Prediction of Chemotherapy Toxicity After Four Cycles of R‐ CHOP Treatment for Diffuse Large B‐Cell Lymphoma: Effective Imaging Biomarkers of Body Composition

ABSTRACT Background Previous studies have indicated that diffuse large B‐cell lymphoma (DLBCL) patients with a low baseline skeletal muscle (SM) area are more susceptible to severe toxicity during chemotherapy. However, the predictive role of baseline body composition in determining toxicity risk during the initial frontline treatment remains unexplored. This study aims to identify reliable, non‐invasive biomarkers and validate findings using follow‐up CT scans after four cycles of chemotherapy. Methods We retrospectively included DLBCL patients who received four cycles of R‐CHOP treatment between January 2015 and January 2024, with pre‐treatment abdominal CT scans. We measured the volume, area, and density of SM, visceral adipose tissue (VAT), and subcutaneous adipose tissue (SAT) to assess their predictive potential for chemotherapy toxicity. Subgroup analyses examined longitudinal changes in body composition, and a logistic regression model identified effective imaging biomarkers associated with grade 3/4 toxicity. Results Among the 179 DLBCL patients (mean age 56.96 ± 13.49 years), 46.9% experienced grade 3/4 toxicity. Lower baseline SM volume and density significantly increased the risk of toxicity, particularly in overweight or obese patients ( p < 0.05). ROC analysis identified SM volume as the best predictor, with a cutoff of 2093.71 cm 3 ; patients below this threshold had a 3.34 times higher risk ( p = 0.001). A decrease in SM volume was associated with higher risks of hematological toxicity ( p = 0.022) and neutropenic fever ( p = 0.021). Conclusion Lower baseline SM volume and its reductions during treatment are associated with an increased risk of grade 3/4 toxicity, particularly in overweight or obese patients. Body composition measurements serve as effective imaging biomarkers.