Noninvasive diagnosis of liver fibrosis in adults

Watch a video presentation of this article

Watch the interview with the author

Abbreviations

ALT

alanine aminotransferase

APRI

aspartate aminotransferase/platelet ratio index

AST

aspartate aminotransferase

AUROC

area under receiver operating characteristic

2D SWE

two‐dimensional shear‐wave elastography

ELF

enhanced liver fibrosis

HCV

hepatitis C virus

MRE

magnetic resonance elastography

MRI

magnetic resonance imaging

TIMP

tissue inhibitor of metalloproteinase

ULN

upper limit of normal

US

ultrasound

VCTE

vibration controlled transient elastography

Noninvasive methods incorporating serum‐based markers (biological properties) or radiology‐based techniques using elastography (physical properties) are increasingly being used to indirectly assess liver fibrosis.

Noninvasive Serum Markers

Types

Serum markers can either be specific or nonspecific to the liver. They may either be direct markers of the fibrosis process (e.g., glycoproteins and collagens) or indirect markers of hepatic dysfunction (e.g., prothrombin time, platelet count, and albumin). Some serum marker panels are proprietary (e.g., FibroTest or enhanced liver fibrosis [ELF] score); others combine commonly obtained laboratory tests (e.g., aspartate aminotransferase [AST]/platelet ratio index [APRI]), patient characteristics (e.g., FIB‐4, nonalcoholic fatty liver disease fibrosis score, or BARD score), or are results of multivariable analysis (Figure 1 and Table 1).

Figure 1.

Calculation of APRI. Abbreviation: ULN, upper limit of normal.

Table 1.

Selected Serum Markers for Assessment of Fibrosis in Patients With Hepatitis C

Nonspecific to HCV	Components	Equation	Cutoffs
Indirect
AST/ALT ratio	AST, ALT	AST/ALT	0.8 and 1
FIB‐4	Platelet, AST, ALT, age	(Age × AST)/(platelets × (sqr (ALT))	1.30 and 2.67	0.84
APRI	Platelet, AST	AST/AST ULN/platelets × 100	0.5 and 1.5	0.83
FibroTest	Age, sex, alpha‐2 macroglobulin, alpha‐2‐globuin, gamma‐globulin, apolipoprotein A1, GGT, total bilirubin	Commercial panel		0.86
Forns index	Platelet, GGT, age, cholesterol	7.811 − 3.131 × In(platelet count) + 0.781 × In(GGT) + 3.467 × In(age) − 0.014 × (cholesterol)		0.87
Direct
ELF	Hyaluronic acid, TIMP‐1, type III collagen	Commercial panel		0.78

Abbreviations: ALT, alanine aminotransferase; GGT, gamma glutamyl transpeptidase; TIMP, tissue inhibitor of metalloproteinases; ULN, upper limit of normal.

Diagnosis

Most markers accurately identify patients with cirrhosis or normal liver parenchyma but inaccurately differentiate between intermediate stages of fibrosis. In hepatitis C virus (HCV) studies, the area under receiver operating characteristic (AUROC) for diagnosis of cirrhosis at the respective cutoff was 0.84 (APRI) versus 0.86 (FibroTest), 0.86 (APRI) versus 0.87 (Forns index), 0.89 (APRI) versus 0.88 (platelet count), and 0.86 (APRI) versus 0.84 (FIB‐4).1 In a separate multicenter study, addition of FibroTest to APRI decreased the number of biopsies required from 71% to 53.5% for the diagnosis of cirrhosis.2

Limitations

Serum markers are easily repeatable, noninvasive, performed in an outpatient setting, and applicable to large populations.3 Limitations include generalizability, cost of proprietary tests, overestimation in certain conditions (e.g., elevated bilirubin in Gilbert's syndrome or hemolysis), and an inability to discriminate between intermediate stages of fibrosis.3

Elastography

Types

Elastography evaluates the intrinsic property of the liver parenchyma, or liver stiffness, with elevated stiffness associated with presence of advanced fibrosis. Elastography‐based techniques can be either ultrasound (US) based or magnetic resonance imaging (MRI) based (Figure 2). Elastography (strain or shear wave) requires either mechanical excitation of tissue by manual compression, use of acoustic radiation force impulse, or controlled external vibration. Shear‐wave elastography monitors the propagation of shear waves in tissues. Shear waves propagate faster in stiffer tissue. Some common techniques include vibration controlled transient elastography (VCTE), two‐dimensional shear‐wave elastography (2D SWE), and magnetic resonance elastography (MRE) (Figure 3). For US‐based techniques, results are often expressed in kilopascals, with a value of approximately 5 kPa or less indicating normal liver and a value greater than 12 to 14 kPa implying cirrhosis. For MRE, results are approximately one‐third the values of transient elastography, with values less than 2.9 kPa denoting normal liver and more than 5.2 kPa indicating presence of cirrhosis.

Figure 2.

Selected noninvasive markers of fibrosis (elastography). Reproduced with permission from Clinical Gastroenterology and Hepatology.12 Copyright 2015, American Gastroenterological Association.

Figure 3.

Elevated liver stiffness suggesting presence of cirrhosis by (A) VCTE, (B) 2D‐SWE, and (C) MRE. (A) Courtesy of FibroScan (Echosens). (B) Courtesy of SuperSonic Imagine. (C) Courtesy of S. Venkatesh, Mayo Clinic Rochester, MN.

Diagnosis

All modalities are better at diagnosing cirrhosis as compared with identifying significant fibrosis (stage 2 or higher). As compared with liver biopsy, at a cutoff of 14.5 kPa, VCTE had a sensitivity of 83% and a specificity of 89% for cirrhosis.4, 5 Among nonalcoholic fatty liver disease patients undergoing MRE, a cutoff greater than 3.63 kPa (AUROC 0.92) had 86% sensitivity and 91% specificity for discriminating advanced fibrosis from stage 0 to 2 fibrosis.6 VCTE and 2D SWE had similar performance characteristics for diagnosis of cirrhosis (AUROCs were 0.98 and 0.96).7 Studies comparing MRE and VCTE have conflicting reports on the superiority of either test, especially for intermediate stages.8, 9 MRE has higher completion rates (94% versus 84%).

Limitations

Association of absolute liver stiffness values with intermediate stages of fibrosis is suboptimal. Rates of incomplete examinations and unreliable examinations vary by technique, especially among obese patients. Acute inflammation, cholestasis, nonfasting state, elevated portal pressure, and hepatic congestion may overestimate liver stiffness.

Radiological markers have a superior diagnostic accuracy as compared with serum markers.10, 11 As a prognostic tool, a meaningful change in liver stiffness or serum marker from baseline may help identify patients at highest risk for progression with serial assessments over time.

Clinical Application (HCV)

A treatment‐naive patient is seen in the outpatient setting for hepatitis C. She undergoes noninvasive testing either by use of serum markers or radiological markers. Based on local availability, her APRI is 6 and liver stiffness measurement is 5.5 kPa (MRE) and 15.5 kPa (2D SWE), which are all suggestive of cirrhosis. She undergoes antiviral therapy and achieves sustained virological response. Given the presence of cirrhosis, she undergoes appropriate management including surveillance for esophageal varices, as well as screening for hepatocellular carcinoma. At 1 year after successful completion of therapy, she returns to the clinic and undergoes elastography. Her liver stiffness measurement is 5.0 kPa (MRE) and 13.5 kPa (2D SWE), suggesting improvement. She continues to undergo appropriate surveillance. In the future, she may undergo stiffness measurement every 1 to 2 years for prognosis and decisions on interval of follow‐up.

Summary

Cutoffs for presence of advanced fibrosis or cirrhosis vary by technique and patient population. All modalities are better at the diagnosis of cirrhosis as compared with significant fibrosis (stage 2 or higher). Serum markers overall have high negative predictive value and are better at ruling out presence of advanced fibrosis. As compared with serum markers, radiological markers have higher diagnostic accuracy for detection of cirrhosis. For radiological markers, unreliable and incomplete examinations limit their use, especially in obese patients. Finally, the translation of findings and implications in the adult population need to be comprehensively developed and validated in the pediatric population.