Scholarlink

Background <p>The response to immunotherapy is limited in advanced biliary tract cancer (BTC). Response prediction is a serious challenge in the clinic.</p> Methods <p>This study included 60 patients with advanced BTC who received anti-PD-1 treatment. Among these patients, 30 were subjected to 520 gene panel sequencing, and 50 were subjected to multiplex circulating cytokine testing. The entropy and mutation features were analysed via the optimized pipeline based on our previous work. The repeated LASSO algorithm was used to identify the optimal features. The associations between sequence features and cell communications were explored by analysing single-cell transcriptome data from BTC (GSE125449). Cox regression was used to develop the integrated model. Time-dependent C-index, Kaplan‒Meier, and receiver operating characteristic (ROC) curves were used to assess the prediction performance.</p> Results <p>TP53, NRAS, FBXW7, and APC were identified as prognosis-related genes. The average C-indices of sequence entropy (0.819) and mutation (0.817) for overall survival (OS) were significantly greater than those of tumour mutation burden (TMB, 0.392) and mutation score (0.638). Single-cell transcriptome data revealed that TP53, KRAS, and NRAS were enriched in plasmacytoid dendritic cells (pDCs) and that the communication between pDCs and macrophages was mediated through the CXCL signalling pathway. The integrated model (EM-CXCL10) showed powerful predictive performance for survival status (AUC: 0.863, 95% CI: 0.643–0.972) and objective response rate (AUC: 0.990, 95% CI: 0.822–1.000).</p> Conclusions <p>This study constructed a multidimensional strategy that might indicate the prognosis of BTC immunotherapy, enabling the recognition of BTC patients who would benefit from immunotherapy, thereby guiding personalized clinical decision-making.</p> <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

Alterations in genomic features and the tumour immune microenvironment predict immunotherapy outcomes in advanced biliary tract cancer patients

  • 期刊论文

  • 发表日期：

    2025年4月10日

查看全文
  • Chenjun Huang,
  • Zhiquan Qiu,
  • Honglian Huang,
  • Xiao Xiao,
  • Fei Du,
  • Jun Ji,
  • Xuewen Xu,
  • Xiaoqing Jiang,
  • Ying Wang,
  • Chunfang Gao

摘要

Background

The response to immunotherapy is limited in advanced biliary tract cancer (BTC). Response prediction is a serious challenge in the clinic.

Methods

This study included 60 patients with advanced BTC who received anti-PD-1 treatment. Among these patients, 30 were subjected to 520 gene panel sequencing, and 50 were subjected to multiplex circulating cytokine testing. The entropy and mutation features were analysed via the optimized pipeline based on our previous work. The repeated LASSO algorithm was used to identify the optimal features. The associations between sequence features and cell communications were explored by analysing single-cell transcriptome data from BTC (GSE125449). Cox regression was used to develop the integrated model. Time-dependent C-index, Kaplan‒Meier, and receiver operating characteristic (ROC) curves were used to assess the prediction performance.

Results

TP53, NRAS, FBXW7, and APC were identified as prognosis-related genes. The average C-indices of sequence entropy (0.819) and mutation (0.817) for overall survival (OS) were significantly greater than those of tumour mutation burden (TMB, 0.392) and mutation score (0.638). Single-cell transcriptome data revealed that TP53, KRAS, and NRAS were enriched in plasmacytoid dendritic cells (pDCs) and that the communication between pDCs and macrophages was mediated through the CXCL signalling pathway. The integrated model (EM-CXCL10) showed powerful predictive performance for survival status (AUC: 0.863, 95% CI: 0.643–0.972) and objective response rate (AUC: 0.990, 95% CI: 0.822–1.000).

Conclusions

This study constructed a multidimensional strategy that might indicate the prognosis of BTC immunotherapy, enabling the recognition of BTC patients who would benefit from immunotherapy, thereby guiding personalized clinical decision-making.