Related Experiment Videos
Development of a multivariable prediction model for tumor response and treatment tolerance based on body composition
Soohyeon Lee1, Dong Hyun Kang1, Tae Sung Ahn1
1Department of Surgery, College of Medicine, Soonchunhyang University Cheonan Hospital, Cheonan, Republic of Korea.
Background:
The treatment paradigm for locally advanced rectal cancer (LARC) is evolving toward total neoadjuvant therapy (TNT), emphasizing strategies to maximize tumor regression and enable organ preservation. Identifying host-related biomarkers that predict response to neoadjuvant chemoradiotherapy (CRT) has therefore become increasingly important. Although body composition has been associated with cancer prognosis, the distinct roles of skeletal muscle and adipose tissue dynamics in treatment tolerance and tumor response remain unclear. This study aimed to develop a multivariable prediction model for tumor response and treatment tolerance based on body composition dynamics during CRT.
Methods:
In this single-center retrospective cohort study conducted at a tertiary referral hospital, 138 patients with LARC who underwent neoadjuvant CRT were analyzed. Baseline and post-CRT body composition parameters (skeletal muscle and adipose indices at the L3 level) and clinicolaboratory variables were evaluated as predictors. The primary outcome was high TRG (tumor regression grade 3-4), and the secondary outcome was CRT-related adverse events (CRT-AEs). A multivariable logistic regression model was developed, with continuous variables standardized. Model performance was assessed using the area under the receiver operating characteristic curve (AUC) and decision curve analysis. Internal validation was performed using 1,000 bootstrap resamples.
Results:
High TRG was observed in 33.3% of patients. Multivariable analysis demonstrated that the absence of CRT-AEs, higher baseline subcutaneous fat index, and a smaller reduction in visceral fat during CRT were independently associated with high TRG (all P<0.05). In contrast, a greater reduction in skeletal muscle index (ΔSMI) during CRT was significantly associated with CRT-AEs. The prediction model incorporating these variables showed moderate discriminative performance (AUC =0.707; 95% CI: 0.619-0.794).
Conclusions:
Body composition appears to play compartment-specific roles during neoadjuvant CRT in rectal cancer. Skeletal muscle status primarily reflects treatment tolerance, whereas adipose tissue dynamics are more closely associated with tumor regression. These computed tomography (CT)-based metrics may serve as practical, noninvasive indicators to support risk stratification and personalized treatment strategies in the era of TNT.
