Integration of Clinical, Radiomics, and 2.5D Deep Learning-Based Multiple Instance Learning Features for Predicting Pathological Complete Response in Esophageal Squamous Cell Carcinoma Following Neoadjuvant Immunotherapy and Chemotherapy: A Multicenter Comparative Study
试验速览
- 阶段
- 不适用
- 状态
- 已完成
- 发起方
- 入组人数
- 363
- 试验地点
- 1
- 主要终点
- Pathological Complete Response (pCR) at Surgery
研究概览
简要总结
This multicenter, retrospective cohort study reviews the medical records and CT scans of adults with esophageal squamous cell carcinoma (ESCC) who received neoadjuvant immunotherapy plus chemotherapy before surgery at three hospitals in China. The goal is to develop and validate a computer-assisted model that predicts which patients achieve a pathological complete response (pCR)-meaning no residual tumor is found at surgery-after preoperative treatment. Accurate pCR prediction may help clinicians personalize care and avoid unnecessary treatments in likely non-responders.
The study includes 363 patients. For each patient, routinely collected clinical information and preoperative venous-phase chest CT images were analyzed. From CT images, both radiomics features and features learned by a "2.5D" deep learning approach with multiple-instance learning (MIL) were extracted. These were combined with clinical variables to create a multimodal prediction model. Model performance will be evaluated using standard metrics and validated in internal and external cohorts.
Patients typically received two cycles of taxane-platinum chemotherapy (paclitaxel with cisplatin or carboplatin) combined with camrelizumab every 2-3 weeks before surgery; CT scans were performed within 14 days prior to starting therapy. Surgery (R0 resection) was performed 6-8 weeks after treatment, and pCR was determined by the postoperative pathology report.
This is an observational study; no treatments are assigned by protocol. The study was approved by the Ethics Committee of Nanjing Medical University, with informed consent waived due to the retrospective design.
详细描述
Design and Setting. Multicenter, retrospective cohort study conducted at three affiliated hospitals in China. A total of 363 consecutive ESCC patients met eligibility criteria and were split into a training cohort (n=107), internal validation cohort (n=45), and two external test cohorts (n=129 and n=82).
Population. Inclusion criteria: biopsy-confirmed ESCC; locally advanced disease by AJCC 8th edition (cT1N1-T3N0-3M0) on contrast-enhanced CT; completion of standardized neoadjuvant chemo-immunotherapy; availability of high-quality venous-phase chest CT (slice thickness ≤5 mm) within 14 days before therapy; R0 resection 6-8 weeks post-treatment; and a definitive postoperative pathology report documenting pCR. Key exclusions: non-squamous histology, distant metastasis, synchronous malignancies, poor/no venous-phase imaging, slice thickness >5 mm, severe artifacts, incomplete tumor visualization, incomplete treatment, or missing endpoints.
Neoadjuvant Regimen and Imaging. Patients generally received two cycles of taxane-platinum chemotherapy (paclitaxel plus cisplatin or carboplatin) combined with camrelizumab every 2-3 weeks prior to surgery. CT imaging was standardized to venous-phase contrast with 1-5 mm slices; scans without venous phase or >5 mm thickness were excluded. Tumor volumes were delineated by two radiologists; disagreements were adjudicated by a senior radiologist, and features were harmonized via resampling and intensity normalization.
Feature Extraction and Modeling. The pipeline integrated: (1) clinical variables; (2) conventional CT radiomics features (shape, first-order, GLCM, GLRLM, GLSZM, etc.); and (3) 2.5D deep learning slice embeddings aggregated to the patient level using multiple-instance learning (MIL). The 2.5D approach uses adjacent slices in axial/sagittal/coronal planes with ResNet backbones; attention-based MIL plus histogram/BoW-TF-IDF descriptors summarized slice-level predictions. Feature selection used univariate filters, correlation screening, mRMR, and LASSO before training classifiers (logistic regression, SVM, Random Forest, Extra-Trees, LightGBM).
Outcomes and Analysis.
研究设计
- 研究类型
- Observational
- 观察模型
- Cohort
- 时间视角
- Retrospective
入排标准
- 年龄范围
- 18 Years 至 —（Adult, Older Adult）
- 性别
- All
- 接受健康志愿者
- 否
入选标准
- •Biopsy-confirmed esophageal squamous cell carcinoma (ESCC). Locally advanced disease per AJCC 8th ed. (cT1N1-T3N0-3M0) on contrast-enhanced CT.
- •Completed standardized neoadjuvant chemo-immunotherapy (e.g., paclitaxel + cisplatin/carboplatin with camrelizumab every 2-3 weeks) prior to surgery.
- •High-quality venous-phase chest CT (slice thickness ≤5 mm) obtained within 14 days before therapy start.
- •Underwent R0 resection 6-8 weeks after therapy. Availability of a definitive postoperative pathology report to ascertain pCR status.
排除标准
- •Non-squamous histology; distant metastasis (M1); synchronous malignancies. Inadequate imaging quality (no venous phase, slice thickness >5 mm, severe artifacts, or incomplete tumor visualization).
- •Did not complete the full treatment course or had missing endpoints (e.g., no pathological response record or lost to follow-up).
结局指标
主要结局
Pathological Complete Response (pCR) at Surgery
时间窗: At time of surgery after neoadjuvant therapy (~6-8 weeks post-treatment).
pCR is defined as no residual viable tumor in the resected specimen (esophagus and regional lymph nodes) after neoadjuvant immunotherapy plus chemotherapy. pCR status is determined from the postoperative surgical pathology report. This is an observational cohort; treatments were standard-of-care and not assigned by protocol. pCR is abstracted from medical records for all eligible patients.
次要结局
- Diagnostic Performance of the Multimodal Model for Predicting pCR(From baseline CT (≤14 days before therapy start) to surgery (≈6-8 weeks post-therapy); analysis performed at study completion.)
研究者
Zhiyun Xu
Professor of Thoracic Surgery
Nanjing Medical University