Keywords
Acoustic radiation force impulse elastography - alcoholic liver disease - shear wave
velocity, fibrosis stages
Introduction
Alcoholic liver disease in India is rampant with a prevalence of 5% of the total population
and liver cirrhosis mortality rate is as high as 63% according to recent data published
by the WHO in 2014.[1] Chronic liver diseases irrespective of etiology lead to hepatic fibrosis; a dynamic
process with evidence of regression on early intervention, if neglected architectural
distortion and cirrhosis ensues.[2],[3] Thus, noninvasive early diagnosis of liver fibrosis is of paramount importance.
Liver biopsy is the gold standard, but is an invasive, painful procedure, with sampling
error, interobserver variability, and an error rate of 33%.[4],[5] Elastography is an alternative, quantitative technique for liver fibrosis. ARFI
elastography is one of the ultrasound-based elastographic techniques used for grading
of liver fibrosis.
Aims and Objectives
This study was conducted with the aim to evaluate the diagnostic performance of ARFI
elastography against biochemical indices and liver biopsy and also assess the utility
of available shear wave (SW) velocity cutoffs to potentially avoid liver biopsies.
Materials and Methods
Patient selection and description
This institutional cross-sectional study was approved by the institutional ethics
committee and an informed written consent was obtained from all the study participants.
We evaluated a total of 50 patients who fulfilled the inclusion criteria.
Inclusion criteria
Patients aged more than 18 years old with high-risk alcohol consumption (>40 g/day
for men and >20 g/day for women for a cumulative period of more than 5 years)[6] were included in the study.
Exclusion criteria
Patients with liver trauma, established cases of cirrhosis, a known history of chronic
liver diseases of various other etiologies (viz., viral hepatitis, autoimmune hepatitis,
extrahepatic cholestasis, hepatic venous congestion, and other liver diseases) and
patients in whom liver biopsy was not possible (due to medical conditions) were excluded
from the study.
Technique of ARFI elastography
In our study, Acuson S3000 (Siemens Medical Solutions, USA, Inc.) ultrasound machine
equipped with ARFI elastography software was used to obtain SW velocities. ARFI elastography
was performed using the manufacturer guidelines. The patients were asked to visit
with fasting state for 4–6 h. They were trained for resting respiratory position (neither
full inspiration nor full expiration) before the elastography examination. ARFI elastography
was performed in dorsal decubitus position, with the right arm above the head for
adequate intercostal access and breath hold during each measurement. Routine grayscale
ultrasonography of liver was performed first, followed by switching over to the elastography
mode. In the elastography mode, the region of interest (ROI), which is a box of 0.5
× 1.0 cm, was placed 1.5--2.0 cm beneath the Glisson’s capsule perpendicular to the
center of the transducer, avoiding major vessels of the liver. Ten valid measurements
were taken in the right lobe of liver. When the interquartile range of the measurements
was >0.3 × median SW speed, the measurements were considered unreliable.[7],[8] Median SW velocity of the measured values was displayed on the results screen and
used for grading the liver fibrosis.
Biochemical indices
The APRI and FIB-4 indices were calculated based on the age, liver function tests,
and platelet counts that were obtained on the same day.
Liver biopsy
Liver biopsy was performed on the same day. All the biopsies were performed under
ultrasound guidance after SW measurements and were reported by same pathologist for
all patients. The pathologist was also blinded to the ARFI elastography results.
Statistical analysis
All the statistical analyses were performed in SPSS 16.0 software (Statistical Package
for Social Science Inc., Chicago, IL, USA). Descriptive statistics, frequency analysis
for categorical variables, and mean and standard deviations were obtained for all
continuous variables. AUROC curve analysis was done to assess diagnostic performance
of each test (viz., ARFI elastography, aspartate transaminase-to-platelet ratio index,
and fibrosis-4 indices) separately and the results were compared for each test; Kruskal–Wallis
test was performed for analysis of normal variance of the SW velocities and Pearson
correlation coefficient was derived to correlate liver biopsy results with the SW
elastography results.
Results
Our study population comprised 50 male patients ranging from 24 to 62 years (mean
41 ± 9.81 years) fulfilling the inclusion criteria. We did not come across any female
patients. About 44% (n = 22) of our patients were clinically asymptomatic with high risk alcohol consumption.
About 36% (n = 18) of the cases had hepatomegaly with high-risk alcoholism, and 20% (n = 10) of the cases had hepatosplenomegaly and ascites and were referred for confirmation
of cirrhosis.
The study population on liver biopsy comprised 12 patients of cirrhosis (F = 4), 8
patients of severe liver fibrosis (F = 3), 7 patients of significant liver fibrosis
(F = 2), 9 patients of nonsignificant fibrosis (F = 1), and 14 without any evidence
of fibrosis (F = 0).
ARFI elastography positively correlated [Pearson correlation (r): +0.76] with liver
biopsy results in our study. The box plots for measured SW velocities in different
pathological stages of fibrosis as observed in our study population are depicted in
[Figure 1].
Figure 1: Box Plots of the observed median shear wave velocities
Optimal SW velocity cutoff values for different stages of liver fibrosis as suggested
by Friedrich-Rust et al.[9] were used and 24 patients were classified into F ≤ 1 (METAVIR) stage, implying either
no fibrosis or minimal fibrosis. Nine patients were classified to be of F ≥ 2 (significant
fibrosis), 6 patients as F ≥ 3 (severe fibrosis), and 11 patients were grouped as
F = 4 (cirrhotic).
On the basis of APRI scores, 35 patients had no fibrosis, 5 patients had significant
fibrosis, and 6 probably had cirrhosis. APRI score was inconclusive in 4 patients.
Similarly, FIB-4 scores suggested that 31 patients had no severe fibrosis, 10 patients
had severe fibrosis, and in the remaining 9 patients, FIB-4 index was inconclusive.
Highest AUROC was derived for ARFI elastography, followed by FIB-4 index and the APRI
index to grade fibrosis. AUROC was highest for the severe fibrosis versus cirrhosis
of liver, implying the ability of ARFI elastography to better distinguish between
the severe fibrosis versus cirrhosis of liver. The ROC curves for differentiation
of F > 3 and F = 4 stages of liver fibrosis are shown in [Figure 2] and [Figure 3].
Figure 2: ROC curve using APRI Index (Red Line), ARFI elastography (Blue Line), and FIB-4 index
(Green Line) as test curves in diagnosis of Severe Fibrosis and Cirrhosis after applying
the cutoff values suggested by F. Rust et al.[9] with pathological outcomes as reference standard
Figure 3: ROC curve using APRI Index (Red Line), ARFI elastography (Blue Line), and FIB-4 index
(Green Line) as test curves in diagnosis of Severe Fibrosis and Cirrhosis in our study
based on only the pathological outcomes
Discussion
Chronic alcoholism causes hepatic steatosis, alcoholic steatohepatitis, and steatofibrosis
finally causing cirrhosis. Classically, aspartate transaminase (AST):alanine transaminase
(ALT) ratio >1 is a feature of the alcoholic liver cirrhosis that distinguishes it
from other chronic liver diseases. APRI (AST-platelet ratio index) and FIB-4 (fibrosis-4)
indices are commonly used biochemical indices to grade and monitor the stages of liver
fibrosis. On further advancement, steatofibrosis progresses into alcoholic liver fibrosis.
Role of elastography in alcoholic liver disease
Currently, transient elastography (TE), SW elastography, and ARFI imaging are the
three main ultrasound-based elastographic techniques that allow direct and indirect
quantification of liver stiffness.[10] ARFI is a relatively newer elastography technique used for assessment of liver fibrosis.
Bota et al. in 2013 did a meta-analysis comparing the ARFI elastography against TE with the
data available in major research databases. Theyconcluded, “Acoustic radiation force
impulse elastography is a good method with higher rate of reliable measurements and
similar predictive value to TE for significant fibrosis and cirrhosis.”[11] In other studies conducted by Rizzo et al.,[12] Leung et al.,[13] Doherty et al.,[14] and Jaffer et al.,[15] ARFI elastography had clear advantage over the TE even in patients with ascites.
However, Friedrich-Rust et al.[16] and Gerber et al.[17] reported no significant difference in ARFI and TE methods. Hence, ARFI elastography
is easily available radiological tool that can be used in conjunction with other biochemical
tests and indices as an effective tool for acceptable noninvasive assessment of liver
fibrosis.
Friedrich-Rust et al. in their pooled meta-analysis of eight studies on ARFI elastography including a
total of 518 patients suggested that the optimal cutoff for the diagnosis of significant
fibrosis was 1.34 m/s (AUROC = 0.87), 1.55 m/s (AUROC = 0.91) for severe fibrosis,
and 1.80 m/s (AUROC = 0.86) for the diagnosis of liver cirrhosis. When these cutoffs
were used, we observed that AUROC for significant fibrosis was 0.98, for severe fibrosis
it was 0.97, and for cirrhosis it was 0.96. When the meta-analysis cutoff values were
not applied, there was a clear advantage with an AUROC of 0.97 for the diagnosis of
liver cirrhosis. However, in cases of severe fibrosis and significant fibrosis, diagnostic
accuracy was comparable to the biochemical tests as seen from our observations [Table 1].[9]
Table 1
Diagnostic performance of APRI Index vs ARFI Elastography vs FIB-4 Index
|
Number of Patients (n=50)
|
Severe Fibrosis vs. Cirrhosis (F=4)
|
|
APRI
|
FIB-4
|
ARFI elastography (using optimal cut-offs)
|
ARFI elastography (using data obtained in our study)
|
|
Sensitivity
|
0.83
|
0.83
|
1.00
|
0.83
|
|
Specificity
|
0.76
|
0.74
|
0.63
|
0.97
|
|
PPV
|
0.53
|
0.50
|
0.92
|
0.92
|
|
NPV
|
0.94
|
0.93
|
0.97
|
0.97
|
|
AUROC
|
0.83
|
0.83
|
0.97
|
0.97
|
|
95% CI
|
0.651-0.945
|
0.636-0.934
|
0.702-0.930
|
0.810-0.990
|
|
P
|
0.002
|
0.002
|
0.001
|
0.046
|
We have come across very limited literature about ARFI elastography in alcoholic liver
diseases, but it has been validated in liver fibrosis of other etiologies, namely,
viral hepatic fibrosis, nonalcoholic steatohepatitis, etc., Two studies were available
in the literature addressing the use of ARFI elastography in alcoholic liver fibrosis:
one by Zhang et al. in China (2014) on 112 patients and other by Kiani et al. in France (2016) on 83 patients from deaddiction centers.[18],[19] In our study, we observed the SW velocity cutoffs of 1.37 m/s for significant fibrosis,
1.51 m/s for severe fibrosis, and 1.87 m/s for cirrhosis of liver. The cutoffs from
meta-analysis were comparable to our study, whereas the observations of Zhang et al. were lower and observations of Kiani et al. were higher [Table 2]. The probable reason for this variability in measurements could be because of absence
of literature, specifying the depth from the surface at which the measurements are
made. The cutoff values of Zhang et al., who have compared the patients with normal ALT levels (1.23 m/s for F ≥ 2, 1.27
for F ≥ 3, and 1.41 for F = 4) and raised ALT levels (1.33 m/s for F ≥ 2, 1.40 for
F ≥ 3, and 1.65 for F = 4), were still lower than our results. Hence, we also postulate
that the variability in cutoffs can be because of the choice of ROI and the depth
from the surface for the measurements.
Table 2
Comparison of Shear wave velocity cut-offs from available literature
|
Present Study ARFI cut-off (in m/s) [AUROC]
|
M. Friedrich-Rust et
al.[9] ARFI cut-off (in m/s) [AUROC]
|
D. Zhang et al.[18] ARFI cut-off (in m/s) [AUROC]
|
Anita Kiani et al.[19] ARFI cut-off (in m/s) [AUROC]
|
|
F≥2
|
1.37 [0.65]
|
1.34 [0.87]
|
1.27 [0.85]
|
1.63 [0.87]
|
|
F≥3
|
1.51 [0.70]
|
1.55 [0.91]
|
1.40 [0.88]
|
1.84 [0.86]
|
|
F=4
|
1.87 [0.97]
|
1.8 [0.93]
|
1.65 [0.89]
|
1.94 [0.89]
|
Advantages of ARFI elastography
We observed that it was possible to measure liver stiffness in obese patients and
those with ascites, which is not possible with TE.[3],[20],[21] The SW velocities in left lobe of liver were difficult to measure and also were
unreliable owing to the cardiac motion degrading the quality of SWs as observed in
few other studies.[22],[23]
Disadvantages of ARFI elastography
Ultrasound systems for elastography are manufactured by different manufacturers, and
as a result of it, standard values are not available for comparison across different
ultrasound systems. In our study, the reliable measurements could be obtained upto
a depth of 8 cm, which becomes a hindrance in cases of gross ascites.
Limitations of the study
Our study included a small sample size, which would limit statistical power of the
study. The effect of right heart failure and hepatic venous congestion was not considered
as none of our patients had any known cardiac disease. The effect of ALT levels was
not considered in our study, which could potentially affect the choice of cutoff values.
Future directions
A larger cohort of the patients needs to be studied to obtain diagnostic figures.
With recent advancements, magnetic resonance elastography has been shown to have promising
results; so, ARFI elastography needs to be compared with it and diagnostic accuracy
and feasibility need to be studied.
Conclusion
Staging of liver fibrosis helps the patients in a reward system behavior where a patient
is committed to detoxification and a regression or reversal of liver fibrosis encourages
the patients to further comply with treatment. ARFI elastography has an independent
diagnostic ability to perform with AUROC of 0.98 for cirrhosis. The available cutoff
values of SW velocities have been well validated in this study with high-diagnostic
performance for significant (F ≥ 2) and severe (F ≥ 3) fibrosis and cirrhosis as well.
Unlike the liver biopsy, ARFI elastography uses a ROI that can be assessed for as
many sample areas as possible and for as many repetitions as required noninvasively.
It can be performed even in patients with altered coagulation factors. ARFI elastography
can be used as a screening tool in asymptomatic patients with suspected liver cirrhosis
when biochemical tests reveal an abnormality or altered biochemical indices (e.g.
APRI and FIB-4). Thus, ARFI elastography of the liver can significantly reduce the
need for liver biopsy.
