Liver stiffness measurements in patients with metabolic dysfunction-associated steatotic liver disease: Updates on the method effectiveness and perspectives

Abstract

Metabolic dysfunction-associated steatotic liver disease (MASLD) is the most widespread chronic liver disease signified by serious life-threatening conditions. The prevalence of MASLD increases along the growing prevalence in obesity and metabolic syndrome. To minimize costs and complications, non-invasive diagnostic tools, including transient elastography (TE), were introduced for assessment of MASLD. TE measures liver stiffness (LS), a clinical marker for the diagnosis of liver fibrosis and cirrhosis. LS measurements are based on ultrasound wave imaging and quantification. Vibration-controlled TE, including FibroScan^®, is commonly used TE methods which can accurately identify the degree of liver fibrosis and cirrhosis progression. TE was reported to predict the progression towards hepatocellular carcinoma, portal hypertension, and varices. However, the accuracy of LS diagnostics alone in patients with MASLD remains controversial. TE measurements have several limitations, including inadequate precision due to focal liver lesions, cholestasis, inflammation, and other pathological and anatomical factors which can lead to the stiffness variability. Overestimations of TE readings were reported in obese patients with body mass index (BMI) over 30 kg/m², and older patients with ascites, diabetes, or hypertension. Not all MASLD patients have high BMI. The prevalence of obesity among MASLD patients varies worldwide, indicating the urgent need for comprehensive diagnostic tools. In patients with MASLD, improved diagnostic accuracy has been demonstrated by combining LS measurements with other blood test-based scores and simple clinical parameters (agile scores based on age, sex, platelet count, aminotransferases, and diabetes). This study reviews the limitations of TE-based diagnostics and discusses the combined scoring algorithm. In conclusion, the sequence of LS measurements along assessment of other important clinical markers is an effective, low-cost, reliable tool to identify and monitor fibrosis progression in MASLD.

Key Words: Non-alcoholic fatty liver disease; Liver fibrosis; Metabolic syndrome; Metabolic dysfunction-associated steatotic liver disease; Liver stiffness measurement; Transient elastography

Core Tip: Clinical diagnostics of metabolic dysfunction-associated steatotic liver disease (MASLD) relies on non-invasive diagnostic tools, such as transient elastography (TE)/FibroScan^®. TE measures liver stiffness, an important clinical marker of liver fibrosis and cirrhosis. However, TE measurements have limited accuracy in obese patients. This study reviews the limitations of TE-based diagnostics and discusses the combined scoring algorithm. Data review indicates that in patients with MASLD reliable diagnostic accuracy can be achieved by combining FibroScan^® measurements with various blood test-based scores, including agile and fibrosis score indicators.

INTRODUCTION

Representing substantial health system burden, chronic liver disease (CLD) is associated with significant morbidities and mortality[1]. CLD is manifested by a progressive accumulation of fibrous tissue and deterioration of liver functions, leading to fibrosis and cirrhosis[1-3]. Alcoholic liver disease, viral hepatitis [hepatitis B virus (HBV) and hepatitis C virus (HCV)], non-alcoholic fatty liver disease (NAFLD)/non-alcoholic steatohepatitis [currently renamed as metabolic dysfunction-associated steatotic liver disease (MASLD)][4,5], and hemochromatosis[6] signify the complexity of CLD. The severe clinical form of MASLD is named metabolic dysfunction-associated steatohepatitis (MASH) (new terminology)[6,7]. New names MASLD and MASH provide a greater emphasis on the function of metabolic factors (such as obesity and diabetes) in the pathology of fatty liver. CLD progression towards decompensated cirrhosis is often accompanied by jaundice, ascites, hepatic encephalopathy, hepatorenal syndrome, or variceal haemorrhage (bleeding)[1,7-9]. Furthermore, cirrhosis may lead to development of hepatocellular carcinoma (HCC)[10].

Early detection and monitoring of advanced fibrosis and cirrhosis are critical for prevention of CLD complications and reduction of mortality. Prior to non-invasive testing, a liver biopsy is used to differentiate cirrhosis from lower fibrosis grades. Liver biopsy is an invasive and expensive procedure with minor risk of sampling error, interobserver variability, complications, and will not meet the diagnostic demands of highly prevalent liver disease. However, liver biopsy is considered as a standard criterion for advanced fibrosis assessment[11], although it was associated with the surgical complications. Cost-comparing study indicated reliability and advantages of non-invasive methods for assessment of liver fibrosis[12]. The alternative diagnostics relies on specific blood markers and radiologic investigations, including transient elastography (TE) or FibroScan^®. TE is a non-invasive, reliable, simple, and less costly method[13,14]. Other types of liver stiffness (LS) measurements include point shear wave elastography (pSWE) and two-dimensional (2D)-shear wave elastography (SWE) which combine imaging with ultrasound techniques[14,15]. The pSWE and 2D-SWE methods were developed and introduced after FibroScan^® and remain less validated method compared to TE. Magnetic resonance elastography (MRE) has been used for accurate assessment of liver fibrosis using specific tissue mechanical parameters[16], although MRE limitations has been noted[17]. MRE is not reviewed in this study as it is costly and not widely available technique[17,18].

FibroScan^® device is often used to determine liver steatosis according to controlled attenuation parameter (CAP) which correctly indicates the level of liver fat in vast majority of patients with CLD[13,14]. TE is the most used test for the assessment of liver fibrosis[13,15,18]. However, LS diagnostic accuracy remains to be confirmed in patients with MASLD. The incidence of MASLD is growing, representing serious health and financial challenges worldwide[19]. The application of TE/FibroScan^® for assessment of fibrosis in obese patients and/or patients with diagnosed metabolic syndrome is currently being in the clinical spotlight and warrants further investigations. This study aims to discuss benefits and limitations of TE/FibroScan^® assessment of the liver steatosis in patients with MASLD. TE technique and impediments associated with the presence of fat and pro-inflammatory conditions in CLD patients will be examined. To define potential areas for future improvement of CLD diagnostics, we critically evaluated perspectives of TE use as combined approach with other clinical methods of hepatological assessment.

GLOBAL PROGRESSION (PANDEMIA) OF HEPATIC STEATOSIS, ITS CAUSES AND ASSOCIATED METABOLIC DYSFUNCTIONS

MASLD prevalence has increased dramatically over the last few decades[20]. Extensive data analysis, involving 8515431 patients (22 countries), demonstrated the global prevalence of MASLD in 2016[21]. Currently, the incidence of MASLD (hepatic steatosis) has reached the pandemic (30%) threshold in some countries and continues to rise[19]. Socioeconomic and genetic factors may contribute the geographic differences in the prevalence of MASLD. Several regions demonstrated the higher prevalence of MASLD (Latin America, the Middle East, North Africa, and South Asia), while Western Europe and the Asia-Pacific countries have shown the lower prevalence rate for this CLD[22]. Obesity, dyslipidemia, aging, and type 2 diabetes mellitus (T2DM) are top contributing risk factors for MASLD (Figure 1)[23,24]. Hepatic steatosis was shown to trigger the development HCC, accenting the key role of early diagnostic for MASLD management[4,5,25].

Figure 1 The risk factors and biological mechanisms of metabolic dysfunction-associated steatotic liver disease. AMPK: AMP-activated protein kinase; CPT1: Carnitine palmitoyltransferase I; ER: Endoplasmic reticulum; HCC: Hepatocellular carcinoma; IL: Interleukin; JNK: C-Jun N-terminal kinase; MASLD: Metabolic dysfunction-associated steatotic liver disease; PPR-α: Peroxisome proliferator-activated receptor-alpha; ROS: Reactive oxygen species; Sirt: Sirtuin; Stat3: Signal transducer and activator of transcription 3; TNF-α: Tumor necrosis factor-alpha.

Although often linked to obesity, MASLD is currently used as the terminology for patients with hepatic steatosis who are not-necessarily obese (defined mostly by waist circumference), but represented with two or more metabolic abnormalities, such as high level of triglycerides [and/or low level of plasma high-density lipoprotein (HDL)-cholesterol][26], high blood pressure, high fasting glucose levels, insulin resistance, or high level of C reactive proteins (reviewed elsewhere)[27,28]. Confirmed abnormal liver enzyme profile is another obligatory part for MASLD diagnosis. Prolonged (> 6 months) elevations of serum alanine aminotransferase (ALT), aspartate aminotransferase (AST), or gamma-glutamyl transferase (not linked to other diseases) along with specific signs of metabolic syndrome are required diagnostic criteria for MASLD[22,23,27]. In MASLD patients, fibrosis stages were defined as F0, F1, F2, F3, and F4 grades, where F3 and F4 represent advanced fibrosis[29,30]. High sensitivity (85%) and specificity (82%) of TE/FibroScan^® was reported in MASLD patients with F3. Even better sensitivity (92%) and specificity (92%) were found using meta-analysis of data for MASLD patients with F4[30]. TE-based diagnostics of F1 and F2 fibrosis grades demonstrated lower accuracy (79% sensitivity and 75% specificity)[31].

A longitudinal study which used liver biopsy indicators (a more expensive approach) reported that one third of MASLD patients progressed to the worse fibrosis stage within a 3-year interval[32], indicating a need for continuous screening. Less-costly TE assessment demonstrated high accuracy in initial assessments of liver fibrosis MASLD patients, confirming the suitability of this method for MASLD monitoring[15,33]. Changes towards higher LS (defined by sequential TE measurements) were also associated with greater risk of liver-related mortality in MASLD patients[34]. Many studies confirmed that TE-based sequential assessment of fibrosis can be used as a reliable predictor of MASLD-linked complications and mortality[35-37].

Adapted for steatosis quantification, CAP [measured in decibels per meter (dB/m)] is calculated using TE radiofrequency data. However, in MASLD patients with body mass index > 30 kg/m², older patients with ascites, diabetes, or hypertension, several studies have shown a false increase in fibrosis due to higher steatosis-induced stiffness[38]. High BMI was shown to influence CAP numbers[39,40]. In obese MASLD patients, the correct TE data acquisition is impacted by the presence of fat tissue, inflammation, and other factors (see subsection 5)[38,41,42]. Therefore, for fibrosis assessment in overweight patients, several multistep approaches were recommended (discussed in subsections 6 and 7), although some studies indicated that CAP is suitable for identification of even mild steatosis conditions in MASLD[39,40].

Notably, nearly half of all MASLD patients (obese) requires additional TE method adjustments and multistep measures for fibrosis assessment[43]. Accumulated data indicated that about 45% of MASLD patients have “lean” (normal, non-obese), BMI (BMI < 23 kg/m²: Below overweight or obesity threshold) and closely associated with T2DM, or other metabolic abnormalities[39,44,45]. Currently, there is limited research on non-obese MASLD, and its pathogenesis and treatment methods remain unclear. It has been reported that non-obese type MASLD patients are at a higher risk of progressing to advanced liver fibrosis[46]. This observation warrants further investigations. It has been reported recently that in non-obese MASLD patients, there are elevated levels of urea nitrogen, uric acid, triglycerides, and fasting blood glucose (compared to normal controls)[47]. Alternatively, HDL and apolipoprotein A1 were downregulated[47].

Interestingly, the products of fat metabolism, free fatty acids (FFA) (such as palmitic acid, oleic acid, linoleic acid, and arachidonic acid) were also significantly elevated in non-obese MASLD patients[47,48]. Disrupted balance between specific circulating and intracellular FFAs may lead to the accumulation of fat not only in liver (lipotoxicity), but also in cardiovascular system[49]. High FFA indicates liver cell damage. It remains to be investigated how increased FFA levels may impact the TE/FibroScan^® data acquisition and whether TE maybe used to assess oxidative liver damage and inflammation. Nevertheless, non-obese MASLD patients may directly benefit from TE-based monitoring. Less expensive and easy-accessible TE diagnostics in non-obese MASLD patients is not impacted by the presence of fat tissue, although pro-inflammatory conditions may (potentially) influence TE measurements[50]. The pathogenesis of liver steatosis in non-obese MASLD patients is most likely multifactorial and may be provoked by several metabolic dysfunctions, including T2DM.

The relationship between T2DM and MASLD are complex and not completely understood[39]. Higher glucose level in circulation was associated with increased LS[51]. It is well known that one of the main liver functions is to balance and preserve normal blood glucose[52]. Active glucose metabolism, glucose storage, and endogenous glucose production from glycogen (gluconeogenesis) were observed in liver and muscles during normal ontogenesis[53]. Many studies indicated that the prevalence of liver steatosis is very high among patients with T2DM[39,54]. In patients with advanced liver fibrosis, liver cells are no longer efficient in glucose metabolism, leading to increased blood glucose (the main T2DM marker)[55]. Notably, less insulin is absorbed by hepatocytes from the circulation (chronic hyperinsulinemia, insulin resistance) in MASLD patients. Therefore, T2DM conditions could be both sequential cause (a trigger) of the fatty liver disease or MASLD-induced complication. Epidemiological data confirmed that MASLD patients are more likely to progress towards T2DM[56].

Persistent immune disturbances can also lead to aberrant liver cell metabolism, resulting in the accumulation of lipids (FFA and cholesterol) in liver cells. Excessive lipid accumulation in the liver (lipotoxicity) is marked by mitochondrial dysfunction, increased reactive oxygen species levels, and chronic inflammation[57], which, in turn, contributes progression towards liver steatosis and, potentially, liver cancer[58]. The recent study demonstrated that approximately 30% of chronic HBV (viral inflammation) patients with MASLD have a high level of fibrosis progression[59]. A strong positive correlation was found between higher LS and hepatic portal hypertension in HCV patients[60]. Indeed, LS values increase gradually alongside the increment in liver cirrhosis and portal hypertension[61]. Thus, aggravated inflammation cascade stimulates MASLD progression.

Distinct immune microenvironment of MASLD was shown to accompany the progression of liver steatosis and fibrosis (Figure 1)[62-64]. Macrophage-derived pro-inflammatory cytokines (such as tumor necrosis factor-alpha and interleukin-1β) were shown to promote lipid deposition in hepatocytes[58,65]. Infection or diet/lifestyle—triggered dysregulation in lipid metabolism stimulated accumulation of FFAs and modified immune cell landscape[66]. The observed changes included exhaustion or decline of Kupffer cell (KC) (liver resident macrophages) functions[66]. KCs are not only key regulators of inflammation in liver during normal ontogenesis, but also contribute fibrotic signalling cascades in MASLD[44,67,68]. Aside from KCs, atypical T cell activation and exhaustion profiles were reported in MASLD patient and rodent models[69,70]. The detected T cells failed to demonstrate the immune surveillance functions, although liver tissue damage was observed[70,71]. The diverse characteristics of liver microenvironment in patients with steatosis/fibrosis suggest the high level of heterogeneity of MASLD.

Considering the spectrum of contributing pathologies and various stages of steatosis and cirrhosis, MAFLD is highly heterogenous disease. The mechanisms for MAFLD development, progression stages, and response to therapy are also likely to be different[53,55-57]. Therefore, it may require complex approach in diagnostics and monitoring. That is why combined approach is more likely to be the most effective (discussed in sub-section 6 of this review) and may cover the larger spectrum of conditions involved.

NON-INVASIVE METHODS OF LS ASSESSMENT: FOCUS ON TE

Increased collagen content is the major indicator of growing tissue stiffness which can be manifested as fibrosis [a deposition of fibrous (dense) material]. Elastography (ultrasound-based approach) techniques is used to measure LS and estimate the level of fibrosis in liver tissue[57,72]. FibroScan^® device (Echosens, Paris, France) is commonly used for TE assessment. TE/FibroScan^® method works on two physical principles: (1) Strain displacement; and (2) Shear wave imaging and quantification. The latter includes pSWE. It measures the velocity of low frequency (50 Hz) elastic shear waves propagating through the liver. The stiffer the tissue, the faster the elastic shear wave propagates. TE can generate quantitative data presented as elasticity (kPa)[73].

The successful testing of FibroScan^® for assessment of LS was reported since 2003[74-76]. TE helps to differentiate advanced steatohepatitis and hepatic fibrosis from simple (lower grade) steatosis. Derived in association with LS, CAP is a reliable marker for quantification of liver fat. CAP demonstrated promising area under the receiver operating characteristic curve (AUROC) > 0.80 and high sensitivity (80%) and specificity (80%)[77,78]. However, fibrosis (collagen fiber deposition) develops gradually, not linearly[38,39]. At the earlier stages, fibers are deposited with a slower increment. Alternatively, during advanced CLD, deposition of fibers happens with higher speed (exponentially). Therefore, the accuracy of TE measurements and reference standards can be negatively impacted by these non-linear differences and random accumulation of collagen[74]. Consequently, TE assessment is combined with other scores to improve the diagnostic accuracy[79-81].

Other non-invasive tests, including the Fibrosis-4 index (FIB-4)[79] and the Enhanced Liver Fibrosis test (ELF^™)[80], are often used together to increase the power of assessment[81]. When used alone, specificity and sensitivity of FIB-4 or ELF^™ may fluctuate according to the set of cut-offs values[82]. In MASLD patients, FIB-4 with ELF^™ before FibroScan^® assessment were recommended for the diagnosis of early and mild fibrosis[42,83,84]. However, every method has its own clinical limitations. For instance, in older patients, a high rate of false positive scores were generated using FIB-4[85]. To avoid mistakes, the cut-offs rules were introduced to fit different clinical situations, although the efficiency of this approach remains to be confirmed in larger clinical studies. Multiple investigations indicated that TE is very reliable methods for diagnosing advanced (3-4 grade) fibrosis stages, compared to early (1-2 grade) stages in patients with CLD[76,86]. The differentiation between several early fibrosis stages can be challenging for nearly all indicated methods.

CLINICAL APPLICATION OF TE: ADVANTAGES, LIMITATIONS, AND VARIATIONS IN METHODOLOGY

Two strongest TE advantages include the acquisition of continuous (quantitative) data and low-cost operation (easy in operation ultrasound-based technique). Comparing to TE, liver biopsy is more expensive procedure which is used to establish an important immunohistochemistry (IHC)-based (semiquantitative) clinical disease staging. A side from pathological characteristics, IHC/liver biopsy diagnostics is prone to biases associated with the location and quantity of fibrosis, because diseased tissue may not be evenly distributed in the liver. However, LS assessment by TE is also influenced (confounded) by tissue pathophysiology, such as portal hypertension, biliary obstruction, local inflammation, and passive venous congestion (thrombosis). Furthermore, LS assessment is recommended to be done in the fasting state, as liver physical condition is affected by postprandial status (Figure 2)[87]. Therefore, the diagnostic accuracy can be reached in some cases only after careful consideration of all measured characteristics, including liver imaging, liver function tests, and other available pathology markers. The role of coexisting confounders (other diseases, including metabolic dysfunctions) should be carefully evaluated. The impact of various pathologies on TE-based (and not only) assessment of liver fibrosis remains largely under-investigated.

Figure 2 Factors that define transient elastography diagnostic accuracy. ALT: Alanine aminotransferase; AST: Aspartate aminotransferase; HCC: Hepatocellular carcinoma; TE: Transient elastography.

To resolve complications associated with anatomical differences and body types, the FibroScan^® devices are fitted with several different transducers (S, M, and XL). To acquire reliable data in overweight patients, it is required to use the XL transducer[88], although it may not be able to solve all diagnostic problems. To further increase the method sensitivity, CAP is combined with blood test data (such as serum transaminases) to generate a FibroScan^®-AST (FAST) score (Figure 2)[76,83,89]. According to the meta-analysis, FibroScan^® method allows to detect F3+ stage fibrosis with high sensitivity (85%) and specificity (82%). The accuracy of this method in detection of cirrhosis is even higher (sensitivity and specificity = 92%)[90]. However, the accuracy gets lower in patients with BMI > 35 kg/m², representing a serious impediment for proper diagnostics[91-93]. In patients with morbid obesity (bariatric surgery cohort), one recent observational study reported that LS/CAP measurements using FibroScan^® (XL probe) were technically successful in only 59% of cases[94], indicating only moderate achievements with this technique. Notable, data collected in bariatric cohorts is controversial, as another study with 190 patients (BMI: 40.2 kg/m² ± 6.6 kg/m²) reported better FibroScan^®/CAP achievements with technically reliable measurements in 88% of patients[95]. Excellent data were reported in smaller cohort with 126 obese patients (BMI ≥ 30 kg/m²) in Turkey (97% of successful measurements)[96]. The study with Brazilian MASLD cohort (287 patients, BMI ≥ 32 kg/m²), which used combined FAST score, also indicated reliable diagnostic outcome[97]. FAST score was also evaluated retrospectively and indicated similar (reliable high sensitivity) findings in Japanese cohort[98]. All recent studies indicated that unsuccessful FibroScan^® measurements in MASLD cohorts were associated with increased body weight and high BMI (> 40 kg/m²)[94,95,97].

TE/imaging of liver fibrosis can be acquired using not only ultrasound, but also magnetic resonance (MR) techniques. MR-based elastography is a very reliable, but expensive method which was reviewed recently[16]. MR technique was successfully used to distinguish MASLD from MASH[17,18,99]. However, lower cost and the easiness of the ultrasound TE device application led to the preferential use of this method in clinics. Ultrasound-based TE is represented by several approaches, such as pSWE and 2D-SWE, which can be measured using devices from various manufacturers. All of them seem to produce similar outcomes[86,89]. For instance, a recent study compared six different ultrasound-adopting TE systems[100]. Reliable correlation coefficients and similar diagnostic conclusions were reported in human patients, suggesting a good agreement across the compared systems[100]. However, the quantitative TE measures somewhat different characteristic of liver pathology which cannot be easily linked to IHC staging, indicating a need for large matching IHC-TE investigations[18,101].

OTHER TE LIMITATIONS AND ASSOCIATED DIAGNOSTIC IMPEDIMENTS

The difference in measured parameters between TE and IHC-based diagnostics leads to a limited usability of TE to predict histopathological stages of CLD, although the problem is not associated with the TE accuracy. Ultrasound-based techniques reflect exactly the tissue structure and may be impacted mostly by the level of operator experience, which is minimized during training. The difficulties with liver tissue visualization and confounding anatomical structures were mostly resolved[16]. However, the calculated fibrosis staging is partially impacted by low sampling volume for standards, the high level of natural tissue variability, and insufficient correlation of liver histopathology with CLD severity/outcome. These impediments may complicate TE clinical interpretation in some CLD patients.

Physical characteristics of liver tissue influence TE data and interpretations. The velocity of shear wave propagation is defined not only by the severity of fibrosis, but also by several other factors. For the ideal propagation of the signal, the medium should be homogeneous, has uniform density and high level of elasticity. Obvious absence of these conditions in the human liver makes TE measurements prone to errors. Therefore, increased viscosity and tissue density impact TE imaging. Increased viscosity was observed in patients with biliary obstruction, acute hepatitis[102], and congested liver[103]. Higher tissue density was reported in patients with inflamed liver tissue and HCC (including liver tissue necrosis)[104], MASLD[105], cholestasis[106], and hepatic storage disease[107]. To disentangle LS measurement problem, TE guidelines have been developed by American Gastroenterological Association (AGA)[76] and Baveno VI groups[108]. The guidelines suggested LS cut-off points which were designed not only to help with cirrhosis assessment, but also to estimate the need for endoscopy, while excluding the compensated advanced CLD (cACLD). Extending the Baveno VI group-based guidelines, the World Federation of Ultrasound in Medicine and Biology introduced the “Rule of Five” protocol[109]. All limitations are described in Table 1 below[13,15,16,19,34,37,38,74,87,95,101,110-133].

Table 1 Factors that impacts transient elastography accuracy and controlled attenuation parameter in chronic liver disease diagnostics.

Ref.	TE accuracy and controlled attenuation parameter-impacting factors	Problem solving recommendations
Coco et al[110], Arena et al[111], Sasso et al[112], Kumar et al[113], Jung et al[114], Mi et al[115]	Hepatic inflammation (including viral loading), liver condition in general (aspartate aminotransferase, alanine aminotransferase levels and activities)	To recognize inflammation, it was recommended to apply damping ratio and shear loss modulus. Apply dispersion slope, attenuation coefficient, and shear wave speed imaging markers. Combined/multivariate tests (blood analysis, liver biopsy)
Sagir et al[116] Millonig et al[117], Ozturk et al[118], Trifanov et al[119], Huang et al[120]	Hepatic congestion (venous pressure, portal vein thrombosis, bilirubin, protein accumulation) and liver damage (including intoxication by alcohol, heavy metals etc.)	CVD-related tests, administer diuretics, use histopathology as reference standard, combined/multivariate tests (blood analysis, liver biopsy) are yet to be tested and characterized
Petta et al[34], Younossi et al[37], Yin et al[87], Yang et al[95], Cassinotto et al[101], Millonig et al[121], Sharma et al[122], Ali et al[123]	Cholestatic liver diseases (steatosis)	TE probe should be selected according to BMI (less failure rate with XL probe in high BMI patients), proper cut-off values should be defined, incorporation of hemoglobin A1c and alkaline phosphatase with liver-stiffness measurement improves accuracy in detecting significant fibrosis, better IHC matching (or different, such as blood marker-based) should be designed
Huang et al[19]	Focal liver lesions/HCC	N/A, HCC protocol should be followed
Blank et al[124], Nogami et al[125]	Distance between skin and liver; peripheral or abdominal fat (obesity/high BMI)	Automatic selection of the probe, the choice of XL or M probe should be standardized using novel SmartExam computational method
Lin et al[126]	Ascites	N/A, a proper disease-related protocol should be activated
Ozturk et al[16]	Anatomical factors/musculoskeletal deformity, age/sex	The FibroScan® system uses A and transmission metasurface mode maps which guide the operators to find the ideal location in the liver tissue. Regular calibration is reqiured, correct choice of proper controls
Loustaud-Ratti et al[127], Lanzi et al[128]	Amyloid deposition in liver	Reversing the ligation of the bile duct
Mederacke et al[129], Arena et al[130]	Fasting glucose (diabetes)/recent food ingestion	At least 4 hours fasting before procedure
Barr et al[15], Pennisi et al[131], Boeckmans et al[132]	Heart condition, beta-blockers, total cholesterol/triglycerides, intense physical exercise	It is recommended to apply heart disease case-finding strategies, CVD-related tests, proper cut-off values should be defined for this category of patients
Lemmer et al[13], Barr et al[15], Boursier et al[38]	Respiratory movements	Patients are asked to hold the breath and minimize movements. Measurements are repeated up to 3 and more times
Bassal et al[74], Hudson et al[133]	Operator-related biases	Regular operator training, several repeated readings per a patient to achieve the average reading value and minimize the measurement variability

BMI: Body mass index; CVD: Cardiovascular diseases; HCC: Hepatocellular carcinoma; N/A: Not applicable; TE: Transient elastography.

Later these guidelines were re-visited; and the range was tighter adjusted. The new rules were introduced by Society of Radiologists in Ultrasound in 2020. Accordingly, the current framework for TE data assessment is as follows: (1) Normal: LS ≤ 5 kPa; (2) Not cACLD: LS < 9 kPa (given the absence of other CLD clinical signs); (3) Suggestive cACLD: 9 kPa > LS < 13 kPa (further testing is required); (4) Likely cACLD: LS > 13 kPa; and (5) Significant portal hypertension: LS > 25 kPa[14,15]. These (current) rules are designed to guide the clinical decision, helping to avoid unnecessary and expensive endoscopic investigations. However, IHC/histopathology remains the main evidence for final diagnosis in many advanced CLD cases.

All ultrasound clinical techniques are biased by the level of operator training and experience. The operator-related biases were demonstrated in Hudson et al[133] study which estimated the accuracy of TE measurements in the same liver segments by different professionals. A decade ago, to reach the required quality of the image and measurement, it was necessary to repeat the probe reading (interquartile range of the median value < 30%) ten times in 1 patient[133,134]. This requirement significantly extended the time of assessment per a patient (although, commonly, only 3 repeats are required). Aside from operator-related problems, various patient-associated factors may impact TE imaging. The reflection of the sound (reverberation) from the internal tissue elements was associated with appearance of natural artifacts. For instance, limited penetration of the sound and resulting faulty images were linked to interaction of the sound with ribs (ribs shadow), vessel structure/content-related artifacts, the liver capsule, high body wall thickness (fat deposition), and patient’s motion[134]. To standardize the image acquisition and alleviate the variability, the imaging protocol was amended by Quantitative Imaging Biomarkers Alliance[134,135]. TE ultrasound systems should be checked for accuracy and calibrated periodically.

Considering the patient structure and composition variability, the main limitation of TE technique was associated with the increased inaccuracy of data in patients with high BMI (overweight/obese and metabolic syndrome diagnosed cases)[123,135]. In obese patients, the formation of thickened subcutaneous fat tissue leads to the sound attenuation, limited wave induction, and reduced or obstructed detection of sound/wave propagation. This is also results in the failure to define shear wave speed or estimated Young’s modulus values for specific image pixels (inadequate TE color filling)[16]. Sometimes, focal steatosis can be mistakenly identified as solid lesions, while other pathologies can mimic MASLD. For instance, localized accumulation of fat in liver can be often associated with vascular anomalies, use of specific steatogenic drugs, inherited metabolic disorders, or HCC[7,91,136].

Nevertheless, very high reliable AUROC values (used to distinguish between disease and non-disease) were reported for TE application in MASLD patients[137-139]. For instance, in patients with advanced fibrosis (F3-F4) (452 patients, MASLD was confirmed by biopsy/IHC), AUROC value for TE was 0.83[34,38,79]. Good LS accuracy for diagnosis of steatosis grades 2 and 3 was reported by Eddowes et al[137] (AUROC of 0.80). Despite this, recent meta-analyses demonstrated that higher grades of liver steatosis (MASLD/MASH/steatosis stages with liver inflammation) could not be differentiated properly by TE in patients with morbid obesity[139]. IHC-based steatosis grades did not always match well with CAP data[113,116,125]. For comparison, effective TE diagnostics of steatosis was reported in patients with viral hepatitis[50,129,130,136], but not in patients with MASLD[121,140]. Therefore, different CAP cutoffs were introduced for patients with high BMI, diabetes, high liver enzymes (AST), and sex (steroid hormones level). It has been suggested that during potential MASLD screening in obese patients, CAP values should be studied and adjusted/compared using several reference methods (not only the traditional IHC reference standard)[16,118,140]. TE adjustment (calibration and addition of other scores) may be required to achieve higher TE sensitivity in all patients with various metabolic pathologies[38,126,127,140].

SEQUENTIAL TESTING AND COMBINED SCORING METHODS/COMBINED APPROACH TO FIBROSIS GRADING

TE/FibroScan^® was designed to assess liver elasticity and possible development of liver fibrosis. This non-invasive assessment of liver steatosis and fibrosis is important for risk-stratification in patients with MASLD. Moreover, it is very useful to monitor disease trajectory and provide re-evaluation of liver conditions over time, particularly after therapies. Therefore, non-invasive TE assessment guidelines were set to the highest standards by AGA and other professional organizations, which currently develop and adjust the protocol for TE diagnostic procedures and disease monitoring steps[76,135]. However, MASLD encompasses a spectrum of metabolic pathologies which require a multidisciplinary diagnostic approach. Consequently, multicomponent investigation (diagnostics) method should be adapted for liver assessment in patients with MASLD[47,91,141]. Sequential combination of tests (multistep approach) allows to increase the predictive values for TE[141-143].

Currently available non-invasive tests, including TE/FibroScan^® demonstrated suboptimal values for sensitivity and specificity in obese MASLD patients in many studies[126,135,141,144], confirming a need for sequential combinations of assessments. Currently available serum biomarkers for recognition of low-grade steatosis from advanced disease (MASH), such as ALT-levels and cytokeratin-18 fragments, were also reported as suboptimal[37]. However, better sensitivity and specificity was reported for sequential combined tests. For instance, the Bordeaux algorithm suggested a combination of tests which includes AST-to-platelet ratio index (APRI) and FibroTest (FibroScan^®) plan[135,138,144,145]. The FAST score was validated globally and demonstrated a very good functional reliability (AUROC ranging from 0.74 to 0.95 in different cohorts)[76,139,140]. Another successful diagnostic approach recommended to calculate FIB-4 and APRI for the initial assessment, exclude the patients with unlikely fibrosis, and focus on those patients who has FIB-4 > 1.5. Those patients should be monitored using TE/FibroScan^® and/or other relevant methods[50]. The time of initial assessment is the critical point when clinicians can identify a patient at the increased risk and define the appropriate pathway for the disease monitoring and treatment, minimizing the potential risk of disease-related complications.

Modified APRI was introduced recently as a better diagnostic parameter, named NAFLD fibrosis index (NFI), for assessment of fibrosis in MASLD patients[118,146]. NFI formula combines 8 indicators, including age, platelet count, postprandial plasma glucose level, conjugated bilirubin, ALT, AST, total iron binding capacity, and the level of parathyroid hormone[138,141]. Interestingly, for the identification of advanced fibrosis NFI demonstrated improved diagnostic accuracy, compared to various clinical scoring formulas (FIB-4, APRI, and BMI-AST-to-ALT ratio-diabetes mellitus). However, the NFI performance characteristics did not exceed scoring effectiveness of FibroScan^®[138]. Another reliable method to increase diagnostic accuracy is to combine TE with several serum-based indicators/scores (such as hemoglobin A1c and alkaline phosphatase values)[123]. Blood serum-based and clinical characteristics score was constructed recently using the triglyceride-glucose index[147]. The index was effectively used to identify MASLD[147], but it should be combined with TE to score fibrosis. However, sequential TE-based monitoring of therapeutic interventions in MASLD and MASH remains under investigated.

Recently, a novel algorithm for assessment and monitoring of patients with MASLD has been suggested[143]. Using retrospective data analysis, it was demonstrated that FibroScan^® provides additional information and complement the FIB-4 scores. Best practice advice for patients with MASLD and advanced cirrhosis/fibrosis indicated that serial FibroScan^® assessments combined with blood tests should be implemented[143]. The study suggested that a FIB-4 < 1.3 should be used as a negative indicator for hepatic fibrosis, while FIB-4 > 1.3 should be combined with 2 or more non-invasive assessments (serum biomarkers and imaging-based scores) for further disease staging and risk stratification. To optimize prediction of the disease grade and identify patients with advanced fibrosis, the study suggested using clinical data (such as physical examination, endoscopic, and biochemical reports) and adapt periodical LS monitoring using TE/FibroScan^® method[143].

Other promising scoring tests for MASLD include agile 3+ and agile 4 which were effective in predicting HCC, HCC/cirrhosis-related death, various hepatic decompensation (ascites, variceal hemorrhage, hepatic encephalopathy), or progression towards liver transplantation[126]. Agile scoring was tested in a large cohort study which included 16603 patients with MASLD in several countries (the United States, Europe, and Asia)[126]. Furthermore, 'liquid' biomarkers were recently reported to be successful for diagnostics of liver fibrogenesis. The test is based on the scoring of collagen-derived markers [N-terminal propeptide of collagen (PRO)-C3 or PRO-C6]. The method was defined as 'multi-omics' technology associated with testing of proteomics and microRNAs levels. The “liquid” biopsy test aims to estimate intrahepatic disease level and can be compared to the liver biopsy/IHC. For comparison, FibroScan^® and other ultrasound-based methods (such as MR-based elastography and proton density fat fractioning), were defined as “dry” testing methods and demonstrated very high level of accuracy in the identification of patients at the highest risk of steatosis and fibrosis[148]. However, proper clinical testing is warranted to assess and compare the effectiveness of novel tests alone and in combinations with FibroScan^®. To improve MASLD management, the complimentary use of “liquid” and “dry” biomarkers should be adjusted according to the clinical indications.

CONCLUSION

Considering the complexity of confounding factors and wide spectrum of MASLD, a prospective artificial intelligence assistance should be sought for the choice of proper protocols, risk stratification, and assessment of prognosis and treatment choices.