Prediction of pCR based on clinical-radiomic model in patients with locally advanced ESCC treated with neoadjuvant immunotherapy plus chemoradiotherapy

Xiaohan Wang; Guanzhong Gong; Qifeng Sun; Xue Meng

doi:10.3389/fonc.2024.1350914

Prediction of pCR based on clinical-radiomic model in patients with locally advanced ESCC treated with neoadjuvant immunotherapy plus chemoradiotherapy

Front Oncol. 2024 Mar 20:14:1350914. doi: 10.3389/fonc.2024.1350914. eCollection 2024.

Authors

Xiaohan Wang¹, Guanzhong Gong², Qifeng Sun³, Xue Meng¹

Affiliations

¹ Department of Radiation Oncology, Shandong Cancer Hospital and Institute, Shandong First Medical University and Shandong Academy of Medical Science, Jinan, China.
² Department of Radiotherapy, Shandong Cancer Hospital and Institute, Shandong First Medical University and Shandong Academy of Medical Science, Jinan, China.
³ Department of Thoracic Surgery, Shandong Provincial Hospital Affiliated to Shandong First Medical University, Jinan, China.

Abstract

Background: The primary objective of this research is to devise a model to predict the pathologic complete response in esophageal squamous cell carcinoma (ESCC) patients undergoing neoadjuvant immunotherapy combined with chemoradiotherapy (nICRT).

Methods: We retrospectively analyzed data from 60 ESCC patients who received nICRT between 2019 and 2023. These patients were divided into two cohorts: pCR-group (N = 28) and non-pCR group (N = 32). Radiomic features, discerned from the primary tumor region across plain, arterial, and venous phases of CT, and pertinent laboratory data were documented at two intervals: pre-treatment and preoperation. Concurrently, related clinical data was amassed. Feature selection was facilitated using the Extreme Gradient Boosting (XGBoost) algorithm, with model validation conducted via fivefold cross-validation. The model's discriminating capability was evaluated using the area under the receiver operating characteristic curve (AUC). Additionally, the clinical applicability of the clinical-radiomic model was appraised through decision curve analysis (DCA).

Results: The clinical-radiomic model incorporated seven significant markers: postHALP, ΔHB, post-ALB, firstorder_Skewness, GLCM_DifferenceAverage, GLCM_JointEntropy, GLDM_DependenceEntropy, and NGTDM_Complexity, to predict pCR. The XGBoost algorithm rendered an accuracy of 0.87 and an AUC of 0.84. Notably, the joint omics approach superseded the performance of solely radiomic or clinical model. The DCA further cemented the robust clinical utility of our clinical-radiomic model.

Conclusion: This study successfully formulated and validated a union omics methodology for anticipating the therapeutic outcomes of nICRT followed by radical surgical resection. Such insights are invaluable for clinicians in identifying potential nICRT responders among ESCC patients and tailoring optimal individualized treatment plans.

Keywords: esophageal squamous cell carcinoma; immunotherapy; inflammatory biomarkers; neoadjuvant therapy; radiomics.

Grants and funding

The author(s) declare that financial support was received for the research, authorship, and/or publication of this article. This work has been funded by National Natural Science Foundation of China (81972864 and 82172720), Science and Technology Support Plan for Youth Innovation of Colleges and Universities of Shandong Province of China (2019KJL001).