A clinico-radiological model integrating baseline CT-derived body composition to predict prognosis in immunotherapy-treated metastatic colorectal cancer.

This study aims to investigate the association between baseline CT-derived body composition parameters and progression-free survival (PFS) in patients with metastatic colorectal cancer (mCRC) receiving immunotherapy, and to develop a personalized clinico-radiological prognostic model. We retrospectively enrolled 107 mCRC patients treated with immune checkpoint inhibitors. Baseline abdominal CT images at the third lumbar vertebra were segmented to quantify visceral fat area (VFA), subcutaneous fat area (SFA), skeletal muscle area (SMA), and skeletal muscle radiodensity (SMD). Independent prognostic factors for PFS were identified via univariable and multivariable Cox regression analyses and further validated using the Least Absolute Shrinkage and Selection Operator (LASSO).

Exploratory subgroup interaction analyses were conducted to evaluate prognostic consistency. Finally, a predictive nomogram was developed and performance was assessed via C-index, time-dependent receiver operating characteristic (ROC) curves, calibration plots with 1000 bootstrap resamples, and decision curve analysis (DCA). Multivariable and LASSO Cox regression analyses identified age, number of metastatic organs, baseline CA 19-9 level, VFA, and SMD as potential independent prognostic factors for PFS. Higher VFA and SMD were significantly associated with prolonged PFS.

Notably, subgroup analyses revealed that the protective effect of high VFA was more evident in patients with baseline liver metastasis (P for interaction