Objective
To investigate the role of preoperative body composition analysis for muscle and adipose tissue distribution on long-term oncological outcomes in patients with middle and low rectal cancer (RC) who received curative intent surgery.
Methods
A total of 155 patients with middle and low rectal cancer who underwent curative intent surgery between January 2014 and December 2016 were included for the final analysis. Skeletal muscle area (SMA), skeletal muscle radiodensity (SMD), visceral fat area (VFA) and mesorectal fat area (MFA) were retrospectively measured using preoperative CT images. To standardize the area according to patient stature, SMA was divided by the square of the height (m2) and the skeletal muscle mass index (SMI, cm2/m2) was obtained. Each median values of the distribution in male and female served as cut-off point for SMI, SMD, VFA, and MFA, respectively. Univariate and multivariate analysis were performed to evaluate the association between body composition and long-term oncological outcomes. Overall survival (OS) measured in months from the day of primary surgery until death for any cause. Disease-free survival (DFS) was defined as the interval between surgery and tumor recurrence. The Kaplan-Meier method with log-rank testing was used to validate prognostic biomarkers. Intraclass correlation coefficient (ICC) was used to evaluate interobserver and intraobserver reproducibility for SMA, SMD, MFA,VFA.
Results
During the follow-up period, 42 (27.1%) patients had tumor recurrence; 21 (13.5%) patients died. The sex-specific median value of SMI was 28.6 cm2/m2 for females and 48.2 cm2/m2 for males. The sex-specific median value of SMD was 34.7 HU for females and 37.4 HU for males. The sex-specific median value of VFA was 123.1 cm2 for females and 123.2 cm2 for males. The sex-specific median value of MFA was 13.8 cm2 for females and 16.0 cm2 for males. In the Cox regression multivariate analysis, SMI (P = 0.036), SMD (P = 0.022), and postoperative complications grades (P = 0.042) were significantly different between death group and non-death group; SMD (P = 0.011) and MFA (P = 0.022) were significantly different between recurrence group and non-recurrence group. VFA did not show any significant differences. By the Kaplan-Meier method with log-rank testing, DFS was significantly longer in patients with high-MFA (P = 0.028) and shorter in patients with low-SMD (P = 0.010), OS was significantly shorter in patients with low-SMI (P = 0.034) and low-SMD (P = 0.029).
Conclusions
Quantitative evaluation of skeletal muscle mass and adipose tissue distributions at initial diagnosis were important predictors for long-term oncologic outcomes in RC patients. SMD and SMI were independent factors for predicting OS in patients with middle and low rectal cancer who had radical surgery. SMD and MFA were independent factors for predicting DFS in patients with middle and low rectal cancer who had radical surgery.