Myosteatosis predicts survival after surgery for periampullary cancer: a novel method using MRI

David P J van Dijk, Frans C H Bakers, Sebastian Sanduleanu, Rianne D W Vaes, Sander S Rensen, Cornelis H C Dejong, Regina G H Beets-Tan, Steven W M Olde Damink, David P J van Dijk, Frans C H Bakers, Sebastian Sanduleanu, Rianne D W Vaes, Sander S Rensen, Cornelis H C Dejong, Regina G H Beets-Tan, Steven W M Olde Damink

Abstract

Background: Myosteatosis, characterized by inter- and intramyocellular fat deposition, is strongly related to poor overall survival after surgery for periampullary cancer. It is commonly assessed by calculating the muscle radiation attenuation on computed tomography (CT) scans. However, since magnetic resonance imaging (MRI) is replacing CT in routine diagnostic work-up, developing methods based on MRI is important. We developed a new method using MRI-muscle signal intensity to assess myosteatosis and compared it with CT-muscle radiation attenuation.

Methods: Patients were selected from a prospective cohort of 236 surgical patients with periampullary cancer. The MRI-muscle signal intensity and CT-muscle radiation attenuation were assessed at the level of the third lumbar vertebra and related to survival.

Results: Forty-seven patients were included in the study. Inter-observer variability for MRI assessment was low (R2 = 0.94). MRI-muscle signal intensity was associated with short survival: median survival 9.8 (95%-CI: 1.5-18.1) vs. 18.2 (95%-CI: 10.7-25.8) months for high vs. low intensity, respectively (p = 0.038). Similar results were found for CT-muscle radiation attenuation (low vs. high radiation attenuation: 10.8 (95%-CI: 8.5-13.1) vs. 15.9 (95%-CI: 10.2-21.7) months, respectively; p = 0.046). MRI-signal intensity correlated negatively with CT-radiation attenuation (r=-0.614, p < 0.001).

Conclusions: Myosteatosis may be adequately assessed using either MRI-muscle signal intensity or CT-muscle radiation attenuation.

Copyright © 2018 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.

Source: PubMed

3
Abonneren