Page 71 - Read Online

Ballestri et al. Metab Target Organ Damage 2023;3:13 https://dx.doi.org/10.20517/mtod.2023.21 Page 3 of 7

Table 1. Principal published studies on US-FLI from 2012 to 2023
Author Method Findings Conclusion
Ballestri et al. [9] 53 individuals were submitted to both US-FLI was correlated with metabolic parameters, This study was the first to
US and LB associated with hepatic histology. US-FLI, propose US-FLI and validate
independently predicted NASH (OR 2.236; P = it vs. liver histology in
0.007) and US-FLI < 4 ruled out severe NASH (NPV NAFLD/NASH arena
= 94%)
Ballestri et al. [11] 352 patients were submitted to both In patients with NAFLD: (A) US-FLI ≥ 2 detected US-FLI detects mild-
US and LB: 173 had HCV, 23 HBV, 123 steatosis ≥ 10% (sensitivity 85.7%. specificity moderate hepatic steatosis
NAFLD 87.5%) (AUC 0.893) and ≥ 5%-10% (sensitivity (≥ 10% on histology)
(70.7% of whom had NASH), and 33 95.9%-97.5%; specificity 66.7%) accurately and is correlated
other liver diseases (AUC 0.956). (B) US-FLI was correlated with WC, to histological and metabolic
BMI, HOMA and the number of traits of the MetS. variables in CLD owing to
(C) US-FLI was strongly correlated with steatosis different etiologies, notably
extent (rho = 0.883; P < 0.001 in the whole sample). including NAFLD
Moreover, among NAFLD patients, it was moderately
correlated with the severity of lobular inflammation
(rho = 0.490; P < 0.001); ballooning degeneration
(rho = 0.485; P < 0.001); strongly correlated with
Brunt’s inflammatory grading (rho = 0.622; P <
0.001); and weakly correlated with portal fibrosis and
stages of fibrosis
[12]
Liu et al. 117 children (10-18 years) were At multivariate analysis, US-FLI score was associated Among children with obesity,
submitted to anthropometric and with WHR, WHtR, UA, adiponectin, and M30 levels US-FLI is correlated with
laboratory assessment. Hepatitis was (all P < 0.05) among children with obesity anthropometric measures
defined by ALT > 40 units/L. LB was US-FLI ≥ 6 was the best cut-off value for predicting and laboratory tests
not performed in any patients hepatitis in children with NAFLD [PPV = 71.4%; AUC US-FLI ≥ 6 identifies raised
= 0.710 (95% CI: 0.572-0.847); P = 0.005] ALT among children with
NAFLD
Nelson et al. [7] LB and US metrics were available in Poor gallbladder wall visualization was specific for US-FLI accuracy may be low
208 adult obese individuals (mean NASH (89%), and vessel blurring was sensitive for in differentiating steatosis
age 47 years; age range 22 to 72; BMI NASH (93%) from NASH among
32.8 ± 5.1) with normal liver, bland US-FLI ≤ 4 ruled out NASH (NPP 88%; sensitivity individuals with obesity and
steatosis, or NASH (n = 14; 89; and 91%) US-FLI ≥ 5
105, respectively) At LRA, vessel blurring predicted NASH (P ≤ .01).
However, when the US-FLI score was ≥ 5, it
performed poorly in differentiating steatosis from
NASH (AUC = 0.649)
[13]
Xavier et al. 31 patients were initially evaluated for (A) Inter-observer agreement on the total US-FLI US-FLI is highly reproducible
assessing inter-observer score was excellent [average Interclass Correlation and accurately discriminates
reproducibility; 96 additional patients Coefficient of 0.972(95% CI: 0.949-0.986)] among different steatosis
with NAFLD were submitted to the grades
assessment of anthropometric, (B) US-FLI ≤ 3 had a NPV 100% for steatosis > S2
clinical, laboratory parameters, US and US-FLI ≥ 6 points had a PPV of 94.0% for US-FLI ≤ 3 rules out
and TE. Cut-off for steatosis > S1 was steatosis > S2 significant steatosis and
268 dB/m and > S2 was 280 dB/m. scores ≥ 6 points have a PPV
LB was not performed (C) AUC of FLI vs. US-FLI in discriminating the same of 94,0% for steatosis > S2
CAP cut-offs was significantly different for both cut-
offs (P < 0.001), indicating that FLI, compared to US- US-FLI performs better than
FLI, displayed a weaker capacity to differentiate both FLI in discriminating different
grades of steatosis steatosis grades
Sourianarayanane 11 normal livers; 24 bland steatosis; US-FLI ≥ 6 ruled in NASH and US-FLI ≤ 3 ruled out By confirming that it is a
and McCullough [10] and 78 NASH individuals were NASH (sensitivity 81% and 100%, respectively). In useful tool in excluding
submitted to both US and LB multivariate analysis, the difficult visualization of the NASH, this study brings US-
gallbladder wall was the only independent predictor FLI to the community level
of NASH (LR 4.2, 95%CI: 1.07-8.7 P = 0.0226) and supports its use in
epidemiological studies

ALT: alanine transaminase; AUC: area under the curve; BMI: body mass index; CI: confidence interval; FLI: fatty liver index; HBV: hepatitis B virus;
HCV: hepatitis C virus; HOMA: homeostasis model of insulin resistance; LB: liver biopsy; LR: likelihood ratio; LRA: logistic regression analysis;
M30: caspase-cleaved cytokeratin fragment of cytokeratin 18; MetS: metabolic syndrome; NAFLD: nonalcoholic fatty liver disease; NASH:
nonalcoholic steatohepatitis; NPV: negative predictive value; PPV: positive predictive power; rho: Spearman’s coefficient; TE: transient
elastography; UA: uric acid; US: ultrasonography; US-FLI: ultrasonographic fatty liver indicator; WC: waist circumference; WHR: waist-to-hip;
WHtR: weight-to-height ratio.

available to this end [15,16] .