The primary goal of this observational study is to assess the predictive value of a model developed through a multiparametric analysis of clinical images, histopathological data, CT scans, and molecular analysis. This analysis will be conducted on a retrospective cohort recruited at CHU Brest (training group). The model will then be validated using external cohorts from other centers (test group).

Around 20% of patients with colorectal cancer present with synchronous liver metastases at the time of diagnosis. In recent years, there has been increasing interest in the high-throughput extraction of quantitative data from medical images, a technique known as radiomics. Radiomic analysis can involve both qualitative and quantitative imaging parameters, which may be used independently or integrated into multiparametric prediction models.

This is a multicenter, retrospective study designed to investigate the demographic, clinical, and radiological features, treatment regimens, and outcomes of patients diagnosed with colorectal cancer. The study will involve the collection of anonymized clinical data, including age, sex, diagnosis date, treatment details, results from histopathological and genetic analyses, as well as pre- and post-treatment medical imaging typically performed as part of the management of colorectal cancer patients, along with survival and tumor recurrence data.

PRIMARY EVALUATION CRITERIA Prediction of recurrence/treatment response based on the radiomic model.

SECONDARY EVALUATION CRITERIA The impact of CT scan image acquisition and reconstruction parameters on the prognostic model.

Inclusion Criteria:

Patients who have undergone surgery for colorectal cancer.

Exclusion Criteria:

• Patients under legal protection (guardianship, curatorship, etc.)
• Refusal to participate (expressed opposition)
• Individuals with conditions that could interfere with the study protocol or impact data integrity

The study population consists of patients diagnosed with colorectal cancer (CRC) who are undergoing various stages of treatment and care. The cohort includes individuals who have been surgically treated for colorectal cancer, with or without additional therapies such as chemotherapy, radiotherapy, or immunotherapy. The population may also include patients who have undergone routine follow-up care after treatment.

All collected data that underlie results in a publication

Data will be available beginning three years and ending fifteen years following the final study report completion

Data access requests will be reviewed by the internal committee of Brest UH. Requestors will be required to sign and complete a data access agreement