This presentations reviews the rationale, principles, and practical application of MRI-based methods for diagnosing hepatic fibrosis. MR elastography is addressed in detail.
Chronic liver disease is caused by many conditions, including chronic viral hepatitis, metabolic disease, exposure to toxic chemicals, autoimmune disease, chronic venous congestion and others. A common mechanism in most of these etiologies is hepatocellular injury, leading to necroinflammation and activation of hepatic stellate cells. The macromolecular composition of the extracellular matrix becomes modified, with cross-linking and accumulation of proteins, including collagen, which can be identified as fibrosis microscopically. With ongoing chronic injury, the healing response results in coalescence of fibrotic areas and scar formation which distorts the anatomy of the liver, initially at a microscopic level but eventually progressing to affect the gross morphology of the liver.
If the progression of liver fibrosis is not halted or reversed by treatment, the condition can advance to cirrhosis and end stage liver failure, which is irreversible and associated with high mortality. Prior to the onset of cirrhosis, liver fibrosis rarely causes any anatomic change and therefore conventional imaging modalities are not sensitive for making this diagnosis.
For decades, the routine method for diagnosing hepatic fibrosis has been liver biopsy. Standardized histologic classifications have been developed for assessing the degree of fibrosis, necrosis, and inflammation. However, liver biopsy is an expensive invasive test with a not-insignificant risk of complications, including mortality. The reliability of liver biopsy for assessing liver fibrosis is also affected by sampling error and subjectivity of interpretation.
A number of imaging-based methods have been proposed and evaluated for noninvasive assessment of liver fibrosis. Conventional anatomic imaging with ultrasonography, computed tomography, and MRI can reveal nodularity, atrophy, and other features indicative of advanced liver fibrosis or cirrhosis but these modalities are not helpful for detecting and assessing earlier stages of the disease, when it is reversible. Parenchymal texture analysis with these modalities has also been evaluated and generally shows satisfactory diagnostic performance in advanced fibrosis.
Several MRI tissue-characterization techniques have been evaluated as potential methods for non-invasively assessing liver fibrosis. Some the most promising of these methods have included diffusion-weighted imaging, T1 and T2 relaxometry, and MR elastography. A survey of the rationale and current evidence for each of these techniques is beyond the scope of this abstract, but been addressed in a recent review article [1].
Among the MRI-based methods, MR elastography (MRE) currently has the largest evidence base, most standardization across vendors, and widest usage use for diagnosing liver fibrosis. Liver fibrosis isassociated with increased hepatic tissue stiffness and MRE is an MRI-based technique for quantitatively assessing tissue stiffness [2]. It was first introduced as an FDA-approved product in the US in 2009 and since then it has been made available by several manufacturers as an upgrade to their MRI systems. The main application of MRE at this time is non-invasive assessment of liver fibrosis [3-6]. As of 2017, more than 800 MRI systems around the world had been equipped for MRE.
MRE is based on the physical principle that the propagation characteristics of mechanical waves within various materials are determined by their mechanical properties [2]. The technique consists of three steps: (1) generating mechanical waves in the region of interest, (2) imaging propagating mechanical waves, and (3) processing the information to calculate the mechanical properties. For assessing liver disease, mechanical waves are typically generated at 60 Hz in the upper abdomen with a flat disk-shaped vibration source that is placed against the body wall. During imaging, synchronous cyclic motion-sensitizing gradients are used with a modified phase-contrast MRI pulse sequence to acquire snapshots of the propagating waves, depicting displacements as small as fractions of microns. The acquired data are then automatically processed with an inversion algorithm to generate cross-sectional images showing the mechanical properties of tissues (typically shear stiffness) on a color scale.
The MRE acquisition is performed during breath-holding at end expiration and takes 12-15 seconds for each slice. This acquisition is typically repeated four times, for a total acquisition time of less than one minute. MRE is usually added to a conventional abdominal MRI protocol (either full or limited) and adds little additional time to the overall examination. Another option is to perform a very limited exam consisting only of MRE and a <60 sec PDFF sequence, which would provide quantitative estimates of fat fraction, iron content, and liver stiffness in an exam that could be accomplished in less than 10 minutes of scanner time, at very low cost.
After automatic processing, the scanner produces color-scaled quantitative images (“elastograms”) depicting tissue shear stiffness in units of kiloPascals (kPa). In addition, the algorithm provides anatomic images corresponding to each of the elastograms and “confidence images” that provide a measure of the reliablility of the tissue stiffness measurement at each image location.
At the Mayo Clinic, the most common indication for MRE is to assess possible hepatic fibrosis in patients who have conditions that are known to lead to this problem, such as fatty liver disease and chronic viral hepatitis. Other indications include follow-up of previously diagnosed fibrosis, staging patients with known cirrhosis, and evaluating patients with unexplained portal hypertension. Because the MRE acquisition can be added to a conventional abdominal MRI protocol with little or no impact on examination time, it does not require a high threshold of suspicion to be included in the protocol.
Since 2006, there have been dozens of published studies assessing the diagnostic performance of MRE in detecting and staging hepatic fibrosis, using biopsy as the reference standard. An MRE-based measurement of hepatic stiffness that is in the normal range (< 2.5 kPa) has a very high negative predictive value for ruling out hepatic fibrosis of any stage. Excellent diagnostic performance for staging hepatic fibrosis has been reported in multiple studies. For instance, a meta-analysis concluded that the sensitivity, specificity, and AUROC of MRE for diagnosing advanced hepatic fibrosis and cirrhosis (≥F3) from less-advanced disease are 92%, 96%, and 0.98, respectively [7]. These metrics are probably at the limit of what is realistic to achieve, given the known limitations of using biopsy as a “gold standard”. Another pooled meta-analysis of 12 published studies, encompassing 697 patients, found that the sensitivity, specificity, and AUROC diagnostic performance for diagnosing stage F3 fibrosis and higher are 85%, 85%, and 0.93 respectively [8]. Most studies that have compared the diagnostic performance of MRE with Transient Elastography (TE) have shown higher accuracy and fewer technical failures with MRE [9-11].
MRE has the same potential confounding factors as quantitative ultrasound-based elastography. Liver stiffness is affected by chronic and acute inflammation. The presence of chronic inflammation can cause considerable overlap in stiffness values between patients with stage F0 and stage F1 fibrosis. Acute hepatitis can be associated with very high liver stiffness values without any degree of fibrosis. Portal hypertension, hepatic venous congestion, and malignant cellular infiltrates can elevate liver stiffness independent of the presence of fibrosis.
The most common reason for technical failure of MRE has been hepatic iron overload, which is not uncommon in patients with liver disease. With conventional gradient echo MRE sequences, very high liver iron content may cause the signal intensity of the liver to be too low to visualize the mechanical waves, resulting in a failure rate of ~4% in clinical populations. The newly-introduced SE-EPI MRE sequences are much less sensitive to iron overload, making these technical failures much less common. Clinical experience has shown that the technical success of MRE is not affected by obesity unless the patient cannot fit in the scanner. The presence of ascites, common in patients with liver disease, does not affect the technical success rate of MRE. The FDA-cleared MRE products from major MRI vendors all use the same mechanical driver hardware, the same default shear wave frequency of 60 Hz, comparable pulse sequences, and the same data processing algorithm to compute tissue stiffness. They all report the magnitude of the complex shear modulus, use the same color scale in the images, and a default 0-8 kPa display. Testing in phantoms and human volunteers has provided preliminary confirmation that liver stiffness data obtained on systems from these three vendors can be compared on a valid basis. More than 10 published studies have assessed the test-retest repeatability of MRE in liver imaging [12]. In general, they have shown that differences in MRE-derived liver stiffness of greater than 19% represent meaningful longitudinal changes. This is a useful level of repeatability because the difference in mean stiffness between normal liver and advanced fibrosis it is approximately 200%.
Further technical developments, especially advances in pulse sequences and processing algorithms for advanced multiparametric 3D MRE are opening up new applications. MRE provides a range of novel quantitative imaging biomarkers that will merit exploration for many years to come.
1. Petitclerc L, Sebastiani G, Gilbert G, Cloutier G, Tang A. Liver fibrosis: Review of current imaging and MRI quantification techniques. J Magn Reson Imaging. 2016 Dec 16. [Epub ahead of print]
2. Muthupillai R, Lomas D, Rossman P, Greenleaf J, Manduca A, Ehman R. Magnetic resonance elastography by direct visualization of propagating acoustic strain waves. Science 1995;269: 1854–1857.
3. Venkatesh SK, Yin M, Ehman RL. Magnetic resonance elastography of liver: technique, analysis, and clinical applications. J. Magn. Reson. Imaging 2013;37: 544–555.
4. Venkatesh SK, Ehman RL. Magnetic resonance elastography of liver. Magn Reson Imaging Clin N Am 2014; 22: 433–446.
5. Yin M, Glaser KJ, Talwalkar JA, Chen J, Manduca A, Ehman RL. Hepatic MR Elastography: Clinical Performance in a Series of 1377 Consecutive Examinations. Radiology. 2016 Jan;278(1):114-24.
6. Venkatesh SK, Ehman RL, eds. In: Magnetic Resonance Elastography. Springer, New York, 2014.
7. Wang QB, Zhu H, Liu HL, Zhang B. Performance of magnetic resonance elastography and diffusion-weighted imaging for the staging of hepatic fibrosis: a meta-analysis. Hepatology 2012; 240-247.
8. Singh S, Venkatesh SK, Wang Z, Miller FH, Motosugi U, Low RN, Hassanein T, Asbach P, Godfrey EM, Yin M, Chen J, Keaveny AP, Bridges M, Bohte A, Murad MH, Lomas DJ, Talwalkar JA, Ehman RL. Diagnostic Performance of Magnetic Resonance Elastography in Staging Liver Fibrosis: A Systematic Review and Meta-analysis of Individual Participant Data. Clin Gastroenterol Hepatol. 2014 S1542-3565.
9. Huwart L, Sempoux C, Vicaut E, et al. Magnetic resonance elastography for the noninvasive staging of liver fibrosis. Gastroenterology 2008;135(1): 32–40.
10. Imajo K, Kessoku T, Honda Y, Tomeno W, Ogawa Y, Mawatari H, Fujita K, Yoneda M, Taguri M, Hyogo H, Sumida Y, Ono M, Eguchi Y, Inoue T, Yamanaka T, Wada K, Saito S, Nakajima A. Magnetic Resonance Imaging More Accurately Classifies Steatosis and Fibrosis in Patients With Nonalcoholic Fatty Liver Disease Than Transient Elastography. Gastroenterology. 2016 Mar;150(3):626-637.
11. Park CC, Nguyen P, Hernandez C, Bettencourt R, Ramirez K, Fortney L, Hooker J, Sy E, Savides MT, Alquiraish MH, Valasek MA, Rizo E, Richards L, Brenner D, Sirlin CB, Loomba R. Magnetic Resonance Elastography vs Transient Elastography in Detection of Fibrosis and Noninvasive Measurement of Steatosis in Patients With Biopsy-Proven Nonalcoholic Fatty Liver Disease.Gastroenterology. 2017 Feb;152(3):598-607.
12. Shire NJ, Yin M, Chen J, et al. Test-retest repeatability of MR elastography for noninvasive liver fibrosis assessment in hepatitis C. J Magn Reson Imaging 2011;34:947–955.