Ultrasound Liver Imaging for Classification of Metabolic Dysfunction-associated Steatotic Liver Disease

Overview

Tissue elasticity and viscosity correlate with pathology. These tissue properties are typically evaluated subjectively using palpation. The purpose of "elastography" is to provide an objective elasticity image that is equivalent to the remote palpation of tissue. The investigators have developed elastography imaging systems based on ultrasound and magnetic resonance imaging and have applied them previously to prostate imaging, breast imaging in patients and liver imaging in healthy volunteers. A first objective of this study is to compare the investigators' ultrasound shear wave absolute vibro-elastography (S-WAVE) technology with the existing clinical standard, FibroScan, and magnetic resonance elastography to quantify liver stiffness in healthy volunteers and in patients suspected of fatty liver disease. A second objective of this study is to compare ultrasound-based liver tissue fat measurement with MRI-based measurements. A third objective of this study is to determine whether ultrasound can be used to assess liver inflammation.

1. Purpose The main purpose of this study is to develop ultrasound methods that can identify the stages of metabolic dysfunction-associated steatotic liver disease (MASLD) with Magnetic Resonance Elastography (MRE), MRI Proton Density Fat Fraction (PDFF) and biopsy as the ground truth. 2. Hypothesis Ultrasound can replace MRI in staging steatosis and fibrosis. Ultrasound can improve the staging of inflammation. 3. Justification A wide range of pathologies can cause changes in the mechanical properties of human tissue. For this reason, finding a way to non-invasively evaluate tissue stiffness is proving clinically useful. Many liver diseases of varying etiologies including hepatitis, fatty liver disease, alcoholism, and others, present with increased liver stiffness from fibrosis. Fibrosis can lead to various cancers of the liver or cirrhosis, sometimes causing death. It has been estimated that the prevalence of cirrhosis is 4.5% to 9.5% of the general population, indicating a much larger group is inflicted with the varying stages of fibrosis. If dramatic increases in liver stiffness from fibrosis can be caught at an early stage, treatment can be tailored for the patient, and results can be monitored to increase the likelihood of reversal and slow damaging effects. Biopsy is traditionally the standard method to diagnose and monitor liver conditions, being widely used since 1958. Some advantages of biopsy include the visualization of fat and iron deposits. Unfortunately, biopsies are not able to show heterogeneity of fibrosis as the reading is often taken from the same location, and consists of only 1/50,000th of the liver volume. Studies have shown discordance in biopsy results from the left to the right lobes in 33% of cases, and 30% misclassification of cirrhotic cases. As well, studies have shown significant inter-observer and intra-observer variation. Finally, a drawback of biopsy that has spurred the transition to non-invasive diagnostic measures, such as elastography, is the associated risk of pain, bleeding, and mortality. Various elastography methods have been developed and applied to the liver. Transient elastography, FibroScan (EchoSens, Paris) is the most common type of elastography currently applied to the liver for the measurement of stiffness, indicating fibrotic activity. It uses ultrasound technology with a unidimensional probe to measure the velocity of shear waves. It has become the standard of care in some clinics worldwide as an affordable device that is diagnostically reliable. Although it is pain-free, quick, and convenient, it is only able to collect a very small sample from the right lobe of the liver so cannot capture any heterogeneity and may misrepresent the state of the whole organ. As well, to the best of the investigators' knowledge, stiffness values obtained from FibroScan readings have not been correlated with Magnetic Resonance Elastography values in any direct comparison in the literature, and they result in less reliable results for earlier stages of fibrosis. Magnetic Resonance Elastography is another type of elastography applied to the liver. It has shown very high diagnostic accuracy in the staging of patients into fibrotic stages, and is used in some private clinics, primarily in the United States. It allows flexibility in its application and in the protocols used to acquire the data. It has shown great repeatability and reproducibility, and is often effective in patients with higher BMIs. The high costs and complexity associated with MRI scans restricts its usefulness in many areas affected by liver disease. The investigators' S-WAVE method is an ultrasound method that will produce two and three-dimensional images, allowing for full visualization of the liver. It is an adjustable multi-frequency technique, which can show large areas of the organ as volumetric elasticity maps. It is affordable and feasible in widespread clinical settings due to the simple hardware needs of an ultrasound device. The ultrasound data that is used to measure tissue stiffness can also be used to estimate tissue fat content. Indeed, fatty tissue tends to attenuate ultrasound more than normal tissue. Other characteristics of the ultrasound data (the radio frequency or RF data) are also changed when tissue becomes fatty. The first stage of MASLD is steatosis, an increase proportion of fat in liver tissue, followed by inflammation, followed by tissue scarring and fibrosis. 4. Objectives * To compare the stiffness results obtained from MRE, S-WAVE, and FibroScan in healthy volunteers and in patients to demonstrate strong agreement between S-WAVE with MRE in assessing fibrosis. * To compare the MRI proton density fat fraction with ultrasound attenuation and other measures, to demonstrate strong agreement between ultrasound and MRI-PDFF in assessing steatosis. * To show that ultrasound can assess the entire MASLD spectrum, including steatosis, fibrosis and inflammation, with biopsy as ground truth. 5. Research Methods 40 subjects with no history of liver disease, mainly to be recruited from a population of graduate students at UBC, and 105 subjects, with various stages of fibrosis, to be recruited at the Gastroenterology Institute of Research Institute (GIRI) in Vancouver, and VGH will undergo a FibroScan, S-WAVE, and MRI scans including PDFF and MRE. The FibroScan and theS-WAVE will be performed by a technician and by a sonographer, respectively, at the GIRI or VGH. The MR examination will be performed at the UBC MRI research centre. Ten Fibroscan readings will be obtained from the right lobe through the patient's intercostal space. FibroScan displays the median stiffness in kPa as well as the interquartile range. A minimum success rate of 80% with an interquartile range of less than 30% stiffness value is required to accept FibroScan readings. Each subject will undergo the process once. S-WAVE is an elastography technique developed at UBC that uses the tissue response to multi-frequency vibrations to estimate the tissue's mechanical properties. The vibrations generate steady-state shear waves in tissue. A series of ultrasound images capture the tissue shear wave motion and a tissue model is used to estimate the underlying mechanical properties. It has already been successfully applied to image the prostate, the breast, the placenta and in the liver. The excitation device used is a board placed underneath the patient on the examination bed. The frequencies applied will range from 45 Hz to 80 Hz. The amplitude selected will be sufficient to produce adequate waves in the region of interest, and within the patient's comfort level. A 3D ultrasound probe will be used to obtain image volumes, while being held in place by the sonographer. In addition, a 2D ultrasound transducer will be swept to obtain image volumes in approximately the same location. Inter-costal and trans-abdominal images will be obtained according to a repeatable protocol. All S-WAVE scans will be performed by two trained and very experienced sonographers, Ms. Vickie Lessoway and Ms. Jan Reid. Magnetic Resonance Elastography will be performed on the 3T Philips Elition MR scanner located at the UBC MRI research centre. The scan will begin with axial T1 and T2 images, followed by Diffusion Weighted imaging. Then MR elastography examination using the eXpresso sequence and concurrent excitation of tissue will be performed next. The eXpresso sequence has been used in phantom, liver, and prostate studies. The excitation system is an electromagnetic actuator that has been built in house and comprises of loops of wire arranged into a coil. Vibrations result when current is passed through the wires in the presence of the magnetic field of the MR scanner. It is synchronized with the MRI imaging sequence. The shaker will be placed over the right rib cage of the subject whilst lying in a supine position. It will be strapped in place with a velcro belt to avoid any movement. The diameter of the shaker plate is 5 to 10 cm wide. The system is very similar to that used in previous studies elsewhere and to image the prostate and the liver. The full MR imaging session will last less than 90 minutes. Processing of MRE and VE data occurs offline to extract mechanical property and fat content information. These can then be displayed visually and compared. Both involve processing of the raw ultrasound echo data. 6. Data Processing All S-WAVE data will be processed by software developed in the Robotics and Control Laboratory at UBC. This software is able to estimate a stiffness value at all imaged points based on a tissue model. The output is a volumetric map of tissue stiffness that can be superimposed on the corresponding B-mode image. Regions of Interest will be selected based on homogenous areas with few blood vessels. The raw MRE images will be processed with an in-house developed software which filters the data and applies models to calculate the dynamic shear modulus (Gd) and viscosity (Gl) for each pixel. These mechanical properties form a new image that are then superimposed on the standard T2-weighted images for easy visualization. Regions of interest with few blood vessels will be chosen. The average and standard deviation of Gd and Gl will be calculated. Tissue fat content will be determined by standard MRI PDFF sequences. Tissue fat content will be determined from ultrasound based on quantitative analysis of the raw ultrasound RF data. Ultrasound image processing will be optimized to match the MRI images (MRI as ground truth for fibrosis and steatosis assessment). Ultrasound image processing will be optimized to predict inflammation, with biopsy results as ground truth.