Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2025 Aug:51:9-17.
doi: 10.1016/j.clnu.2025.04.019. Epub 2025 May 13.

Automated three-dimensional body composition analysis identifies visceral adipose tissue radiodensity as a predictor of mortality and recurrence in colorectal cancer

Dinh Van Chi Mai  1 , Ioanna Drami  2 , Edward T Pring  3 , Laura Gould  4 , Jason Rai  3 , George Malietzis  5 , Phillip Lung  3 , Vincent Chow  6 , Karteek Popuri  7 , Mirza F Beg  6 , Thanos Athanasiou  8 , John T Jenkins  3 ; BiCyCLE Research Group  9
Affiliations

Automated three-dimensional body composition analysis identifies visceral adipose tissue radiodensity as a predictor of mortality and recurrence in colorectal cancer

Dinh Van Chi Mai et al. Clin Nutr. 2025 Aug.

Abstract

Background: Artificial intelligence enables automated three-dimensional (3D) volumetric body composition (BC) analysis from computed tomography (CT), opposed to single third lumbar vertebra (L3) slices alone. This study aimed to identify relationships between 3D BC and postoperative colorectal cancer (CRC) outcomes utilising Voronoi Health Analytics' Data Analysis Facilitation Suite software.

Method: Abdominopelvic skeletal muscle (SM), subcutaneous, and visceral adipose tissue volume and radiodensity were quantified from preoperative CT in 750 patients undergoing CRC resection. Non-contrast enhanced muscle, subcutaneous and visceral adipose tissue radiodensity (Hounsfield units) were adjusted using institution-specific correction values. Volumes were divided by abdominopelvic length, and (height)2, providing size-adjusted indices. Sex-specific cut-offs were defined for BC. Multivariate Cox regression assessed overall, and disease-free survival.

Results: Sarcopenia [HR 1.389 (1.071-1.800), p = 0.013], myosteatosis [HR 1.526 (1.163-2.002), p = 0.002], visceral obesity [HR 2.020 (1.458-2.798), p <0.001], subcutaneous adipose tissue quantity depletion [HR 1.499 (1.117-2.011), p = 0.007], and elevated visceral adipose tissue radiodensity [HR 1.728 (1.171-2.548) p = 0.006] predicted poorer overall survival. Sarcopenia [HR 1.513 (1.019-2.247), p = 0.040] and elevated visceral adipose tissue radiodensity [HR 1.763 (1.148-2.708), p = 0.010] predicted disease recurrence. Elevated visceral adipose was significantly associated with T4 tumour stage [RR 1.621 (1.211-2.489), p = 0.004]; and lymphovascular invasion [RR 1.472 (1.043-2.078), p = 0.030].

Conclusion: As established in studies using L3 CT-slices, sarcopenia and myosteatosis in the 3D abdominopelvic region also predict CRC survival outcomes. Whilst further validation is required, elevated abdominopelvic visceral adipose tissue radiodensity is a potentially novel predictor of both mortality and recurrence. This was linked to a more aggressive tumour phenotype in this study.

Keywords: Adipose tissue; Body composition; Colorectal cancer; Myosteatosis; Sarcopenia.

PubMed Disclaimer

Conflict of interest statement

Conflict of interest Mirza Beg, Karteek Popuri, and Vincent Chow are co-founders of Voronoi, a spinout company from Simon Fraser University (Vancouver, Canada) producing DAFS: https://www.voronoihealthanalytics.com/