Computed Tomography Scan and Clinical-based Complete Response Prediction in Locally Advanced Rectal Cancer after...

Area of Science:

• Oncology
• Radiology
• Medical Imaging

Background:

• Locally advanced rectal cancer (LARC) treatment involves neoadjuvant chemoradiotherapy (nCRT) followed by surgery.
• Assessing treatment response is crucial for patient follow-up.
• Radiomics analysis of pretreatment CT images can offer insights into treatment response.

Purpose of the Study:

• To investigate radiomics patterns from pretreatment CT images in rectal cancer.
• To explore the relationship between radiomics features and treatment response criteria.
• To evaluate the predictive capability of radiomics for treatment response in LARC.

Main Methods:

• Fifty patients with rectal adenocarcinoma undergoing nCRT and surgery were included.
• Radiomics features were extracted from pretreatment CT images.
• Random forests modeling was used to analyze combined clinical and radiomics data.

Main Results:

• The study included 50 patients (46% male, 54% female).
• No significant difference in treatment response was observed based on age or gender.
• Combined clinical and radiomics data achieved high predictive accuracy (AUC 0.97-0.99) for responders and non-responders.

Conclusions:

• Random forests modeling using combined radiomics and clinical data can predict neoadjuvant treatment response in LARC.
• This approach offers an acceptable level of accuracy in identifying responders and non-responders.
• Radiomics holds potential as a non-invasive tool for treatment response assessment in rectal cancer.