A combined predicting model for benign esophageal stenosis after simultaneous integrated boost in esophageal squamous cell carcinoma patients (GASTO1072)

Weitong Liu; Chengbing Zeng; Siyan Wang; Yizhou Zhan; Ruihong Huang; Ting Luo; Guobo Peng; Yanxuan Wu; Zihan Qiu; Derui Li; Fangcai Wu; Chuangzhen Chen

doi:10.3389/fonc.2022.1026305

A combined predicting model for benign esophageal stenosis after simultaneous integrated boost in esophageal squamous cell carcinoma patients (GASTO1072)

Front Oncol. 2022 Dec 22:12:1026305. doi: 10.3389/fonc.2022.1026305. eCollection 2022.

Authors

Weitong Liu^{1

2}, Chengbing Zeng¹, Siyan Wang¹, Yizhou Zhan¹, Ruihong Huang¹, Ting Luo^{1

3}, Guobo Peng¹, Yanxuan Wu¹, Zihan Qiu⁴, Derui Li¹, Fangcai Wu¹, Chuangzhen Chen¹

Affiliations

¹ Department of Radiation Oncology, Cancer Hospital of Shantou University Medical College, Shantou, China.
² Department of Radiation Oncology, Jieyang People's Hospital, Jeiyang, China.
³ Department of Radiation Oncology, Shenshan Central Hospital, Sun Yat-Sen Memorial Hospital, Sun Yat-Sen University, Shanwei, China.
⁴ Department of Otolaryngology-Head and Neck Surgery, The First Affiliated Hospital of Shantou University Medical College, Shantou, China.

Abstract

Purpose: We aimed to develop a combined predicting model for benign esophageal stenosis (BES) after simultaneous integrated boost (SIB) with concurrent chemotherapy in patients with esophageal squamous cell carcinoma (ESCC).

Methods: This study included 65 patients with EC who underwent SIB with chemotherapy. Esophageal stenosis was evaluated using esophagograms and the severity of eating disorders. Risk factors were investigated using univariate and multivariate analyses. Radiomics features were extracted based on contrast-enhanced CT (CE-CT) before treatment. The least absolute shrinkage and selection operator (LASSO) regression analysis was used for feature selection and radiomics signature construction. The model's performance was evaluated using Harrell's concordance index and receiver operating characteristic curves.

Results: The patients were stratified into low- and high-risk groups according to BES after SIB. The area under the curves of the clinical model, Rad-score, and the combined model were 0.751, 0.820 and 0.864, respectively. In the validation cohort, the AUCs of these three models were 0.854, 0.883 and 0.917, respectively. The Hosmer-Lemeshow test showed that there was no deviation from model fitting for the training cohort (p=0.451) and validation cohort (p=0.481). The C-indexes of the nomogram were 0.864 and 0.958 for the training and validation cohort, respectively. The model combined with Rad-score and clinical factors achieved favorable prediction ability.

Conclusion: Definitive chemoradiotherapy could alleviate tumor-inducing esophageal stenosis but result in benign stenosis. We constructed and tested a combined predicting model for benign esophageal stenosis after SIB. The nomogram incorporating both radiomics signature and clinical prognostic factors showed favorable predictive accuracy for BES in ESCC patients who received SIB with chemotherapy.

Trial registration number and date of registration: Registered in www.Clinicaltrial.gov, ID: NCT01670409, August 12, 2012.

Keywords: chemoradiotherapy; esophageal cancer; esophageal stenosis; radiomics; radiotherapy.

Associated data

ClinicalTrials.gov/NCT01670409

Grants and funding

This study was funded by Shantou University Medical College Clinical Research Enhancement Initiative, N0201424 (to CC); Science and Technology Special Fund of Guangdong Province of China, 2019-132 (to CC); Strategic and Special Fund for Science and Technology Innovation of Guangdong Province of China, 180918114960704 (to CC); Innovative Research Group Project of the National Natural Science Foundation of China, 82173079 (to CC).