Quantitative imaging outperforms molecular markers when predicting response to chemoradiotherapy for rectal cancer

Area of Science:

• Oncology
• Radiology
• Molecular Imaging

Background:

• Rectal cancer treatment involves chemoradiotherapy (CRT).
• Predicting treatment response is crucial for optimizing patient outcomes.
• Current methods may not fully capture treatment response dynamics.

Purpose of the Study:

• To investigate the integration of imaging and molecular data for predicting response to neoadjuvant CRT in rectal cancer.
• To evaluate the combined predictive performance of ¹⁸F-FDG PET/CT, diffusion-weighted imaging (DWI), and T2-weighted MRI volumetry.

Main Methods:

• Eighty-five rectal cancer patients underwent preoperative CRT.
• ¹⁸F-FDG PET/CT and DWI were acquired at three time points: before CRT (TP1), during CRT (TP2), and before surgery (TP3).
• Multivariate models were developed combining imaging, cytokine, and gene expression data to predict ypT0-1N0 response.

Main Results:

• A ypT0-1N0 response was achieved by 26% of patients.
• Combined imaging parameters (¹⁸F-FDG PET/CT, DWI, T2-volumetry) yielded the highest predictive performance (AUC 0.81±0.03).
• Models using TP1 and TP3 imaging data achieved 75% sensitivity, 94% specificity, and 80% PPV for predicting response.

Conclusions:

• Combining ¹⁸F-FDG PET/CT, DWI, and T2-weighted MRI volumetry before CRT and before surgery aids in predicting treatment response.
• This multimodal imaging approach may assist in selecting rectal cancer patients for organ-preservation strategies.