Synopsis

Keywords: Liver, Liver

Non-alcoholic fatty liver disease (NAFLD) is often associated with abnormal metabolic syndrome. In this study, we investigated dynamic glucose enhanced imaging of the liver using T1rho MRI after glucose ingestion. We hypothesize this approach can be used to assess metabolic activities in the liver. Sixteen young volunteers and four patients with NAFLD were recruited in this study. The preliminary results suggest that the proposed approach has the potential to detect metabolic variations between normal subjects and subjects with fatty liver.

Introduction

Dynamic glucose-enhanced (DGE) imaging of the liver after the intake of glucose can be used to detect the metabolic changes in the liver. Such a technique has significant potential in the assessment of liver diseases. The ¹⁸F fluorodeoxyglucose (FDG)-PET imaging (1-2) and ¹³C-labeled glucose MRS (3) have been reported for DGE imaging to measure the glucose metabolism in the liver. The FDG-PET imaging results show potential in detecting NASH (1-2). However, these methods require extra hardware or specially labeled glucose which brings radiation risks to subjects. Chemical Exchange Saturation Transfer (CEST) and Chemical Exchange Spin-Lock (CESL) techniques have been developed for DGE imaging (4-5). However, the application of these methods in the liver has not been reported in previous studies. In this work, we report DGE imaging using a breath-hold black blood T1rho mapping technique (6) and show preliminary results on healthy young volunteers and patients with NAFLD.

Method

The workflow designed for this study is illustrated in Figure 1. All subjects were asked to fast overnight before the MRI exam. Two repeated baseline T1rho scan were performed at the beginning of the MRI exam. After that, subjects drank natural D-glucose (75g, 30% solution, DEXTROSOL, Germany) with a straw and without moving the scan position. Multiple T1rho scans were followed and lasted for about 50 minutes. A total of 20 subjects were recruited in this study, including four patients with NAFLD confirmed by biopsy (age: 53.3±10.6 years, BMI: 24.3±2.6 kg/m2, three male) and 16 young volunteers (age: 26.4±2.8 years, BMI: 22.5±4.3 kg/m2, eight male). Three patients (2 male and 1 female) were confirmed with NASH by liver biopsy.
This study was conducted under the approval of the institutional review board. All scans were conducted using a 3.0T MRI scanner (Philips Achieva TX, Philips Healthcare, Best, Netherlands). A 32-channel cardiac coil (Invivo Corp, Gainesville, USA) was used as the receiver and the body coil was used as the RF transmitter. The detail of black-blood T1rho imaging pulse sequence was described in our previous study (6). The T1rho imaging parameters include: resolution 2×2 mm, slice thickness 7 mm, time of spin-lock (TSL) 0, 30, 50 ms, frequency of spin-lock (FSL) 400 Hz, TE/TR 17/2000 ms. Spectral Attenuated Inversion Recovery (SPAIR) and Double inversion recovery (DIR) were applied to suppress the fat and blood signal, respectively. Two coronal slices were acquired in each T1rho scan. At the beginning of exam, an mDIXON Quant^TM (Philips Healthcare) scan was performed to measure liver fat fraction. We measured the difference between the peak R1rho and the baseline R1rho before glucose ingestion, denoted as ∆R1rho and use its mean from the ROIs within the right lobe of the liver for analysis. The statistical difference between two groups was tested by one-way ANOVA. A p-value < 0.05 is considered statistically significant.

Results

Liver R1rho value increases after glucose ingestion among most young volunteers. The increase of liver R1rho is significantly reduced among most patients and volunteers with fatty liver. Figure 2 shows examples of dynamic R1rho changes from a healthy volunteer without fatty liver (fat fraction < 5%) and a biopsy-confirmed NASH patient, respectively. Figure 3 compared ∆R1rho between the two groups with or without fatty liver, and the two groups with or without biopsy-confirmed NASH, respectively. The fatty liver is based on the fat fraction measured using mDixon Quant.

Discussion and Conclusion

In this preliminary study, the dynamic change of glucose-enhanced liver R1rho was different between healthy volunteers and subjects with fatty liver. We observed such a difference is more significant in the male subjects than the female subjects.
The preliminary result is consistent with the results previously reported using PET imaging (2). Our study shows dynamic glucose-enhanced liver R1rho has the potential to reflect the metabolism difference between healthy and subjects with fatty liver diseases.
In our protocol, the subjects drink natural glucose with the dosage same as that used in standard clinical glucose tolerance test (75g). Our studies show our approach is capable to detect the DGE signal with this dosage. In addition, we used blood suppression technique when measuring T1rho in the liver (6). This is expected to reduce the confounding effect from the glucose level changes in the blood when measuring metabolism of the liver.
One limitation of this study is small sample size. The influence of gender and age on R1rho dynamic changes requires further investigation. Further larger cohort studies are required to investigate the robustness and sensitivity of this method.

Acknowledgements

This study was supported by a grant from the Innovation and Technology Commission of the Hong Kong SAR (Project MRP/046/20X).