Ultrasound Elastography of the Liver: What the Clinician Needs to Know

With the recent US Food and Drug Administration approval of several ultrasound elastographic techniques for the evaluation of liver stiffness, additional information can now be provided to the hepatologist to aid in the appropriate treatment and follow-up of patients with chronic liver disease. Increased understanding of how liver stiffness results can be helpful to clinicians is important for radiologists to appropriately devise imaging protocols and report results to improve patient treatment and outcome. Cirrhosis is the end stage of chronic liver disease, characterized by progressive parenchymal fibrosis. It can result from any etiology that causes chronic inflammation of the liver, including alcohol, hepatitis B and C, nonalcoholic steatohepatitis, autoimmune hepatitis, and cholestasis. Cirrhosis is a considerable source of morbidity and mortality in the United States, with a prevalence of approximately 0.27%, or 633,323 adults, and a 2-year mortality rate of 26.4% compared to 8.4% in propensity-matched controls.1 Clinically, there are at least 2 distinct stages of cirrhosis: compensated and decompensated. Decompensated cirrhosis is a relatively easy clinical diagnosis with poor prognostic implications. Findings include variceal hemorrhage, ascites, encephalopathy, and jaundice. Compensated cirrhosis, on the other hand, is often not clinically evident. Detecting compensated cirrhosis is important because median survival is substantially better in this population, and interventions can be taken to prevent progression. In a prospective cohort study of 257 patients with compensated cirrhosis of different etiologies, median survival for all patients was approximately 9 years, whereas it was significantly lower in patients who developed a decompensating event (ascites, jaundice, encephalopathy, or hemorrhage), in whom the median survival was only 1.6 years.2 Therefore, the term “cirrhosis” by itself really does not adequately characterize an individual patient clinically anymore. Patients with compensated cirrhosis can be further risk stratified on the basis of the presence or absence of varices. Bruno et al3 demonstrated that in patients with hepatitis C, the probability of both death and decompensation increases if varices are present. The difference in 1-year mortality increases from 1% in those without varices to 3% in those with varices.4 One of the strongest predictors for the development of varices is increased portal pressure. However, measurement of pressure via direct catheterization of the portal system is difficult to obtain, as the portal vein is not easily accessed, and doing so can result in complications. The hepatic venous pressure gradient (HVPG) is a safer and reproducible method to predict the likelihood of developing varices. The HVPG is the difference between the free hepatic venous pressure (a measurement of systemic blood pressure) and the wedged/occluded hepatic venous pressure (a measurement of sinusoidal pressure). Normal HVPG values are typically within the range of 3 to 5 mm Hg. An HVPG value of greater than 10 mm Hg is strongly associated with the development of varices and is termed “clinically significant portal hypertension.”5 Patients with an HVPG of less than 10 mm Hg were shown to have a 90% chance of not developing clinical decompensation in a 4-year period.6 In patients with posttransplant recurrent hepatitis C, the HVPG was shown to predict the development of decompensation, as patients who had an HVPG of less than 6 mm Hg were much less likely to develop ascites or encephalopathy. An HVPG of greater than 6mm Hg has been shown to be a more effective predictor of short-term decompensation, and a resultant increase in mortality, than the fibrosis score.7 A wedged HVP has been shown to correlate with liver biopsy in terms of progression of fibrosis and portal hypertension.8 The HVPG can also be used to monitor changes in fibrosis after therapy. For example, in posttransplant patients with recurrent hepatitis C, the HVPG reflected changes in histologic fibrosis after antiviral therapy.9 Reducing a patient's HVPG by greater than 10% at 1 year protects against the development of varices.10 A decrease of greater than 10% also improves outcomes in patients with large varices that have not bled, with responders also showing a lower probability of developing ascites.11 There are several different methods for performing elastography on the liver, including transient elastography (TE), radiation force impulse (ARFI) point shear wave elastography (SWE) and 2-dimensional SWE, and real-time elastography. Real-time elastography has limited literature, is offered by only one vendor, and requires a specific image analysis program. One of the major differences is that ARFI, SWE, and real-time elastography and are image-based techniques, meaning that they are incorporated into a standard ultrasound machine compared to TE, for which no image is available. This factor offers several benefits, including being able to visualize the interrogated region (excluding masses, large blood vessels, and bile ducts) while conducting elastographic measurements, thus ensuring the proper position. It also allows for subjective evaluation of the liver, which may correlate with advanced fibrosis. Last, it can be used to screen for hepatocellular carcinoma and other complications.12 One of the major benefits of TE, on the other hand, is the convenience of being able to obtain measurements without a sonographer in a clinical setting. A detailed review of each of these techniques, including the basic science of each technique, can be found elsewhere.13 The efficacy of these methods has been well studied. Acoustic radiation force impulse SWE techniques and TE have been shown to have similar diagnostic accuracy for substantial fibrosis and cirrhosis across several studies.14-20 Acoustic radiation force impulse SWE measurements have been shown to be more reliable for liver stiffness than TE.14 When it comes to the severe stages of fibrosis, TE may have higher sensitivity than ARFI SWE.21 The accuracy of SWE increases with the increasing severity of fibrosis and may be even better than TE for fibrosis stage F2 or greater.22 Shear wave elastography may have a higher technical success rate and better diagnostic value than TE for clinically significant portal hypertension.23 Using either the HVPG or the reference standard, liver biopsy, to screen for cirrhosis and varices is impractical and invasive, with the risk of complications. In addition, there are several other pitfalls to liver biopsy, including interobserver and intraobserver variability in histologic interpretation, sampling errors, and unequal distribution of fibrosis in the liver.24, 25 Thus, a noninvasive test, such as elastography, is preferred, as it can safely be repeated over a short time and has the potential to monitor changes in fibrosis in response to treatment. A detailed analysis determined similar accuracy and cost savings for point SWE.26 Judging the efficacy of elastography centers around several questions: Can it accurately determine whether cirrhosis is present? If cirrhosis is present, can it predict the presence of clinically significant portal hypertension and gastroesophageal varices? Can the response to treatment be measured or predicted? Elastography has consistently been shown to be accurate in staging chronic liver disease and has even been touted as a “crystal ball” that can help clinicians predict the outcomes of patients and help with therapeutic decisions.27 Two-dimensional SWE, for example, is a reliable method for determining liver stiffness measurements in nearly all patients and effectively predicts cirrhosis.28 Transient elastography is both an accurate and reproducible way of assessing liver fibrosis, and even possibly steatosis, in patients with a spectrum of liver disease and may even be able to predict outcomes.29 Strain elastography has been used to evaluate liver fibrosis but has not gained acceptance, and most centers now use ARFI shear wave techniques, mechanical shear wave techniques (FibroScan; Echosens, Paris, France) or magnetic resonance elastography. The recommended value for liver stiffness to rule in cirrhosis is greater than 2.2 m/s (>15kPa), and the optimal cutoff to distinguish nondiseased liver from noncirrhotic chronic liver disease is 1.5 m/s (<7 kPa), with slight variability depending on the vendor.30, 31 Using a low cutoff value below which the patient is most likely normal, a high cutoff at which the patient has severe fibrosis or cirrhosis, and a middle group that has mild to moderate fibrosis triages patients into appropriate treatment groups: a group with no or minimal fibrosis, which may not need any additional follow-up depending on the initiating cause; a group with mild to moderate fibrosis, which needs continued follow-up; and a group with severe fibrosis, which needs immediate clinical evaluation.32 However, others have suggested that the usefulness of elastography can be maximized by not just using a single cutoff but, rather, a continuum.30 Diagnosing cirrhosis clinically can be accomplished by laboratory tests, including the complete blood count and liver function tests, as well as evaluating for signs of complications of cirrhosis. However, TE has been shown to be superior to clinical signs, laboratory parameters, and traditional sonography for the detection of cirrhosis. For example, elastography detected an additional 10% to 16% of patients with chronic liver disease who would have been missed otherwise.33 Transient elastography also increases the accuracy of biopsy and aspartate aminotransferase-to-platelet ratio index analyses in identifying patients with cirrhosis.34 Although certain serum marker tests, such as the enhanced liver fibrosis test, perform well in assessing the degree of fibrosis, elastography is even more reliable when it comes to assessing severe fibrosis and cirrhosis.35 Transient elastography has been shown to help diagnose clinically occult cirrhosis, thus leading to early therapy and surveillance.36 In particular, TE has been shown to be able to identify subclinical cirrhosis in patients with chronic hepatitis B who are at risk for hepatocellular carcinoma and would thus benefit from screening.37 Elastography has been shown to be a reliable tool for assessing the degree of fibrosis. Real-time tissue 2-dimensional SWE outperforms laboratory tools, such as the aspartate aminotransferase-to-platelet ratio index and the fibrosis-4 index, in predicting severe fibrosis.38 Liver stiffness measurements using vibration-controlled TE are able to differentiate severe from nonsevere liver fibrosis with a high level of confidence in patients with primary sclerosing cholangitis. These baseline measurements and interval changes can then be used as prognostic factors.39 Elastography can also be used in conjunction with clinical markers, such as the Forn index, to increase the predictive value when the fibrosis stage is S2 or greater.40 Two other populations may also benefit from the use of elastography: patients with nonalcoholic steatohepatitis and children. For example, patients with simple steatosis can be differentiated from those with nonalcoholic steatohepatitis, and the degree of fibrosis can be evaluated.41 Now, with new drugs that can reverse nonalcoholic fatty liver disease, elastography may be helpful in determining the time to treat and monitor drug efficacy. Reliable liver stiffness measurements obtained with TE can thus target patients with severe fibrosis, who could benefit from intervention.42 Estimation of liver stiffness with elastography is also possible in pediatric and adolescent patients, as shear wave measurements of liver stiffness correlate strongly with the degree of fibrosis in this population as well.43, 44 Several studies have determined that point SWE and 2-dimensional SWE have similar if not higher accuracies in evaluation of liver stiffness.28, 45 In addition to liver stiffness, elastography can also assess another useful measurement: spleen stiffness. Both liver and spleen stiffness increase with progression of liver fibrosis. In cirrhotic patients, the difference between liver and spleen stiffness decreases.46 This factor is another potential tool that a clinician can use to assess the presence and degree of fibrosis. As previously discussed, the presence of varices is a considerable risk factor for the development of death and decompensation, so being able to detect varices early has the potential to affect patient care substantially. Liver stiffness measurements by elastography are able to predict the presence of esophageal varices, as well as determine whether large varices are present.47 Two-dimensional SWE is accurate for diagnosing clinically significant portal hypertension as well, given that certain reliability criteria are met. These criteria include a standard deviation-to-median ratio of less than 0.10 and a depth of less than 5.6 cm.48 FibroScan, alone or in combination with the platelet count, may be able to predict both the presence and severity of esophageal varices in patients with hepatitis B or C cirrhosis.49 Liver stiffness has been shown to correlate with the HVPG in patients with hepatitis C virus and stage F3 or F4 fibrosis.50 Spleen stiffness may correlate even better with the HVPG than liver stiffness and is able to predict clinically significant portal hypertension and varices.51 Spleen stiffness has been shown to directly correlate with the presence, severity, and bleeding risk of esophageal varices. For example, in 340 patients undergoing endoscopy for evaluation of esophageal varices, spleen stiffness measurements by ARFI were used to rule out the presence of varices and thus may serve as an initial noninvasive screening test.52 In another study, spleen stiffness alone or in combination with liver stiffness was able to reliably and reproducibly exclude the presence of varices.53 Using a spleen stiffness cutoff value of 3.4 m/s is an effective way to distinguish between high- and low-risk varices.54 The ratio of spleen stiffness to liver stiffness may even be able to accurately distinguish whether portal hypertension is idiopathic or due to underlying chronic liver disease.55 One of the most promising uses of elastography is the ability to monitor changes in liver fibrosis, which can allow for treatment surveillance, risk stratification, and monitoring of complications.56 An accurate and noninvasive test has a tremendous advantage in this situation, in which multiple invasive procedures become particularly impractical, and elastography has been shown to be effective in this setting. For example, liver stiffness values significantly decreased, reflecting improvement in fibrosis, in patients with chronic hepatitis B after a 3-year period of treatment with entecavir. Higher baseline liver stiffness values were also the only independent predictors of a significant decline in fibrosis.57 In patients receiving dual or triple therapy for hepatitis C, pretreatment liver stiffness values may be useful for predicting which patients will respond favorably as well.58 Liver stiffness by TE also accurately predicts the risk of death or hepatic complications in patients with chronic liver disease, which can determine the prognosis and guide management.59 In patients with chronic hepatitis B, for example, liver stiffness values were predictive of liver-related events during a 2- year treatment period with entecavir.60 This finding suggests a potential role for long-term monitoring of dynamic changes in fibrosis in patients being treated with entecavir. Liver stiffness values at a complete virologic response are also useful for predicting future liver-related events.61 Therefore, elastographic evaluation at a complete virologic response may become an important and widely used prognostic examination for cirrhotic patients at that point in their treatment. Cohort studies indicate that hepatocellular carcinoma is currently the major cause of liver-related death in patients with compensated cirrhosis.62 The 5-year cumulative incidence varies with the inciting etiology, with hepatitis C virus at 30% in Japan and 17% in the West; hereditary hemochromatosis, 21%; hepatitis B virus, 15% in endemic areas and 10% in the west; alcoholic cirrhosis, 8%; advanced biliary cirrhosis, 4%; and nonalcoholic steatohepatitis, 2% to 12%. This variance is another reason to identify patients with compensated cirrhosis, as these patients need to have routine screening for hepatocellular carcinoma. There are limited studies on the effect of treatment on the development of hepatocellular carcinoma. One study found an adjusted risk ratio for development of hepatocellular carcinoma in treated hepatocellular carcinoma patients of 0.516.63 Therefore, treatment may the risk of hepatocellular it does not to and treated patients may need routine screening for hepatocellular carcinoma. of the elastographic is important to reliable and reproducible Several liver stiffness and being able to predict the effect or the effect can prevent in For example, patients be an elastographic as liver stiffness measurements increase significantly after also needs to be taken into as liver stiffness measurements decrease in patients who are from A factor in accurate liver stiffness is measurements in 1 the substantial in liver stiffness that can between measurements in a Liver affect the cutoff values used for of fibrosis, with lower in patients with values for predicting severe esophageal varices are also by the etiology of liver disease, with higher for alcoholic cirrhosis The index, in particular, a value of greater than has a with liver stiffness measurements after for the fibrosis the ARFI leading to a such as in patients with may also effect the of liver stiffness measurements in that this without severe liver disease has a liver stiffness at but when the measurements are taken in the patient has a stiffness value with cirrhosis. are also associated with a higher likelihood of a or liver stiffness For example, a cohort of more than patients that and a higher index increase the chance of an liver stiffness However, liver stiffness measurements were reliable in of Several other studies have suggested that may to more of steatosis and in the liver may to of that may result in measurements include an the presence of hepatitis C higher and to performing the elastographic examination can also affect liver stiffness For example, the of in the liver can affect with and showing the least between repeated The has been shown to have the with liver fibrosis SWE predicts when at the biopsy or in patients with hepatitis The depth of in the liver is also with optimal values of 5 to 6 the for and 3 to for The measurement also be taken at least 1.5 to 2 below the liver and have similar and accuracy for liver stiffness can affect the reliability of elastography, as the diagnostic accuracy of TE increases with In in which an is to obtain liver stiffness values with the traditional the may be In the has been shown to be able to obtain reliable measurements in of patients with measurements by the Spleen stiffness measurements also on the of the and a period is recommended to improve of spleen stiffness are less reproducible than liver stiffness There is in the the of measurements that are to accurately assess liver stiffness. studies used a median of 10 measurements for However, studies suggested that a single measurement may be all that is to assess for fibrosis, and there is no of accuracy compared to multiple have shown that 3 measurements and using the than are whereas at the other 6 measurements were as the optimal for intraobserver evaluation is to determine the appropriate of measurements for accuracy of liver stiffness. It is recommended that when this technique, 10 measurements be taken that a of measurements would be The reliability of liver stiffness measurements can be by the range by the The value can be used to distinguish between 3 of liver stiffness reliable reliable 0.10 and or if the liver stiffness evaluation median is m/s and reliable the if liver stiffness evaluation median is m/s These have been shown to be able to that used criteria such as more than 10 measurements and a greater than success also increase the of patients with accurate measurements without diagnostic for cirrhosis and portal For the to be 10 measurements be liver analysis is a the the liver in of to substantial fibrosis. It has been shown to be more effective than subjective analysis and similar to that of and could serve as a tool to standard sonography in the However, the usefulness for evaluating changes on follow-up limited at this patients and patients with ascites often a for elastography because of the and elastographic However, elastography on sonography can be into a clinical and used to evaluate tissue and greater This could be a useful for assessing fibrosis in this patient Last, magnetic resonance elastography has been shown to have even better diagnostic accuracy for staging fibrosis than TE, but the of cost and of the examination usefulness for repeated monitoring at this As the elastographic techniques that sonography also allow for the screening of liver in a patient population at risk for hepatocellular carcinoma. may be able to be with elastography. For example, a that elastography has high sensitivity and for between and However, the is with a studies that ARFI SWE differentiate the of as there is a considerable of stiffness values between and 2 that even within a single could be whereas others are The combination of increased liver stiffness and values with and may useful as hepatocellular carcinoma was detected with accuracy in these patients with hepatitis C In this patient with liver disease from one is with a stiffness value of m/s kPa), whereas another is with a stiffness value of m/s Several new and effective for hepatitis B and C have shown the ability to affect disease progression and even reverse For example, a virologic response for in fibrosis is found in to in only to and 16% to studies have shown complete of fibrosis or portal inflammation after a virologic treatment of hepatitis B with is also with of cirrhosis in of patients and of fibrosis in after 5 of There are even promising for nonalcoholic fatty liver disease with which has been effective in of fatty liver and fibrosis, further is these elastography has become even more for ability to stage fibrosis. fibrosis is important for prognostic determining when patients need for hepatocellular carcinoma or varices, a baseline for monitoring the treatment the treatment as treatment in hepatitis C drug The interval for monitoring changes with elastography on several including the underlying etiology of fibrosis, disease severity, of of and efficacy of and patient Cirrhosis can be into groups: compensated and which have substantially different Although more difficult to detect compensated cirrhosis has a better and early treatment may prevent progression. Although invasive such as liver biopsy and the HVPG are effective for evaluating fibrosis and the risk of varices, their invasive their The has shown that elastography is an effective and noninvasive tool for staging fibrosis, predicting clinically significant portal hypertension and varices, and monitoring the response to treatment. the different elastographic methods to monitor liver stiffness, those that include a greater benefit because of their ability to an appropriate region of for liver stiffness measurements, assess for signs of advanced fibrosis, and screen for hepatocellular carcinoma in an As new and effective for the of cirrhosis become more the use of ultrasound elastography to increase as