Thymic Carcinoma, Thymic Epithelial Tumor, Thymoma, Thymoma and Thymic Carcinoma
Conditions
Keywords
Artificial Intelligence, Digital pathology
Brief summary
Thymic epithelial tumors are rare neoplasms in the anterior mediastinum. The cornerstone of the treatment is surgical resection. Administration of postoperative radiotherapy is usually indicated in patients with more extensive local disease, incomplete resection and/or more aggressive subtypes, defined by the WHO histopathological classification. In this classification thymoma types A, AB, B1, B2, B3, and thymic carcinoma are distinguished. Studies have shown large discordances between pathologists in subtyping these tumors. Moreover, the WHO classification alone does not accurately predict the risk of recurrence, as within subtypes patients have divergent prognoses. The investigators will develop AI models using digital pathology and relevant clinical variables to improve the accuracy of histopathological classification of thymic epithelial tumors, and to better predict the risk of recurrence. In this multicentric and international project three existing databases will be used from Rotterdam, Maastricht and Lyon. For all models one database will be used to build AI models, and the other two for external validation. The ultimate goal of this project is to develop AI models that support the pathologist in correctly subtyping thymic epithelial tumors, in order to prevent patients from under- or overtreatment with adjuvant radiotherapy.
Interventions
DIAGNOSTIC_TESTArtificial Intelligence Diagnostics
AI Diagnostics uses advanced algorithms for precise histological image analysis to help diagnose disease, including subtype.
DIAGNOSTIC_TESTRecurrence Prediction Tool
This AI tool evaluates thymic tumour data and other clinical data and calculates the risk of recurrence, with the aim of analysing whether there is an association with specific subtypes of thymic epithelial tumours and clinical data.
Sponsors
Maastro Clinic, The Netherlands
Hospices Civils de Lyon
Erasmus Medical Center
Study design
Observational model
COHORT
Time perspective
PROSPECTIVE
Eligibility
Sex/Gender
ALL
Inclusion criteria
Participants with specific diagnoses are eligible for inclusion in the study. The eligible diagnoses include various subtypes of thymoma and thymic carcinoma, specifically: * Thymoma A * Thymoma AB * Thymoma B1 * Thymoma B2 * Thymoma B3 * Thymic Carcinoma Inclusion is based on a consensus diagnosis with a level of agreement less than 70%. This criterion is applied during the training phase of the model. Recurrence Criteria: Participants with a documented recurrence outcome within a 5-year period are considered eligible for this aspect of the study. This criterion is primarily applied during the validation phase.
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| WP1 - Databases/Data Pre-processing | M1-M18 | The EMC-dataset includes 179 TET-patients classified by experienced TET-pathologists. Cases with good agreement between pathologists will be used for training AI-models. Evaluation includes digitized pathology slides assessed by an international expert-panel. The MUMC-database (137 patients) and CHUL-database (181 patients) provide additional data, including clinical variables. Relevant factors include age, gender, tumor volume, stage, completeness of resection, autoimmune disorders, and treatment details. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| WP2 - Deep Learning-Model for TET Classification and Recurrence Prediction | M6-M32 | This outcome aims to create an AI-framework with two principal goals. First, investigate TET-subtypes using four different models emphasizing cell type, morphological structures, and a combination. Second, classify patients based on recurrence outcome within 5 years. An ablation study will be conducted with state-of-the-art deep learning classifiers (ResNet, Inception). |
Other
| Measure | Time frame | Description |
|---|---|---|
| WP3: Clinical Evaluation | M6-M36 | AI-models 1-3 will be built and validated on the EMC-database, while AI-model 4 will be built on the MUMC+-database and validated on both. Model performance will be assessed using sensitivity, specificity, negative/positive predictive value. Decision analysis curves will quantify the clinical benefit, identifying patient groups with the largest utility. |
Countries
Netherlands
Contacts
Primary ContactAnna Salut Esteve Domínguez
Outcome results
None listed