Comparison of volumetric and single-slice computed tomography body composition metrics for colorectal cancer survival

Abstract

Background: Body composition is associated with colorectal cancer (CRC) survival. However, body composition measurements have traditionally relied on single-slice, axial imaging. Fully automated volumetric body composition analysis is widely available, but associations with CRC survival have yet to be examined in detail. Methods: Among a nested case-control sample of CRC patients with existing CT scans, volumetric and single-slice body composition analysis was performed, including total area and proportional skeletal muscle (SM), visceral adipose tissue (VAT), subcutaneous adipose tissue (SAT), and intermuscular adipose tissue (IMAT). Body composition was measured from the T12 vertebra to the sacrum, with the mid-L3 level used for single-slice analysis. We used multivariable Cox regression models to estimate associations between height-indexed volumetric and single-slice body composition metrics with all-cause mortality. Results: The mean age of the 121 enrolled patients was 61, and 38 (31 %) died over a mean follow-up of 7.7 years. The T12–sacrum, T12–L3, and L3–sacrum volumetric measurements were correlated with each other and their corresponding mid-L3 metric (all ρ > 0.8). In adjusted models, the T12–sacrum VAT index proportion yielded the strongest association with CRC survival (per 1-SD increase, HR: 2.07, 95 % CI: 1.13, 3.80, P = 0.02). Mid-L3 and volumetric composition metrics showed similar associations with CRC survival. Conclusions: Volumetric body composition metrics are associated with CRC survival but did not outperform single-slice metrics in predicting CRC survival. Proportional metrics, which account for total abdominal muscle and adipose tissue area, may be a novel computational technique for assessing body composition.