (WBC), ALB/GLB, platelet-to-lymphocyte ratio (PLR), lymphocyte-to-monocyte ratio (LMR), neutrophil-to-platelet ratio (NPR), platelet-to-albumin ratio (PAR), and CRP-to-lymphocyte ratio (CLR) were the independent predictive factors for CIP. Based on these risk factors, a predictive nomogram model was constructed. The C-index for the nomogram model in predicting the probability of CIP at 1 year, 1.5 years, and 2 years was 0.704, 0.718, and 0.725, respectively. The average C-index (SD) was 0.712 (0.004), and the average AUC (SD) was 0.733 (0.005), calculated through 100 iterations of 10-fold cross-validation. The calibration curves demonstrated good concordance, and the DCA indicated that the model had good clinical utility.</p> Conclusions <p>The nomogram was accurate in predicting the occurrence of CIP in patients with lung cancer. This study provides a reference for screening CIP high-risk patients and for individualized 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

Nomogram-based prediction of checkpoint inhibitor pneumonitis in lung cancer patients

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年2月12日

查看全文
  • Dan Tao,
  • Haike Lei,
  • Lisi Sun,
  • Lulu Wang,
  • Wei Zhou,
  • Ying Wang,
  • Yongzhong Wu

摘要

Background

Checkpoint inhibitor pneumonitis (CIP) is an uncommon but clinically severe adverse event that can seriously compromise the quality of life and can be potentially life-threatening in lung cancer patients receiving immune checkpoint inhibitors (ICI). However, there is still a lack of effective predictive models to predict the occurrence of CIP. The aim of this study was to develop a novel scoring system for predicting the risk of CIP based on a nomogram model.

Methods

We retrospectively screened patients with lung cancer who received ICI treatment at our hospital. The independent risk factors of CIP were identified by the univariable and multivariable analysis of the COX hazard regression model and were integrated to develop a nomogram predictive model. The receiver operating characteristic (ROC) curve, the concordance index (C- index), and the calibration curve were used to evaluate the discrimination and prediction accuracy of the model. The clinical utility of the model was evaluated by decision curve analysis (DCA).

Results

A total of 2082 cancer patients were included in the analysis. In the final multivariate Cox regression analysis identified that sex, body mass index (BMI), chemotherapy, radiotherapy, C-reactive protein (CRP), CD4/CD8, white blood cell (WBC), ALB/GLB, platelet-to-lymphocyte ratio (PLR), lymphocyte-to-monocyte ratio (LMR), neutrophil-to-platelet ratio (NPR), platelet-to-albumin ratio (PAR), and CRP-to-lymphocyte ratio (CLR) were the independent predictive factors for CIP. Based on these risk factors, a predictive nomogram model was constructed. The C-index for the nomogram model in predicting the probability of CIP at 1 year, 1.5 years, and 2 years was 0.704, 0.718, and 0.725, respectively. The average C-index (SD) was 0.712 (0.004), and the average AUC (SD) was 0.733 (0.005), calculated through 100 iterations of 10-fold cross-validation. The calibration curves demonstrated good concordance, and the DCA indicated that the model had good clinical utility.

Conclusions

The nomogram was accurate in predicting the occurrence of CIP in patients with lung cancer. This study provides a reference for screening CIP high-risk patients and for individualized treatment strategies.