This prospective, multicenter diagnostic accuracy study will develop and validate a standardized 3.0-T magnetic resonance imaging (MRI) acquisition protocol and structured reporting system for primary esophageal squamous cell carcinoma (ESCC). Approximately 500 adults scheduled for curative esophagectomy will be enrolled consecutively into two clinically defined cohorts. In Cohort A, patients proceeding directly to surgery will have MRI-based T staging compared with postoperative pathological T stage. In Cohort B, patients receiving neoadjuvant therapy before surgery will undergo paired MRI assessment, and MRI-based tumor regression grade will be compared with pathological tumor regression grade in the resected primary tumor bed. Imaging will be independently assessed by two radiologists, with adjudication by a third reader. The study will evaluate diagnostic performance, reader agreement, image quality, protocol adherence, reporting completeness, and consistency across participating centers and MRI vendors. Lymph-node staging and nodal treatment response are outside the research scope.
Age 18 to 80 years, inclusive
Histologically confirmed esophageal squamous cell carcinoma based on endoscopic biopsy or other tissue examination
Primary tumor located in the thoracic esophagus
Planned curative esophagectomy
Adenocarcinoma, neuroendocrine carcinoma, adenosquamous carcinoma, or another non-squamous histological type
Cervical esophageal cancer, a gastric primary tumor extending into the esophagus, or multiple primary lesions that prevent reliable imaging-pathology matching
Definite distant metastasis, unresectable T4b disease, or another reason that prevents acquisition of a surgical pathological reference standard
Previous esophageal cancer surgery, radiotherapy, or another treatment that substantially altered local esophageal anatomy
Adults aged 18-80 years with histologically confirmed thoracic esophageal squamous cell carcinoma who are scheduled for curative esophagectomy at seven tertiary hospitals in China. Eligible patients will be enrolled consecutively. Cohort membership is determined by the clinical treatment pathway (upfront surgery or neoadjuvant therapy followed by surgery) selected by the treating multidisciplinary team independently of study participation. MRI examinations are performed as part of routine clinical care.
Cohort
Prospective
Standardized 3.0-T Esophageal Magnetic Resonance Imaging
Diagnostic testA standardized 3.0-T esophageal MRI framework comprising large-coverage and tumor-oriented T2-weighted imaging, high-resolution oblique axial T2-weighted imaging, diffusion-weighted imaging with apparent diffusion coefficient maps, and three-dimensional T1-weighted imaging before and after gadolinium contrast administration. Cohort A has one preoperative examination within 14 days before surgery. Cohort B has a baseline examination within 14 days before neoadjuvant treatment and a post-treatment examination within 14 days before surgery. Images are evaluated with predefined structured mrT or mrTRG criteria.
Exact agreement between MRI-based and pathological T stage in Cohort A
Proportion of evaluable Cohort A participants for whom the four-category structured MRI T stage (mrT1, mrT2, mrT3, or mrT4a) exactly matches postoperative pathological T stage (pT1, pT2, pT3, or pT4a). Linear weighted kappa and its 95% confidence interval will also be reported.
Area under the ROC curve of mrTRG for pathological good response in Cohort B
Area under the receiver operating characteristic curve and 95% confidence interval for four-level MRI tumor regression grade (mrTRG 1-4) to identify pathological good response, defined as modified Ryan pTRG 0-1 versus pTRG 2-3. The prespecified imaging dichotomy is mrTRG 1-2 versus mrTRG 3-4.
Agreement within one T-stage category in Cohort A
Proportion of evaluable participants for whom mrT and pT differ by no more than one ordered T-stage category.
MRI T-stage overstaging and understaging rates in Cohort A
Proportions of evaluable participants in whom mrT is higher than pT (overstaging) or lower than pT (understaging).
Sensitivity of mrT at prespecified pathological T-stage thresholds in Cohort A
Sensitivity (percentage) of structured MRI T staging for detecting pT2 or higher and pT3-pT4a disease. Results will be reported separately for each prespecified threshold using postoperative pathological T stage as the reference standard.
Interreader agreement for MRI T stage in Cohort A
Cohen kappa or weighted kappa calculated from the original independent mrT assessments of the two radiologists.
Difference in exact MRI-to-pathology T-stage agreement across prespecified subgroups in Cohort A
Percentage-point differences in exact mrT-pT agreement across center, MRI vendor, tumor location, image quality category, and reader-experience category. Each subgroup comparison will be reported separately; mixed-effects and leave-one-center-out sensitivity analyses will assess robustness.
Difference in AUROC for pathological good response across prespecified subgroups in Cohort B
Differences in area under the receiver operating characteristic curve for mrTRG identification of pathological good response (modified Ryan pTRG 0-1) across center, MRI vendor, tumor location, image quality category, reader-experience category, and neoadjuvant-treatment type. Each comparison will be reported separately; mixed-effects and leave-one-center-out sensitivity analyses will assess robustness.
Click on the lead sponsor to view all of their trials.
This trial is not recruiting at the moment. You can still explore other options:
