Histologic and noninvasive estimates of liver fibrosis

Serum Biomarkers

Today, many of the current serum biomarker algorithms adopted into the clinical setting include a combination of peptides or proteins derived from fibrogenic cells, extracellular matrix remodeling, and simple biochemical tests that estimate disease severity. Most of the available noninvasive tests for fibrosis have been validated in hepatitis C, with relatively fewer studies in other important CLDs, with the algorithm dependent on the underlying liver disease. The AST-to-Platelet Ratio Index (APRI) and the FIB-4 are easily calculated indices.7, 8 Other patented marker panels (FibroTest, Fibrometer, Hepascore) are available (Table 2). The most widely used and validated are the APRI and the FibroTest, mainly in viral hepatitis C. A recent systematic review including 172 studies conducted in hepatitis C reported median areas under the receiver operating characteristic curve (AUROCs) of 0.79 and 0.86 for FibroTest and of 0.77 and 0.84 for APRI, for significant fibrosis and cirrhosis, respectively.9 Stepwise or sequential algorithms combining serum markers and imaging assessment appear to improve diagnostic accuracy for significant fibrosis and may reduce the need for biopsy. In the context of NAFLD, simple algorithms such as the NAFLD fibrosis score may differentiate advanced nonalcoholic steatohepatitis, but a substantial percentage of patients will have indeterminate values.10 This algorithm has now been recommended in practice guidelines as being clinically useful in identifying NAFLD patients with advanced disease. It is noteworthy that noninvasive techniques were initially developed for binary disease staging outcomes (for example METAVIR F0-1 versus F2-F4). Whether they are adapted to follow the dynamic changes of fibrogenesis or fibrosis regression is an important question that remains. Indeed, certain biomarkers such as granulocyte monocyte colony-stimulating factor, interleukin-12, interleukin-2, and matrix metalloproteinase 13 might be linked with changes in fibrosis stage, but this requires further validation.11

Next-generation, high-throughput technologies have potential utility for providing more accurate blood biomarkers. Indeed, blood protein/peptide profiling may help to identify a novel biomarker for fibrosis, but the complex dynamic plasma protein range, issues regarding cost, variability in sample phenotype and depletion methods, technical reproducibility, and high-throughput capability still need to be addressed. MicroRNAs that regulate posttranscriptional gene expression appear to also regulate the fibrogenic cascade, and several have been proposed as potential markers of fibrosis in exosome-enriched fractionated serum.12 Epigenetic changes through DNA methylation also have a role in fibrosis and disease dynamics and might be potentially helpful. At this stage, clinical validation of such biomarkers that are detectable in blood or urine is still awaited. Long, noncoding RNA transcripts (>200 nucleotides) and genomic, metabolomic, and proteomic profiling of circulating extracellular vesicles (such as microvesicles, membrane particles, exosomes, or apoptotic bodies) implicated in disease pathogenesis and intercellular signaling represent novel approaches that hold promise for refined “liquid biopsy” biomarker development in the future.

Imaging Techniques

The introduction of liver stiffness measurement by using a variety of ultrasound (US)- or magnetic resonance (MR)-based techniques has been a major advance for the work-up of patients with CLD. Monodimensional US transient elastography (TE) (FibroScan; Echosens, Paris, France) is currently the most widely used and best-validated technique because it is a fast, safe, and reproducible procedure that can be performed at the bedside.13 TE more accurately detects cirrhosis (AUROC values, 0.87-0.98; correct classification ranging from 85% to 94%) than significant fibrosis (AUROC values, 0.75-0.93; correct classification from 57% to 90%).14 More recently, 2-dimensional elastography techniques embedded into conventional US machines (sonoelastography) have been introduced, such as Acoustic Radiation Force Impulse imaging (Acuson 2000 Virtual Touch Tissue Quantification; Siemens Healthcare, Erlangen, Germany) and ShearWave elastography (Aixplorer; Supersonic Imagine, Aix-en-Provence, France) (Table 3).15 Liver stiffness can be also measured by means of MR elastography using a modified phase-contrast method to image the propagation characteristics of the shear wave in the liver.16 It allows analysis of almost the entire liver, but MR elastography is thus far too costly and time-consuming to be used in routine practice.

In addition to static evaluation, liver stiffness can be used to identify patients with cirrhosis who are at risk of disease progression. However, the diagnostic accuracy of TE is too low for identification of patients with esophageal varices in clinical practice.17 More recently, spleen stiffness has received much attention as a potential better predictor than liver stiffness, for analysis of portal hypertension. However, results remain conflicting. Several novel imaging technologies, such as perfusion, MR diffusion-weighted imaging, intravoxel incoherent motion, or acoustic structure quantification are being developed, but their use remains limited thus far to research settings.

Noninvasive methods for fibrosis assessment are now well implanted in clinical practice. However, we must keep in mind that fibrosis is only one of the many histopathologic features among a spectrum of liver injury observed in CLD, and these features are assessed together in a comprehensive manner only by liver biopsy. Therefore, liver biopsy will stay in a hepatologist's armamentarium for many years.