Scholarlink

Background and Objective <p>Endoscopic ultrasound (EUS)-guided drainage with metal stents is a standard therapy for walled-off necrosis (WON). However, some patients require re-interventions such as stent unclogging, necrosectomy or percutaneous drainage. This study aimed at identifying predictors of re-intervention to facilitate risk stratification and optimize management.</p> Methods <p>This study included consecutive patients who underwent EUS-guided WON drainage with metal stents between January 2023 and December 2024 at a tertiary referral center. Demographic, clinical and radiological data was collected prospectively. Multi-variate logistic regression identified independent predictors of re-intervention. Model performance was evaluated using receiver operating characteristic (ROC) curve, calibration plot and decision curve analysis. Model’s diagnostic performance was evaluated in the validation cohort.</p> Results <p>Of 500 patients (83.2% male), 28.6% had alcohol-related and 10% had biliary pancreatitis. Re-intervention was required in 24% (<i>n</i> = 120), primarily for stent unclogging (85.8%), nasocystic tube placement (70%) and necrosectomy (50%). Independent predictors of re-intervention&#xa0;included WON size (odds ratio [OR] 1.38; 95% confidence interval [CI], 1.25–1.52), paracolic extension (OR 9.96; 95% CI, 1.77–55.98) and solid debris content (OR 1.10; 95% CI, 1.08–1.13). The model demonstrated good discrimination (area under the curve [AUC] = 0.85) and calibration (Hosmer–Lemeshow <i>p</i> = 0.19). The overall accuracy of the model in the validation cohort was 86.67% (95% CI, 73.21–94.95%). Adverse events occurred in 6% of patients, including bleeding (<i>n</i> = 11), stent migration (<i>n</i> = 13) and sepsis-related death (<i>n</i> = 6).</p> Conclusion <p>Larger WON size, paracolic extension and higher solid debris content are independent predictors of re-intervention after EUS-guided drainage. Early identification of these factors may allow for individualized step-up therapy and improved clinical outcomes.</p> Graphical Abstract <p></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

Re-intervention in endoscopic ultrasound–guided walled-off necrosis drainage with metal stents: Identifying high-risk morphologies

  • 期刊论文

  • 发表日期：

    2026年2月26日

查看全文
  • Nitin Jagtap,
  • Krithi Krishna Koduri,
  • Sana Fathima Memon,
  • Pradev Inavolu,
  • Sundeep Lakhtakia,
  • Shujaath Asif,
  • Jahangeer Basha,
  • Rupjyoti Talukdar,
  • Mohan Ramchandani,
  • Rakesh Kalapala,
  • Vivek Sreekanth,
  • Jagdeesh R. Singh,
  • Tharani Putta,
  • G. Venkat Rao,
  • D. Nageshwar Reddy

摘要

Background and Objective

Endoscopic ultrasound (EUS)-guided drainage with metal stents is a standard therapy for walled-off necrosis (WON). However, some patients require re-interventions such as stent unclogging, necrosectomy or percutaneous drainage. This study aimed at identifying predictors of re-intervention to facilitate risk stratification and optimize management.

Methods

This study included consecutive patients who underwent EUS-guided WON drainage with metal stents between January 2023 and December 2024 at a tertiary referral center. Demographic, clinical and radiological data was collected prospectively. Multi-variate logistic regression identified independent predictors of re-intervention. Model performance was evaluated using receiver operating characteristic (ROC) curve, calibration plot and decision curve analysis. Model’s diagnostic performance was evaluated in the validation cohort.

Results

Of 500 patients (83.2% male), 28.6% had alcohol-related and 10% had biliary pancreatitis. Re-intervention was required in 24% (n = 120), primarily for stent unclogging (85.8%), nasocystic tube placement (70%) and necrosectomy (50%). Independent predictors of re-intervention included WON size (odds ratio [OR] 1.38; 95% confidence interval [CI], 1.25–1.52), paracolic extension (OR 9.96; 95% CI, 1.77–55.98) and solid debris content (OR 1.10; 95% CI, 1.08–1.13). The model demonstrated good discrimination (area under the curve [AUC] = 0.85) and calibration (Hosmer–Lemeshow p = 0.19). The overall accuracy of the model in the validation cohort was 86.67% (95% CI, 73.21–94.95%). Adverse events occurred in 6% of patients, including bleeding (n = 11), stent migration (n = 13) and sepsis-related death (n = 6).

Conclusion

Larger WON size, paracolic extension and higher solid debris content are independent predictors of re-intervention after EUS-guided drainage. Early identification of these factors may allow for individualized step-up therapy and improved clinical outcomes.

Graphical Abstract