Application of Ultrasound and Shear Wave Elastography Imaging in a Rat Model of NAFLD/NASH

Summary

This protocol describes the use of an enhanced ultrasound technique to non-invasively observe and quantify liver tissue changes in rodent models of nonalcoholic fatty liver disease.

Abstract

Nonalcoholic Steatohepatitis (NASH) is a condition within the spectrum of Non-Alcoholic Fatty Liver Disease (NAFLD), which is characterized by liver fat accumulation (steatosis) and inflammation leading to fibrosis. Preclinical models closely recapitulating human NASH/NAFLD are essential in drug development. While liver biopsy is currently the gold standard for measuring NAFLD/NASH progression and diagnosis in the clinic, in the preclinical space, either collection of whole liver samples at multiple timepoints during a study or biopsy of liver is needed for histological analysis to assess the disease stage.

Conducting a liver biopsy mid-study is an invasive and labor-intensive procedure, and collecting liver samples to assess disease level increases the number of research animals needed for a study. Thus, there is a need for a reliable, translatable, non-invasive imaging biomarker to detect NASH/NAFLD in these preclinical models. Non-invasive ultrasound-based B-mode images and Shear Wave Elastography (SWE) can be used to measure steatosis as well as liver fibrosis. To assess the utility of SWE in preclinical rodent models of NASH, animals were placed on a pro-NASH diet and underwent non-invasive ultrasound B-mode and shear wave elastography imaging to measure hepatorenal (HR) index and liver elasticity, measuring progression of both liver fat accumulation and tissue stiffness, respectively, at multiple time points over the course of a given NAFLD/NASH study.

The HR index and elasticity numbers were compared to histological markers of steatosis and fibrosis. The results showed strong correlation between the HR index and percentage of Oil Red O (ORO) staining, as well as between elasticity and Picro-Sirius Red (PSR) staining of livers. The strong correlation between classic ex vivo methods and in vivo imaging results provides evidence that shear wave elastography/ultrasound-based imaging can be used to assess disease phenotype and progression in a preclinical model of NAFLD/NASH.

Introduction

Non-alcoholic fatty liver disease (NAFLD) is a metabolic condition characterized by an excessive buildup of fat in the liver and is quickly becoming a leading liver ailment worldwide with a recently reported global prevalence of 25%¹. Non-alcoholic steatohepatitis (NASH) is a more progressed stage of the spectrum of NAFLD, characterized by excess liver fat with progressive cellular damage, inflammation, and fibrosis. These ailments are often silent, undetected via blood tests or routine examinations, until considerable damage has already occurred to a patient's liver. Currently, the gold standard to diagnose NASH in patients is through histological examination of patient-derived liver biopsy samples. Similarly, preclinical researchers who work to understand the pathogenesis of NASH/NAFLD as well as the drug development industry rely on in vivo wedge biopsy of liver samples or terminal euthanasia of satellite cohorts for histology to measure steatosis, inflammation, and fibrosis.

For example, liver wedge biopsy has been a standard technique for assessing steatohepatitis and fibrosis while using the GUBRA NASH model². The liver wedge biopsy method is invasive and laborious in small animals³. The use of wedge liver biopsy in the middle of a study represents an added experimental variable in a disease model, which often increases the number of animals that are needed. With these factors in mind, non-invasive imaging techniques that can be used to reliably assess steatosis and fibrosis in NASH/NAFLD animal models at early time points prove valuable. Shear wave elastography (SWE) is an ultrasound-based method used to measure the elasticity of soft tissues. The technique measures the propagation of shear waves created by supersonic ultrasound pulses directed at a tissue target, and then calculates a value called E modulus⁴. The velocity of the shear wave is proportional to the degree of tissue stiffness.

Figure 1 and Figure 2 show the imaging area setup and the SWE instrument. The SWE instrument is a single, wheeled unit with two screens and a control panel shown in Figure 2A. The upper monitor (Figure 2B) acts as the computer monitor and displays images and patient directories. The control panel (Figure 2C) is an array of buttons and dials that control general aspects of image capture: freezing screen, saving images, changing from one mode to another. The lower screen (Figure 2D) is a touch screen with additional controls to change settings and acts as a keyboard to input data as needed. The instrument is equipped with a stylus to use on the touch screen if desired. Ultrasound probes attach to the lower front panel of the device. For B-mode and SWE imaging in rodents, the super-linear 6 to 20 MHz transducer was used. This ability to noninvasively measure tissue stiffness makes SWE a valuable tool for the identification and staging of liver fibrosis⁵ in NASH patients, decreasing the need for more invasive methods. SWE has, in fact, been used to measure liver fibrosis in patients and is an FDA-approved method to score fibrosis in the clinic⁶. Using SWE to monitor NASH progression in animal models of the disease would provide a translational tool for the development of treatments and simultaneously improve animal welfare through the reduction of animal subject numbers and refinement of in vivo procedures to minimize pain and distress.

SWE imaging in human patients uses a low-frequency ultrasound transducer⁴, which is not ideal for small animals. Notably, high-frequency SWE techniques have been used to evaluate the efficacy of acetyl-CoA carboxylase inhibition on pathogenesis of NASH in a rat model⁷, and the utility of this technique has been described in carbon tetrachloride rat models of liver fibrosis with successful results when compared to traditional METAVIR histological scoring methods⁸. However, existing literature lacks detailed technique and methodology information on the application of SWE imaging in preclinical models of NASH. As described above, liver steatosis is one of the key features of the NAFLD/NASH condition and is an important stage where intervention can be considered. Thus, assessing liver fat accumulation using an imaging modality is as important as assessing liver fibrosis in preclinical models of NASH/NAFLD.

An ultrasound technique known as the HR index, a ratio of tissue brightness of the liver compared to that of the renal cortex, has been used as a surrogate marker of steatosis in the clinic^9,10. This approach, however, has not been extensively used in preclinical animal models of NAFLD/NASH. This article describes a method of measuring elasticity as well as the HR index as a surrogate marker of hepatic fibrosis and steatosis, respectively, in a choline-deficient, high-fat diet (CDAHFD) rat model of NAFLD/NASH. This model induces rapid steatosis, liver inflammation, and fibrosis, which is measurable within 6 weeks in mice¹¹. The addition of cholesterol (1%) to this diet has been shown to promote fibrogenesis in rats¹², making this model a suitable candidate for validation studies involving shear wave imaging. Overall, this imaging technology can also be applied to a wide range of NASH models/diets where steatosis and/or fibrosis is an endpoint of interest.

Protocol

All animal-involved procedures were reviewed and approved by Pfizer's Institutional Animal Care and Use Committee (IACUC) and conducted in an AAALAC (Assessment and Accreditation of Laboratory Animal Care) International accredited facility.

1. Disease induction

1. Use male Wister Han rats (150-175 g; ~ 6-7 weeks old; total 40 rats) that are free of known rat adventitial pathogens. House the rats in pairs in individually ventilated caging with paper bedding (see the Table of Materials) and maintain them at 22 ± 1 °C, 40-70% relative humidity with a 12:12 h light-dark cycle.
2. Place the rats weighing 150-175 g (~6-7 weeks old) on a choline-deficient, high-fat diet with 1% cholesterol (n = 20) or a standard lab rodent chow (n = 20) depending on the study design.
   ​NOTE: In this study, a total of 40 rats were enrolled with 20 animals per group. At the end of the 6^th week, half of the cohort from each group was necropsied for mid-study histological analysis of liver samples. Thus, the sample size was 10 animals per group for 9^th and 12^th week time points.

2. Instrument setup

1. Set up the imaging area as follows: include a warmed surface to keep the animal warm during imaging (c in Figure 1), and a secured anesthesia nose cone to deliver inhalant anesthesia to maintain a plane of anesthesia throughout the procedure (b in Figure 1).
2. Use an ultrasound probe holder to facilitate moving the ultrasound probe to the desired location and to prevent the probe from resting on the animal.
   1. Use warmed ultrasound gel on the skin where the ultrasound image is acquired.
   2. Maintain the following settings throughout the procedure, which can be adjusted on the touch screen: Acoustic Power 0.0 dB; Tissue Tuner 1540 m/s; Dynamic Range 60 dB; Elasticity range (for SWE mode) < 30 kPa.
3. Attach the ultrasound probe to the rail system in the specialized holder (a in Figure 1).
4. Switch the instrument on and allow it to boot up. Once the monitor is turned on, note the B Mode image with connected transducer details.

3. Subject preparation

1. Make sure the animals are fasted at least 4 h prior to the imaging procedure to prevent gut content from interfering with image acquisition.
   1. After at least 4 h of fasting, place a rat in an isoflurane anesthetic induction chamber until a suitable level of anesthesia is reached, confirmed by no response to toe pinch. Expose the animals to 3-5% isoflurane for 3-5 min to induce anesthesia.
   2. For maintenance anesthesia, keep the animals under 2-3% isoflurane during image acquisition. Apply ophthalmic ointment to protect the eye from drying during anesthesia.
2. Once anesthesia has been achieved, remove an animal from the induction chamber and place it on a warm hot water circulating blanket. Place an anesthetic nose cone over the snout, and shave the animal on its right side, from the ribcage to pelvis. Use chemical depilation cream to remove all the remaining hair in this area.
3. Once hair has been removed, place an animal in left lateral recumbency with upper paws taped above the head on a warm imaging platform (Figure 3A).
4. Press the Patient key on the instrument control panel, and identify the subject according to the study design.
   1. Open the Keyboard function on the instrument by tapping the icon on the touch screen. Type the names as desired.
   2. Tap Exit to exit the patient name screen. Observe that B-mode reopens on the monitor.

4. Image acquisition for hepato-renal (HR) index measurement

1. Apply a small amount of warmed ultrasound gel to the depilated skin region on the animal.
2. Move the ultrasound probe to touch the gel-covered area of the subject (Figure 3B). Once a live B-mode image of the subject's internal organs appears on the monitor, move the ultrasound probe to the area slightly above the hip, just parallel to the lumbar vertebrae (sagittal plane).
3. Using the B-mode display on the monitor, locate the right kidney by identifying the large renal artery and cortex/medulla separation (Figure 4A). In addition, observe part of the liver in a single plane of the image.
   1. Ensure that there are little to no image artifacts such as shadows and air bubbles.
4. Measure a B Mode ratio to obtain the HR index.
   1. Ensure that both renal cortex and liver parenchyma are in the same plane of focus. If needed, adjust focus and gain control to obtain a clear image.
      1. Adjust the focus by turning the Focus knob on the control panel. Adjust the gain by pressing the Auto TGC button once.
   2. Press the Freeze key on the control panel. Ensure that the animal is between breaths when freezing the screen to avoid blurry images.
   3. Once the screen is frozen, tap Measurement Tools on the touch screen. Select B-mode Ratio, a built-in tool that measures relative brightness of a tissue from a selected region of interest. Create a 2 mm circle to select a region of interest (ROI). Adjust the circle size by moving a finger along the outer edge of the trackball on the control panel.
   4. Place the 2 mm circle on the liver image ROI, which should be located to the right of the kidney. Identify liver tissue based on its homogeneous echogenicity and smooth contour.
   5. Once the circle is in place, press the Select button on the control panel, and observe the new circle that appears.
   6. Adjust the size of the new circle to 2 mm, and place it on the image of the kidney cortex. Be sure to keep the depth of the circles on the liver and kidney cortex the same. Once in place, press the Select button on the control panel. Observe the built-in system tool displays the HR index as a B-mode ratio.
   7. Press Save Image to save the image, and observe the saved images that appear as thumbnails on the right side of the monitor.
   8. Press the Freeze button on the control panel to unfreeze the image and return to a live B mode image.
5. Repeat the B Mode ratio measurement 3 times at different depths and planes of tissue. Calculate the average of these three B-mode ratios for each animal and time point.

5. Image acquisition for Shear Wave Elastography

1. Move the probe transversely in the right subcostal area to locate the liver using B Mode. Locate an area of the liver that is mostly parenchyma and free of large blood vessels such as the portal vein and hepatic artery. Once a clear area of the liver has been found, generate a shear elasticity map of the tissue by pressing the SWE button on the control panel.
2. Adjust the size and position of the SWE box below the liver capsule in an area that is free of shadows. Identify the capsule as a bright echogenic line near the top of the liver.
3. Observe that the SWE box transitions to a color map within 5-10 s. Once the box is full and stable, press the Freeze button on the control panel when the animal is between breaths.
   NOTE: The minimum amount of box coverage should be 60-80% to accurately assess the elasticity of the liver.
4. On the touch screen, tap QBox, a built-in system tool that computes elasticity from an ROI on the shear wave elasticity map. Observe the circle and data box that appear on the monitor. Adjust the position of the QBox by tapping the position icon on the touch screen to the desired setup.
5. Adjust the size of the circle to 3 mm by moving a finger along the outer edge of the trackball on the control panel. Using the trackball, position the circle in an area free from shadow with uniform coloring (Figure 5A,B). Take care to avoid known areas of stiffness such as blood vessels or the liver capsule, as well as bleed-down from these structures.
6. When an adequate area is found, press Save Image on the control panel to save the image. Repeat this procedure 3 times in different areas of the liver. Move the probe up and down or sideways on the abdomen to gather SWE-mapped images from different areas of the liver.
7. Once all images have been collected, press End Exam on the control panel, and make note of the patient information screen that appears on the monitor.
8. Remove the tape from the animal's paws, wipe excess gel away, and remove the animal from the imaging stage. Allow it to recover from anesthesia in a warm, dry cage by itself until fully recovered. Monitor each animal to ensure full recovery from anesthesia, indicated by its ability to maintain sternal recumbency
9. Repeat steps in sections 4-5 for each animal in the cohort to be imaged.

6. Image data retrieval and analysis

1. When images for all animals have been collected, turn off the anesthesia.
2. To pull image data from the machine, press the Review button on the control panel, and observe all the scans performed on that instrument that appear on the monitor. Search for the desired scans using the search window in the upper corner of the screen.
3. Select all scans needed for data analysis by checking the box next to the name of the patient via the trackball and Select button. Once all needed scans are highlighted, select Export JPEGs on the touch screen. Export data to a network drive or a portable Universal Serial Bus (USB) drive. Locate the USB ports on the back of the instrument.
4. Once the files have been exported, open individual jpg files of each scan on a workstation computer. Observe all the data on the right side of the image: B Mode Ratio-collect the B Ratio number; Q Box-collect the mean elasticity (kPa) value.
5. Enter all data into a spreadsheet or other database management software, and perform the desired statistical analyses.

7. Histological analysis of liver samples

1. At the end of the 6^th week, perform necropsy on half of the cohort from each group for mid-study histological analysis of the liver samples. Similarly, euthanize the rest of the cohort of animals, and collect liver samples for histological analysis at the 12^th week time point.
2. For ORO staining, fix liver sections in 10% neutral-buffered formalin, and cryopreserve them with sucrose using a refrigerated 30% sucrose solution overnight at a minimum. Cryo-embed the sections in optimal cutting temperature compound, and cryo-section them onto charged slides to prepare for ORO staining.
3. Place the cryo-sections in 100% propylene glycol for 2 min followed by an overnight incubation in 0.5% ORO solution. After removal from the ORO solution, differentiate the sections in 85% propylene glycol for 1 min, rinse in deionized water, and counterstain with Mayer's Hematoxylin-Lillie's modification for 1 min.
   1. Place coverslips on the slides using an aqueous mounting medium and dry them at room temperature.
4. For PSR, deparaffinize formalin-fixed, paraffin-embedded liver section slides, place them overnight in Bouin Fluid, and then stain them using an automated slide stainer as per the manufacturer's protocol with some optimized steps (1% phosphomolybdic acid for 5 min; 0.1% Sirius Red in saturated picric acid for 90 min; 2 x 30 s wash in 0.5% acetic acid). Automatically dehydrate the slides and then mount them with a permanent mounting medium.
5. Capture images of the ORO- and PSR-stained slides using the digital microscopy scanner at 20x magnification, save them in .svs format, and store in the slide manager image database.
6. Analyze the images using custom algorithms created in digital pathology software. Uniformly apply digital pathology software applications with threshold parameters to identify and quantify the liver sections area as well as the ORO- and PSR-stained areas. Export the measurements to a spreadsheet for area percentage calculations.

8. Statistical analysis

1. Perform statistical analysis of the imaging data with two-way ANOVA using Sidak's multiple comparisons test to assess the difference between groups at different time points. Assume significant differences between groups for probability values p ≤ 0.001. In addition, perform correlation of imaging readouts with histological analyses.
2. Use non-parametric statistics to analyze the histological analysis results from this study. Report the group values as median ± semi-interquartile range (sIQR). Assume significant differences between groups for probability values p ≤ 0.001. Use a Mann-Whitney test to compare the quantity of PSR and ORO histochemical stain between different groups.

Results

One hallmark of animals fed CDAHFD is steatosis. Accumulation of fat in the liver changes the echogenic properties of the tissue, which can be quantified by measuring the brightness of the liver and normalizing it to the brightness of the renal cortex from a B-mode image taken in the same plane. The quantified value is expressed as an HR index, which is an indirect measure of steatosis. In Figure 4A, a representative liver image from a control animal shows approximately equal or less brightn...

Discussion

Ultrasound-based imaging, including SWE, can be an invaluable tool for the longitudinal assessment of liver steatosis and stiffness in preclinical models of NAFLD/NASH. This paper describes detailed methodologies on how to acquire high-quality B-mode as well as SWE images of livers for the measurement of the HR index and elasticity using a CDAHFD diet-induced rat model of NASH. Further, the results show excellent correlation of the HR index and elasticity with the gold standard of evaluation-histological assessment of li...

Materials

Name	Company	Catalog Number	Comments
Aixplorer	Supersonic Imagine		Shear Wave Elastography Instrument
Aixplorer SuperLinear SLH20-6 Transducer	Supersonic Imagine		Transducer for Shear Wave Elastography
Alpha-dri bedding			rat cages
Aperio AT2 scanner	Leica Biosystems		Digital Pathology Brightfield Scanner
Compac 6 Anesthesia System	VetEquip		Anesthesia Vaporizer and Delivery System. Any anesthesia delivery system can be used, however.
Manage Imager Database	Leica Biosystems		Digital Pathology
Mayer's Hematoxilin	Dako/Agilent		H&E Staining/Histology
Nair	Church & Dwight		Hair remover
Oil Red O solution	Poly Scientific		Lipid Staining/Histology
Picrosirius Red Stain (PSR)	Rowley Biochemical	F-357-2	Collagen Stain/Histology
Puralube Opthalmic ointment	Dechra Veterinary Product		Lubrication to prevent eye dryness during anesthesia
Tissue-Tek Prisma Plus	Sakura Finetek USA		Automated slide stainer
VISIOPHARM software	Visiopharm		Digital pathology software
	Research Diets	A06071309i	NASH inducing diet
	Purina	5053	Control animal chow
Vevo imaging station	Fujifilm VisualSonics		The Vevo imaging station is used for holding the ultrasound transducer during imaging.
Wistar Han rats	Charles River Laboratories