Keywords:
Metabolic disorders, Diagnostic procedure, MR, Cardiac
Authors:
C. M. Berzovini1, R. Faletti2, M. Gatti2, E. Caramia2, E. Vanni1; 1Torino/IT, 2Turin/IT
Methods and Materials
A prospective single-centre case-control study was performed in our Magnetic Resonance (MRI) Service in collaboration with University Departments of Gastroenterology and Cardiology.
Between October 2014 and February 2016,
33 patients with biopsy proven NAFLD (26 men,
average age 46 years) and 13 healthy controls (7 men,
average age 38 years) were enrolled.
To exclude the presence of diabetes,
metabolic syndrome and hypertension all patients underwent metabolic analysis.
We assessed cardiac function and epicardial fat with transthoracic echocardiography (TTE) and MRI.
MRI was performed,
ECG synchronized,
with Achieva 1.5 T MR System (Philips) and 32-channel coil.
For every patient we acquired Steady State Free Precession (SSFP) axial sequences of the chest and short axis to cover both ventricles evaluating 30 phases of cardiac cycle to study volumes and function.
Overall duration of the exam was averagely about 20 minutes.
14 out of 33 patients underwent also a basal abdominal MRI study with 3D Gradient Echo T1-weighted volumetric sequence for the quantification of visceral fat tissue.
We calculated: volume of epicardial and visceral fat,
left ventricle function parameters (cardiac output,
ejection fraction,
end-systolic and end-diastolic volume and systolic and diastolic speed of peak ejection).
Axial SSFP sequences with OsiriX software (Pixmeo) were used to quantify epicardial fat by manually delimiting a region of interest (ROI) on epicardial fat and using an interpolation and computation function of the overall volume provided by the software (Figure 1).
As for the epicardial fat,
we evaluated visceral fat defining ROIs on the internal side of abdominal muscular or bony structures and selecting fat tissue signal intensity threshold values so we could be able to exclude vascular or visceral structures (parenchymatous or not) from volumetric computing (Fig.
2).
With TTE,
we calculated cardiac function parameters and epicardial fat thickness in the subcostal and long axis projections.
Considering the limited sample size and the type of variables examined,
the statistical analysis used non-parametric tests,
such as the Mann-Whitney and the linear correlation coefficient Pearson test.