Dynamics of body composition in the treatment of locally advanced rectal cancer: the ımpact of adipose and muscle tissue on pathological response

Abstract

OBJECTIVE: The aim of the study was to evaluate the role of body composition parameters measured via computed tomography before and after neoadjuvant chemoradiotherapy in predicting pathological response in patients with locally advanced rectal cancer. METHODS: Eighty-four patients with locally advanced rectal cancer who underwent neoadjuvant chemoradiotherapy followed by curative surgery were retrospectively analyzed. Computed tomography images obtained before and after neoadjuvant chemoradiotherapy were used to assess total adipose area, visceral adipose tissue, subcutaneous adipose tissue, adipose tissue density, psoas muscle area, psoas muscle density, and mesorectal adipose volume and density at the level of the L3 vertebra. Pathological response was determined using the Ryan Tumor Regression Grade system. Baseline values and percentage changes (Delta) during treatment were compared between response groups. RESULTS: Of the 84 patients, 31 (36.9%) showed a pathological response. There were no significant differences in demographic or clinical features between the groups. Among baseline body composition metrics, only psoas muscle density was significantly higher in the response group (p=0.015). Post-treatment analysis revealed a significant difference only in the percentage change in psoas muscle density (Delta PMD) (p=0.023). No other baseline or percentage change values for total adipose area, subcutaneous adipose tissue, visceral adipose tissue, mesorectal adipose volume, or psoas muscle area were significantly associated with pathological response (p>0.05). CONCLUSION: In patients with locally advanced rectal cancer, neither abdominal adipose composition nor mesorectal adipose volume nor psoas muscle area was predictive of response to neoadjuvant chemoradiotherapy. However, psoas muscle density, a potential indicator of muscle quality, may serve as a promising biomarker for predicting pathological response based on both baseline value and treatment-induced changes.