Ultrasonographic fatty liver indicator (US-FLI): a reliable biomarker for non-invasive NAFLD stratification

Defined as liver steatosis occurring in the absence of competing etiologies of steatogenic chronic liver disease (CLD), nonalcoholic fatty liver disease (NAFLD) embraces bland steatosis through nonalcoholic steatohepatitis (NASH), which also features inflammation and hepatocellular ballooning^[1]. Of concern, global NAFLD burdens have increased since 1990^[2], accounting for direct and indirect financial costs.

Alarmingly, compared to individuals with NAFLD, NASH patients exhibit increased odds of developing cirrhosis and mortality owing to liver-related causes^[3], which supports the importance of differentiating NASH from steatosis. This differential diagnosis requires liver biopsy, although this procedure is invasive and not invariably safe, and thus cannot be used in epidemiological studies^[4]. Imaging techniques, such as nuclear magnetic resonance (NMR) proton density fat fraction, are deemed to be the surrogate reference standard for the evaluation of NAFLD^[5]. However, drawbacks of NMR imaging comprise high costs, limited sampling of hepatic regions, susceptibility to motion artifacts from the heart, time consumption, and uneven availability across various institutions and countries^[6]. Taken collectively, issues with liver biopsy and inconvenience with NMR make ultrasonographic techniques an excellent non-invasive imaging modality for use in the NAFLD arena^[7]. Point-of-Care ultrasound (POCUS) refers to the use of echography to diagnose problems wherever a patient is being treated, conducted by a non-radiologist physician who is directly involved in the patient's care^[8]. However, the application of POCUS in the field of NAFLD has been limited so far.

In 2012, Ballestri et al. conceived a novel ultrasound-based scoring system that they named ultrasonographic fatty liver indicator (US-FLI)^[9]. US-FLI was based on the ultrasound semeiotics as follows: (severity of) liver-kidney contrast, posterior attenuation (of ultrasound beam), (difficult) visualization of the diaphragm, vessel blurring, (difficult) visualization of gallbladder wall, and focal sparing^[9]. US-FLI could range from 2 to 8, and NAFLD was identified by a score ≥ 2. [Figures 1 and 2 reproduced, with permission from Ballestri et al.^[9]]. Interestingly, US-FLI was correlated with metabolic variables [homeostasis model (HOMA), insulin, serum uric acid, ferritin, alanine transaminase (ALT), and bilirubin. Moreover, US-FLI was also associated with some histological features [such as steatosis extent and nonalcoholic steatohepatitis (NASH), except for fibrosis]. At multivariate analysis, US-FLI independently predicted NASH (OR 2.236; P = 0.007) and a US-FLI < 4 excluded severe NASH at histology with a negative predictive value of 94%.

Figure 1. Illustration of US-FLI criteria and metrics: liver-kidney contrast, posterior attenuation of ultrasound beam and visualization of diaphragm. Absent -to – minimal liver-kidney contrast, which is normal, would receive a score = 0 (panel A). Conversely, the liver-kidney contrast of mild-moderate severity would be assigned a US-FLI score = 2 (panel B) and severe liver-kidney contrast would receive a US-FLI score = 3 (panel C). Absent posterior attenuation of ultrasound beam (PAUB) is normal and receives a US-FLI score = 0 (panel D). Conversely, PAUB present (highlighted with arrows) receives a US-FLI score = 1, although diaphragm remains visible, which is assigned a US-FLI score = 0 (panel E). Finally, impaired visualization of diaphragm (highlighted with arrows) owing to pronounced PAUB (panel F). These semiotics shown in panel F receive a US-FLI score = 1 each. Reproduced with permission from Ballestri et al.^[9]. US-FLI: Ultrasonographic fatty liver indicator.

Figure 2. Illustration of US-FLI criteria and metrics: vessels blurring, focal sparing, and visualization of gallbladder wall. Normal appearance, with optimal visualization, of contours and lumen (highlighted with arrow) of hepatic veins (panel A). This semeiotics would receive a US-FLI score = 0. Conversely, impaired visualization of contours and lumen of hepatic veins would be assigned a US-FLI score = 1 (panel B). Panel C shows normal appearance of walls of portal branches (highlighted with an arrow) which would receive a US-FLI score = 0 and normal visualization of the gallbladder wall (highlighted with a dotted arrow) which is assigned a US-FLI score = 0. Impaired visibility of walls of portal branches (highlighted with an arrow) would receive a US-FLI score = 1 (panel D). An area of focal sparing (highlighted with an arrow) would be assigned a US-FLI score = 1 (panel E). Impaired visualization of wall of gallbladder (highlighted with a dotted arrow) is assigned a US-FLI score = 1 (panel F). Reproduced with permission from Ballestri et al.^[9]. US-FLI: Ultrasonographic fatty liver indicator.

Sourianarayanane and McCullough have recently validated US-FLI as a marker of NASH in the setting of POCUS delivered by primary care providers^[10]. In their study, these authors found that US-FLI can reliably stratify NASH patients. This important study is only the last adjunct to a series of studies summarized in Table 1.

Taken collectively, the studies illustrated above^[7,9-13] offer a complete outlook of the points of strengths and weaknesses of US-FLI. The former include high reproducibility^[9,13], applicability to the pediatric population^[12] and to the point-of-care setting^[10], its strong association with metabolic derangements^[9,11,12], different grades of hepatic steatosis assessed with transient elastography^[6], and important liver histology endpoints, notably including its capacity to differentiate bland steatosis from true NASH^[7,9,10,11]. As a result, US-FLI outperforms the fatty liver index (FLI)^[13], a validated non-invasive biomarker proposed by Bedogni et al., to help in picking out those subjects to submit to hepatic ultrasonography scanning to confirm steatosis (and whose scope has now been expanded to many other areas)^[14] and represents an invaluable tool for triaging those patients in whom hepatic biopsy for suspected NASH is indicated. Coming to the limitations, US-FLI reportedly has limited accuracy among patient populations with obesity, given that these individuals often exhibit scores ≥ 5^[7]. Moreover, US-FLI was devised at a time when the dichotomy bland steatosis vs. NASH was deemed the key research and clinical question to address. However, over time, assessment of fibrosis has become comparatively more important given that fibrosis dictates the hepatic and extra-hepatic outcomes of NAFLD, and several biomarkers and algorithms are available to this end^[15,16].

With this eminently diagnostic backset, it is somewhat unexpected that US-FLI may also contain some clues to better understand NASH pathomechanics. For example, the finding that a difficult visualization of gallbladder wall is specific for NASH^[7] raises the possibility that accumulation of fat specifically either in the peri-cholecystic area or in the gallbladder wall may identify more aggressive liver histology. The vascular, histogenic or biochemical pathways underlying this phenomenon remain totally unexplored. However, it is possible that the hepato-cholecystic system behaves like a functional unit that is the target of severe metabolic derangements, as demonstrated, for example, by the association of NAFLD with gallstones. A recent umbrella review of 22 meta-analyses established that NAFLD carries an excess risk of gallstones^[17] and, conversely, gallstones are an independent risk factor for NAFLD^[18]. However, whether gallstones are also associated with more severe NAFLD remains controversial, with some studies supporting^[19] and others disconfirming this contention^[20]. Bringing this line of research further, authors have coined the definition of “fatty gallbladder disease” to identify increased gallbladder wall thickness and dysfunction of gallbladder motility, which are associated with NAFLD irrespective of gallstones^[21].

Based on the study by Sourianarayanane and McCullough^[10] as well as the others discussed here^[7,9,11-13], it may be concluded that research on US-FLI needs to be further developed. Several lines of advancements in research can be envisaged. One possibility is to combine US-FLI with simple anthropometric indices or routine liver tests. Moreover, rather than attributing a similar score of 1 to each semeiotics, a more sophisticated analysis may eventually conduce to give a higher score to certain semeiotics such as vessel blurring and gallbladder wall visualization. Additionally, US-FLI must be validated among the cholecystectomized individuals. From a pathogenic perspective, it will be important to understand the histopathological grounds associating difficult visualization of the gallbladder wall with NASH. We highlight that NAFLD is associated with extra-hepatic complications, including a high risk of cardiovascular disease and cancer^{[15,16,22-24]}, although no data on the value of US-FLI in predicting such outcomes are available. US-FLI may also be combined with the measurement of spleen diameter and volume to detect the presence and severity of NAFLD^[25]. Finally, it will be important to understand the accuracy of US-FLI among overweight patients and in the detection of the so-called “at-risk NASH”, i.e., those patients with NAFLD activity score (NAS) ≥ 4, and fibrosis stage ≥ 2^[26].