Vibration-controlled Transient Elastography

A Practical Approach to the Noninvasive Assessment of Liver Fibrosis

Elliot B. Tapper; Nezam H. Afdhal

Disclosures

Curr Opin Gastroenterol. 2015;31(3):192-198. 

In This Article

Vibration-controlled Transient Elastography for the Patient With Chronic Hepatitis C Virus Infection

HCV eradication is within reach.[21] Given the tremendous cost of these medications, if there will be any delay in therapy, the patients with the greatest need should receive priority.[5,22] In cost-effectiveness models, treating all-comers is typically the dominant strategy.[23] Some studies use liver biopsy for risk stratification,[24] others simply treat all without fibrosis assessment (albeit with an interferon-based therapy).[23] However, while such studies have proliferated, they incompletely capture the dilemma of clinical practice with patient waiting lists created as a function of rationing by insurance companies. Still, when examined, the use of VCTE in a treatment algorithm was associated with a highly favorable cost of $6066 per quality-adjusted life year.[23] In contemporary American practice, it is neither safe to assume that most patients would accept interferon-based therapy nor the risk and discomfort of liver biopsy. Risk stratification is an essential component of practice and must be conducted as safely and effectively as possible.

VCTE is a validated and reliable tool for the management of American patients with viral hepatitis.[25] Critically, VCTE performs well, identifying cirrhosis irrespective of low or high BMI with equal area under the receiver-operating curve (AUROC) values (0.92). However, we advocate an approach to the patient with HCV that combines the use of VCTE and serum biomarkers of fibrosis.[12]

By assessing a patient using both noninvasive strategies, the vast majority of patients can be effectively categorized into risk strata within 1–2 clinic visits. The European ANRS HCEP 23 Fibrostar Group argues that the test performance of VCTE and the conventional blood tests are the same and need not be combined. Indeed, the AUROC for the determination of cirrhosis by commonly available tests (in the United States) are FibroSure (0.87), HepaScore (0.89), Fibrosis-4 (FIB-4; 0.84) and FibroScan/VCTE (0.93).[26] These values are reproducible.[27,28] The main reason for using both strategies, however, is that in the event that one cannot obtain a LSM or if that LSM is indeterminate, the flow of clinical decision making is aided by having complementary or confirmatory information from another test.[12]

The LSM values to know are those that carry the strongest prognostic information. First, a cutoff of 7.3 kPa offers excellent (near perfect) negative predictive value (NPV), excluding patients without advanced liver disease, stage 2 or above.[19,29] Second, a cutoff of 12.5 kPa has a PPV of 77%[29] and NPV of 98% for the presence of cirrhosis.[30] Beyond that, a cutoff of 12.5 kPa is significantly and powerfully associated with poor outcomes. In a landmark study, Poynard et al.[28] showed how VCTE and FibroSure can be used to group all patients with chronic HCV into risk categories.[28] Although higher cutoffs captured the patients at highest risk, 12.5 kPa predicted the development of hepatic decompensation and hepatocellular carcinoma.[28] In our hands, a cohort of 667 patients, 67.5% of whom had chronic HCV, the strongest predictor of a hepatic decompensation was an LSM greater than 12.5 kPa [hazard ratio 18.9995% confidence interval (CI) (5.81–62.08), NPV 98%].[31] Similarly, in a cohort of patients undergoing anti-HCV therapy, baseline and serial VCTE were predictive of overall survival. In a multivariate model of mortality risk that adjusted for demographic factors, baseline VCTE (per kPa) was associated with a hazard ratio of 5.76 95% CI (3.74–8.87), increased LSM on follow-up had a hazard ratio of 1.19 95% CI (1.11–1.28) and sustained virological response had a hazard ratio of 0.19 95% (0.05–0.80). In this study, sustained virological response at any LSM or a baseline LSM of less than 7 kPa was associated with a good outcome. The patients with poor outcomes, on the other hand, were those with a 14 kPa or greater at baseline with any LSM increase on follow-up.[27]

processing....