Scholarlink

Background <p>Locally advanced gastric cancer (LAGC) patients undergoing neoadjuvant chemotherapy (NAC) face a substantial risk of postoperative complications (POC), which may impair recovery and subsequent prognosis. This study aimed to develop a comprehensive predictive model incorporating multiple nutrition-related parameters, particularly CT-based body composition, to predict POC in this population.</p> Methods <p>We retrospectively analyzed 403 LAGC patients who underwent radical gastrectomy after NAC between January 2016 and December 2023. The study cohort was randomly divided into a training set and a validation set at a 7:3 ratio. The univariate analysis and subsequent least absolute shrinkage and selection operator (LASSO) regression were used to determine the significant nutrition-centered indicators for development of the predictive model. The performance of the model was evaluated using the area under the receiver operating characteristic (ROC) curve, calibration curve, and clinical decision curve.</p> Results <p>The incidence of POC was 31.9% (90/282) in the training cohort and 32.2% (39/121) in the validation cohort. Several independent predictors of POC were identified by univariate screening and subsequent LASSO regression analyses. Specifically, a higher nutritional risk score-2002 (NRS-2002), lower skeletal muscle index (SMI), lower pre-NAC body mass index (BMI), more advanced clinical stage, higher intramuscular fat index (IFI), higher subcutaneous fat density (SFD), higher visceral-to-subcutaneous fat ratio (VSR), and longer operation time were all associated with an increased risk of POC. Based on these parameters, we constructed a comprehensive nomogram model, which demonstrated robust and reliable predictive performance: the training set yielded an AUC of 0.861 (95% CI, 0.815–0.907), and sensitivity of 0.896; the validation set yielded an AUC of 0.744 (95% CI, 0.654–0.833), and sensitivity of 0.817. The decision curve analysis indicated that the model could provide benefits to NAC-LAGC patients within a specific threshold range.</p> Conclusions <p>We developed and validated a predictive model based on nutrition-focused predictors to predict POC in NAC-LAGC patients. The model enables accurate risk stratification and offers a practical tool to guide individualized nutritional interventions.</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

A nutrition-focused model for predicting postoperative complications in locally advanced gastric cancer after neoadjuvant chemotherapy: a retrospective cohort study

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年4月29日

查看全文
  • Zongsheng Sun,
  • Shanglong Liu,
  • Shunli Liu,
  • Wentao Xie,
  • Longbo Zheng,
  • Zhengzhao Wang,
  • Dongsheng Wang,
  • Ruiqing Liu

摘要

Background

Locally advanced gastric cancer (LAGC) patients undergoing neoadjuvant chemotherapy (NAC) face a substantial risk of postoperative complications (POC), which may impair recovery and subsequent prognosis. This study aimed to develop a comprehensive predictive model incorporating multiple nutrition-related parameters, particularly CT-based body composition, to predict POC in this population.

Methods

We retrospectively analyzed 403 LAGC patients who underwent radical gastrectomy after NAC between January 2016 and December 2023. The study cohort was randomly divided into a training set and a validation set at a 7:3 ratio. The univariate analysis and subsequent least absolute shrinkage and selection operator (LASSO) regression were used to determine the significant nutrition-centered indicators for development of the predictive model. The performance of the model was evaluated using the area under the receiver operating characteristic (ROC) curve, calibration curve, and clinical decision curve.

Results

The incidence of POC was 31.9% (90/282) in the training cohort and 32.2% (39/121) in the validation cohort. Several independent predictors of POC were identified by univariate screening and subsequent LASSO regression analyses. Specifically, a higher nutritional risk score-2002 (NRS-2002), lower skeletal muscle index (SMI), lower pre-NAC body mass index (BMI), more advanced clinical stage, higher intramuscular fat index (IFI), higher subcutaneous fat density (SFD), higher visceral-to-subcutaneous fat ratio (VSR), and longer operation time were all associated with an increased risk of POC. Based on these parameters, we constructed a comprehensive nomogram model, which demonstrated robust and reliable predictive performance: the training set yielded an AUC of 0.861 (95% CI, 0.815–0.907), and sensitivity of 0.896; the validation set yielded an AUC of 0.744 (95% CI, 0.654–0.833), and sensitivity of 0.817. The decision curve analysis indicated that the model could provide benefits to NAC-LAGC patients within a specific threshold range.

Conclusions

We developed and validated a predictive model based on nutrition-focused predictors to predict POC in NAC-LAGC patients. The model enables accurate risk stratification and offers a practical tool to guide individualized nutritional interventions.