Keywords
shear wave elastography - transient elastography - ultrasound - nonalcoholic steatohepatitis
- fibrosis
Introduction
Diffuse parenchymal diseases of the liver are one of the major causes of liver fibrosis
(LF), which leads to cirrhosis, portal hypertension, and hepatocellular carcinoma.[1] They are a major cause of morbidity and mortality in developing and developed countries.
The causative factors of liver diseases (LDs) include infection (hepatitis viruses),
autoimmune disorders, toxins, and metabolic damage. Degree of LF correlates with the
severity of liver parenchymal damage and LDs are curable and reversible if LF is detected
early.[2] Hence, estimating the degree of LF is essential in the evaluation of the severity
of LD as well as in its therapy. Liver biopsy can only assess a very limited part
of the whole liver, while fibrosis is a heterogeneously distributed entity. Hence,
its diagnosis and grading are limited by sampling variability and there is inaccurate
histopathological yield. Liver biopsy is the gold standard for evaluating the extent
of LF, but liver biopsies are associated with higher sampling errors, low repeatability,
and invasive and interobserver variability. Therefore, liver biopsy is not characterized
as the ideal technique for screening, longitudinal monitoring, and assessing the treatment
response.
In the current scenario, percutaneous liver biopsy is the gold standard for assessment
of hepatic fibrosis (HF).[3] Several noninvasive techniques for measurements of liver stiffness, such as real-time
shear wave elastography (SWE) and transient elastography (TE; FibroScan, Echosens,
Paris, France) are now available.[4] The ideal test for the staging of LF should be simple, readily available, inexpensive,
reproducible, accurate, and noninvasive. Given these conditions, ultrasound (US) elastography
has many advantages in becoming the ideal test for quantifying LF with the help of
SWE, which is a relatively new technique.[5]
[6] SWE technology showed wide acceptance and is successfully used in the assessment
of diseases of various tissues and organs.[7]
[8] The diagnostic performance of SWE is comparable or even better than TE in the detection
of portal hypertension in chronic liver disease (CLD).[9] Fibrosis is nonuniform in distribution, so elastographic measurement or biopsy from
one site is not an accurate reflection of the disease process. Previous literature
had shown that liver stiffness measured by SWE reflects the pathological stage of
fibrosis, with good accuracy and diagnostic performance.[10]
[11] However, there is no such study that determines the ideal site of liver biopsy based
on elastography. In view of heterogeneous nature of fibrosis, we performed ultrasonography
(USG) and elastography evaluations before liver biopsy to determine the ideal site
of biopsy and along with it compared the SWE and TE results.
Materials and Methods
Study Design and Study Population
The present study was approved by the institutional review board (IEC/2020/73/MA08)
and it was a prospective nonrandomized single-center tertiary institution study of
adult cohorts. Eligible patients underwent USG-guided nonfocal liver biopsy in the
day care unit of the interventional radiology department between September 2019 and
August 2020. Patients younger than 18 years and those with known cirrhosis were excluded
from the present study. Informed written consent was obtained and patients were divided
into two groups: Group U (USG group) underwent US-guided biopsy, while in Group E (elastography group) both US and elastography guidance was used for targeting the
areas of maximum tissue resistance/velocity and biopsy was planned as shown in the
flowchart in [Fig. 1]. Both techniques were compared with the FibroScan score and histopathological score
(meta-analysis of histological data in viral hepatitis [METAVIR]) of the biopsy sample
in all patients.
Fig. 1 Flowchart summarizing the study methodology.
Transient Elastography Examination
All TE examinations were performed in the supine position with arms in overhead abduction.
The intercostal space providing the best visualization of the liver at the midaxillary
line was chosen using B-mode with the FibroScan machine. The TE probe was placed in
the selected intercostal space perpendicular to the skin surface. An adequate amount
of gel was placed to form better acoustic isolation, and appropriate compression pressure
according to pressure indicator was applied. Measurements were recorded in median
number in kilopascal (kPa), which was then interpreted into LF staging. To ensure
adequate results and readings, the interquartile range (IQR)/median ratio <25% was
considered. Liver stiffness and controlled attenuation parameters (CAP) were analyzed.
Shear Wave Elastography Examination
SWE was performed by using the Toshiba Aplio, Japan, system. It measures the speed
of propagation of shear wave, which is then converted to tissue stiffness using a
computer algorithm. These quantitative values are simultaneously generated with conventional
B-mode images and also mapped as a color-coded two-dimensional elastography of tissue
stiffness. SWE was performed by two radiologists in the interventional radiology department.
The patients were kept nil per os (NPO) for 4 to 6 hours. Before the elastography
examination, patients were trained for neutral breath holding position (neither full
inspiration nor full expiration). First, we performed the routine grayscale USG of
the liver, followed by elastography mode. In the elastography mode, the region of
interest (ROI) was placed approximately 2.0 cm beneath the liver capsule 90 degrees
to the center of the transducer, avoiding major vascular structures of the liver.
The scan box measuring 0.5 × 1.0 cm and largest possible ROI (range: 15–30 mm2) was used.
Before biopsy, three to five elastography measurements were obtained at the following
locations in the right lobe of the liver: (1) right upper lobe segments (segments 7 and 8) and (2) right lower lobe segments (segment 5 and
6). The right intercostal window approach was used for upper lobe measurements, whereas
the intercostal or subcostal approach was used for right lower lobe measurements.
The mean and median liver elasticity values were calculated for each of the four segments.
The obtained measurements were expressed in meter per second (m/s). The SWE speed
was transformed into kilopascal (kPa) using Young's formula (kPa = 3 pv2), where p is tissue density and is constant for liver parenchyma (∼1,000 kg/m3) and v = speed of shear wave. The results were correlated with LF staging (METAVIR scoring).
Liver Biopsy
US-guided nonfocal liver biopsy was performed in the department of interventional
radiology. Informed consent was taken before the procedure and local anesthesia (2%
xylocaine) up to liver capsule was given. All biopsy specimens were obtained using
an 18-gauge core biopsy gun (IB; Medical Device Technologies) from the right lobe
of the liver (specimen length: ∼1.8–2 cm) because left lobe measurements are highly
influenced by the respiratory and cardiovascular movements and the left lobe is the
least favored site for liver biopsy.
Histologic Examination
For staging of LD, the METAVIR scoring system were used.[12] The METAVIR score is an ordinal scale that grades fibrosis (F) from 0 to 4, where
F4 = cirrhosis, F3 = many septa with architectural distortion but no feature of obvious
cirrhosis; F2 = few septa but with maintained parenchymal architecture; F1 = enlarged
portal tract with fibrosis; F0 = no fibrosis. Steatosis (S) was graded into the following
categories: S0 = absent, S1 < 5%, S2 = 5 to 33%, S3 = 34 to 66%, and S4 > 66%. The
necroinflammatory score (A) is the sum of (1) interface hepatitis and/or piecemeal
(score, 0–3) and (2) lobular hepatitis (score, 0–2), which gives the total necroinflammatory
activity score (A0–A3). Statistical analysis of the fibrosis, steatosis, and necroinflammatory
scores was done.
Statistical Analysis
Statistical results were analyzed using Statistical Package for the Social Sciences
software (SPSS) software, version 22.0. The mean values of the SWE velocity were estimated
from the SWE values obtained from four different liver sites. Mann–Whitney U test and t-test were used to find out the correlation between the variables of the two groups.
The liver parenchyma site with the highest positive correlation was identified by
Spearman's correlation test. Online confidence interval (CI) generator was used to
calculate the CIs for the correlation. The diagnostic performance of SWE in differentiating
different stages of fibrosis was evaluated from the area under the curve (AUC) values
of the receiver operating characteristic (ROC) curves. The sensitivity and specificity
of SWE and TE were calculated using optimal cutoff values.
Results
Our study population comprised 127 patients (86 males and 41 females) with age ranging
from 19 to 76 years (mean: 41.2 ± 13.6 years). The patients were divided into two
groups according to the biopsy guidance used; 75 patients in whom only US guidance
was used were included in Group U (US) and 52 patients in whom US and elastographic guidance was used were included
in Group E (elastography). The mean age (years) of patients in groups E and U was 40.75 ± 13.59
and 41.90 ± 12.34 years, respectively. Baseline patient characteristics of the groups
are detailed in [Table 1]. The groups were comparable with respect to etiology, CAP, liver stiffness measurement
(LSM), fibrosis, steatosis, activity, and liver function test (LFT). The causes of
CLD were nonalcoholic steatohepatitis (NASH; n = 52, 40.9%), hepatitis B virus (HBV; 31, 24.4%), hepatitis B virus (HCV; 17, 13.4%),
chronic biliary pathology (primary bliary cirrhosis, PBC and primary sclerosing cholangitis,
PSC; 9, 7.1%), and other diseases including autoimmune hepatitis and nonalcoholic
fatty liver (NAFL; 18, 14.2%). FibroScan (TE) mean liver stiffness measurement (LSM)
was 9.210 ± 5.52 kPa (range: 3.5–42 kPa) and mean CAP was 266.59 ± 56.52 kPa (range:
172–396 kPa).
Table 1
Clinical and demographic characteristics (mean ± standard deviations)
|
Variable
|
Category
|
N = 127
|
Group U (N = 75)
|
Group E (N = 52)
|
p-Value
|
|
Age (y)
|
Male: 86; female: 41
|
41.22 ± 13.06 (19–76)
|
40.75 ± 13.599
|
41.90 ± 12.345
|
0.44
|
|
Liver function test
|
Bilirubin
|
1.809 ± 2.7172 (0.2–16)
|
2.260 ± 4.6806
|
1.158 ± 1.2715
|
0.236
|
|
AST
|
89.26 ± 52.7 (16–455)
|
106.681 ± 66.363
|
64.48 ± 51.262
|
0.146
|
|
ALT
|
96.87 ± 61 (13–521)
|
111.69 ± 159.772
|
75.48 ± 65.097
|
0.267
|
|
AKP
|
107.33 ± 86.705 (13–695)
|
114.97 ± 93.835
|
96.31 ± 74.750
|
0.309
|
|
GGT
|
66.66 ± 83.275 (7–660)
|
69.56 ± 73.846
|
62.48 ± 95.864
|
0.156
|
|
Albumin
|
3.848 ± 0.6347 (1.9–5.1)
|
3.896 ± 0.6420
|
3.780 ± 0.6245
|
0.271
|
|
FibroScan
|
CAP
|
266.59 ± 56.525 (172–396)
|
264.28 ± 56.301
|
269.98 ± 57.243
|
0.595
|
|
LAM
|
9.210 ± 5.5210 (3.5–42)
|
9.571 ± 5.8338
|
8.890 ± 4.2958
|
0.930
|
|
Fibrosis
|
0
|
56 (44.1)
|
24
|
32
|
|
|
1
|
20 (15.7)
|
10
|
10
|
|
|
2
|
24 (18.9)
|
9
|
15
|
|
|
3
|
21 (16.5)
|
7
|
14
|
|
|
4
|
6 (4.7)
|
2
|
4
|
|
|
Activity
|
0
|
75 (59)
|
28
|
47
|
|
|
1
|
8 (6.3)
|
3
|
5
|
|
|
2
|
30 (23.7)
|
15
|
15
|
|
|
3
|
14 (11)
|
6
|
9
|
|
|
Steatosis
|
0
|
73 (57.5)
|
28
|
45
|
|
|
1
|
27 (21.2)
|
11
|
17
|
|
|
2
|
20 (15.8)
|
10
|
10
|
|
|
3
|
7 (5.5)
|
3
|
2
|
|
|
Diagnosis
|
NASH
|
52 (40.9)
|
28
|
24
|
|
|
Chronic hepatitis (HBV)
|
31 (24.4)
|
18
|
13
|
|
|
Chronic hepatitis (HCV)
|
17 (13.4)
|
11
|
6
|
|
|
Chronic biliary pathology
|
9 (7.1)
|
7
|
2
|
|
|
Others
|
18 (14.2)
|
11
|
9
|
|
|
Velocity (m/s), n = 52
|
Segment 5
|
1.9471 ± 0.5 (1.20–3.08)
|
|
|
|
|
Segment 6
|
2.1356 ± 0.6 (1.40–3.88
|
|
|
|
|
Segments 7
|
2.0373 ± 0.6 (1.28–3.81)
|
|
|
|
|
Segments 8
|
2.1058 ± 0.6 (1.36–3.95)
|
|
|
|
Abbreviations: AST, aspartate aminotransferase; ALT, alanine transaminase; AKP, alkaline
phosphatase; CAP, controlled attenuation parameter; GGT, gamma-glutamyl transferase;LSM,
liver stiffness measurement; NASH, nonalcoholic fatty liver disease.
The study population undergoing liver biopsy comprised 56 patients without any evidence
of fibrosis (F = 0). Twenty patients had nonsignificant fibrosis (F = 1), 24 patients had significant LF (F = 2), 21 patients had severe LF (F = 3), and 6 patients had cirrhosis (F = 4). In total, 27.2% patients accounted for stage S0 steatosis. Seventy-three of
the 127 patients (57.4%) in our study had a total activity score of A0.
We noted maximum liver stiffness in segment 6 and took maximum biopsies from the inferior
segments in about approximately 67% as noted in [Table 2]
. There was no significant difference between intersegmental liver stiffness and mean
velocity; however, the biopsy segment velocities show a significant difference with
the mean liver stiffness suggestive of heterogeneous distribution of fibrosis. The
rho (r; Spearman's correlation) value between biopsy segments and mean velocities shows
excellent correlation as described in [Table 2]. Further dividing the right lobe on the basis of Couinaud's classification shows
excellent correlation of all segments (anterior/posterior and superior/inferior) with
V
mean and V
biopsy; however, there is more correlation with the inferior segment (5/6) as described
in [Table 3].
Table 2
SWE velocity value at different segments and its correlation with mean velocity (r value Spearman's correlation) and percentage/frequency of biopsy segment (mean velocity:
2.04 m/s)
|
Segments
|
Mean velocity ± SD
|
Confidence coefficient
|
r value with V
mean
|
Frequency (%)
|
|
Biopsy segment
|
2.31 ± 0.67
|
0.18
|
0.925
|
–
|
|
Segment 5
|
1.95 ± 0.58
|
0.16
|
0.739
|
9 (17.3%)
|
|
Segment 6
|
2.12 ± 0.63
|
0.17
|
0.874
|
26 (50%)
|
|
Segment 7
|
2.04 ± 0.63
|
0.17
|
0.816
|
11 (21.2%)
|
|
Segment 8
|
2.11 ± 0.65
|
0.18
|
0.841
|
6 (11.5%)
|
Abbreviations: SD, standard deviation; SWE, shear wave elastography.
Table 3
Correlation of SWE velocity value at different site with mean velocity (r value Spearman's correlation)
|
Segments
|
Mean ± SD
|
r with V
mean
|
r with V
biopsy
|
|
Anterior (5/8)
|
2.02 ± 0.063
|
0.85
|
1
|
|
Posterior (6/7)
|
2.08 + 0.065
|
0.906
|
1
|
|
Inferior (5/6)
|
2.05 + 0.010
|
1
|
1
|
|
Superior (7/8)
|
2.05 + 0.024
|
1
|
0.875
|
Abbreviations: SD, standard deviation; SWE, shear wave elastography.
The mean velocity in patients with stage F4 fibrosis was 2.82 ± 0.24, F3 was 2.6 9 ± 0.74,
F2 fibrosis was 2.09 ± 0.20, and F1 fibrosis was 1.90 ± 0.19 m/s. Different degrees
of fibrosis showed a significant difference in the level of LS (mean velocity) as
shown in box and whisker plot ([Figs. 2] and [3]). The ROC curve drawn to differentiate fibrosis stage with cutoff value and AUCs
for differentiating fibrosis stage is noted in [Table 4] and [Fig. 4].
Fig. 2 Box and whisker plot shows the mean shear wave elastography (SWE) values in the right
lobe of the liver for various fibrosis stages. The top and bottom lines of each box represent the first and third quartiles (25th and 75th percentiles). The middle lines of each box are the median and the lines of the upper and lower boxes are the 5th and 95th percentiles.
Fig. 3 Receiver operating characteristic (ROC) curve of the diagnostic performance of shear
wave elastography (SWE; velocity) for the prediction of different grades of liver
fibrosis in group E. Graphs show AUCs (area under the receiver operating characteristic
[AUROC]) for mean SWE values at the right lobe. (A) The cutoff (AUC) value for fibrosis stage F0–F1 (normal to mild) is 1.64 (86.7).
(B) The cutoff value for fibrosis stage F1–F2 (moderate) is 1.94 (95.5). (C) The cutoff value for fibrosis stage F2–F3 (severe) is 2.44 (94.6), and (D) the cutoff value for fibrosis stage F3–F4 (cirrhosis) is 2.58 (93).
Table 4
Optimal stiffness cutoff value of SWE in group E according to level fibrosis
|
Parameter
|
Cutoff (m/s)
|
AUC
|
Sensitivity
|
Specificity
|
Asymptotic 95% confidence interval
|
|
Lower
|
Upper
|
|
F0–F1
Normal–mild
|
1.64
|
86.7
|
81
|
80
|
72
|
100
|
|
F1–F2
Mild–moderate
|
1.94
|
95.5
|
88.9
|
85.3
|
90.6
|
100
|
|
F2–F3
Moderate–severe
|
2.44
|
94.6
|
88.9
|
83.7
|
88.3
|
100
|
|
F3–F4
Cirrhosis
|
2.58
|
93.0
|
100
|
72
|
63.3
|
100
|
Abbreviations: AUC, area under the curve; SWE, shear wave elastography.
Fig. 4 Box and whisker plot shows the mean transient elastography (TE) values in the right
lobe of the liver for various fibrosis stage. The top and bottom lines of each box represent the first and third quartiles (25th and 75th percentiles). The middle lines of each box are the median and the lines of the upper and lower boxes are the 5th and 95th percentiles. The error bars show the minimum and maximum values. LSM, liver stiffness measurement.
Liver stiffness measurement according to fibrosis stages: The mean LSM using TE for
fibrosis stages F0, F1, F2, F3, and F4 was 5.57 (4.2–7), 6.25 (4.9–9.1), 8.5 (5.5–13.4),
9.15 (8.0–15.9), and 15.55 (14.5–22.3) kPa, respectively, as described in [Fig. 5]. The areas under the receiver operating characteristic (AUROCs) for TE and SWE in
fibrosis stages F1, F2, F3, and F4 are shown in [Table 5]. TE was good for the diagnosis of fibrosis stages F2, F3, and F4; while SWE was
fair for the diagnosis of fibrosis stages F1 and F2 and fairly equal for the diagnosis
of stages F2 and F3. AUCs in differentiating no or mild fibrosis (F1) from significant
fibrosis (≥F2) were 95.5 with cutoff value of at least 1.94 m/s in the present study
as seen in [Table 5].
Table 5
Comparison between SWE and TE
|
Parameter
|
Group
|
TE cutoff (kPa)
SWE cutoff (m/s)
|
AUC (%)
|
Sensitivity
|
Specificity
|
|
F0–F1
Normal–Mild
|
TE
|
6.1
|
82.4
|
72
|
60
|
|
SWE
|
1.64
|
86.7
|
81
|
80
|
|
F1–F2
Mild–moderate
|
TE
|
7.8
|
93.5
|
90.9
|
84.3
|
|
SWE
|
1.94
|
95.5
|
88.9
|
85.3
|
|
F2–F3
Moderate–severe
|
TE
|
9.0
|
96.4
|
88.9
|
87.7
|
|
SWE
|
2.44
|
94.6
|
88.9
|
83.7
|
|
F3–F4
Cirrhosis
|
TE
|
17.5
|
97.9
|
100
|
94.4
|
|
SWE
|
2.58
|
93.0
|
100
|
72
|
Abbreviations: AUC, area under the curve; SWE, shear wave elastography; TE, transient
elastography.
Fig. 5 Receiver operating characteristic (ROC) curve of the diagnostic performance of transient
elastography (TE; LAM) for the prediction of different grades of liver fibrosis in
all patients. Graphs show AUCs (area under the receiver operating characteristic [AUROC])
for the mean TE values at the right lobe to. (A) The cutoff (AUC) value for fibrosis stage F0–F1 (normal to mild) is 6.25 (84.7).
(B) The cutoff value for fibrosis stage F1–F2 (moderate) is 8.5 (92.5). (C) The cutoff value for fibrosis stage F2–F3 (severe) is 9.1 (95.6). (D) The cutoff value for fibrosis stage F3–F4 (cirrhosis) is 15.5 (97.2).
Discussion
Management and prognosis of CLD are highly dependent on the stage of fibrosis; hence,
management of these patients relies on estimating the degree of fibrosis. Liver biopsy
was one of the earliest and gold standard approaches to evaluate LF.
We assessed the clinical usefulness of LSM using SWE with various CLD in predicting
the accurate site of biopsy and the degree of LF by analyzing the SWE and histopathological
results. The LS values measured by SWE showed significant correlation with severity
of LF (r = 0.88, p < 0.001). Additionally, the present study results indicated that SWE had a high detection
rate for significant (≥F2) and advanced fibrosis (≥F3; AUROC values of 0.95 and 0.94,
respectively) as reported in previous literature.[13]
Biopsy Site
In our study, we only took the stiffness in the right lobe as the left lobe is the
least favored site for liver biopsy.[14]
[15] Also Friedrich-Rust et al[16] reported that the values of the left and right LSM showed no difference statistically.
Elastography findings suggested that there was variable liver stiffness in the right
lobe, which was reflected as a different velocity in the segments of the right lobe.
This suggests a heterogenous distribution of the fibrosis in the liver. This might
be attributed to sinusoidal blood oxygen levels, which in turn are directly related
to the proportion of blood contributed by capsular arterioles and the ratio of systemic
to portal blood perfusing the segment. Liver parenchyma that is closer to the hilum
or has a higher proportion of postprandial portal blood might allow fewer reactive
oxygen species to form and hence a lesser degree of fibrosis.[17]
[18]
Spearman's correlation between biopsy segment velocities shows that the inferior lobe/segments
are better for biopsy, which was denoted by the excellent correlation with mean velocities
(r = 1). SWE stiffness of the biopsy segments better correlates with the histopathological
finding compared with the mean velocities of the right lobe. It is in contrast to
the study performed by Samir et al,[5] who noted that the accurate site of measurement of SWE stiffness is the right superior
or upper lobe.
We compared the variability of the data of elastography in different segments of the
right lobe and we have seen that the coefficient of variance is lowest in segment
5 (29%) compared with other segments (>31%) for diffuse LDs. This result is in concordance
with the study by Ling et al[19] who revealed that the intraindividual measurements of LSM exhibited the lowest variation
in segment 5 (coefficient of variation, CV 27%). Furthermore, they also stated a statistically
significant difference in LSM between segments 5 and 1, 2, 3, 7, or 8 (p < 0.05). In contrast, in our study no significant difference was noticed in between
segments 5 and 6, 7, and 8. Overall, we found that the inferior segment of the right
lobe shows least variability and better correlation with LF.
Shear Wave Elastography
In the present study, we found that the AUROC was over 90% for fibrosis at stages
F1–F2, F2–F3, and F3–F4, indicating that SWE is accurate in assessing LF at different
stages. The AUROC increase with increase in the grade of fibrosis. This study shows
that SWE has high sensitivity and specificity to analyze LF in patients with ≥F2 fibrosis
stage. The AUROC of DOR (diagnostic odds ratio) was 0.90, indicating that it has a
higher diagnostic performance value. In our study, SWE showed a statistically significant
difference in values of the mean LS in different grades of LF. Patients with advanced
LF (METAVIR: F2/F3) had higher LS than those with early stages of fibrosis (METAVIR:
F0/F1). This suggests that SWE has good predicting power for different stages of LF.
Moustafa et al,[20] Cassinotto et al,[21] and Guibal et al[13] showed that LS depends on the stage of fibrosis with significant relationship with
liver biopsy and there was a significant difference in the LS in patients with advanced
fibrosis compared to those in early stage of fibrosis.
The sensitivity for the diagnosis of significant fibrosis (≥F2), severe fibrosis (≥F3),
and liver cirrhosis (F4) was 89.9, 89.9, and 100%, respectively, and the specificity
was 85, 83.5, and 72%, respectively. These results were in concordance with study
by Nierhoff J et al.[22] Comparison with previous studies is summarized in [Table 6].
Table 6
Comparison of shear wave elastography (SWE) cutoff with different studies in kilopascal
(kPa)
|
Etiology
|
≥F2
|
≥F3
|
F4
|
Studies
|
|
Chronic hepatitis B
|
7.1
|
7.9
|
10.1
|
Leung et al[23]
|
|
Chronic hepatitis B
|
8.5
|
11.5
|
18.1
|
Guibal et al[13]
|
|
Nonalcoholic fatty liver disease
|
8.7
|
10.7
|
14.4
|
Tada et al[24]
|
|
Autoimmune liver disease
|
9.7
|
13.2
|
16.3
|
Zheng et al[26]
|
|
Various chronic liver diseases
|
8.6
|
10.5
|
14.0
|
Jeong et al[6]
|
|
Various chronic liver diseases (kPa)
|
11.2
|
17.8
|
19.5
|
Present study
|
|
Shear wave velocity in different grade of fibrosis
|
≥F2
|
≥F3
|
F4
|
|
2D-SWE (ASQ) Toshiba (vendor specification)
|
1.76
|
2.21
|
2.86
|
|
Present study
|
1.92
|
2.44
|
2.58
|
SWE versus TE
The intraclass correlation coefficients (ICC) shows there was an intraobserver agreement
for TE and SWE. There was excellent correlation in repeated measurements by the same
operator for TE and SWE. In a study conducted by Leung et al,[23] the sensitivity and specificity of SWE in the diagnosis of LF were 85 and 92%, respectively.
For diagnosis of liver cirrhosis, the sensitivity and specificity of SWE were 97 and
93%, respectively. These results are quite similar to the results in our study as
noted in [Tables 6]. Tada al[24] stated that SWE can be used similar to TE in the assessment of LF. A strong correlation
between SWE and TE was established by Deffieux et al.[25] In their study for LF staging, the AUROC curves were similar for SWE and TE. Zheng
et al[26] concluded that SWE had a higher sensitivity and specificity for diagnosis of F4
METAVIR stage as compared with lower stages, which is also seen in our results. Our
results are in concordance with result of Tada et al[24] and Ferraioli et al,[27] who observed that both SWE and TE show a similar diagnostic performance in the evaluation
of LS. Previous literature reports no significant association between the liver stiffness
values and steatosis or inflammatory activity within the liver. In our study, we found
that the degree of steatosis shows no significant correlation with the SWE velocity
measurements and this finding is concordant with previous findings in the literature.[28]
[29]
One major disadvantage of SWE is that there are no uniform standard values available
across different US systems and the value of LS varies with different US systems developed
by different manufacturers at the same degree of fibrosis. The velocity at different
levels of fibrosis in the present study and vendor specification regarding the USG
machine elastography is summarized in [Table 6].
The limitations of the present study include a small sample size with a heterogenous
study population. Therefore, etiology-specific studies should be performed. Other
noninvasive techniques such as magnetic resonance (MR) elastography and serum markers
of fibrosis were also not used in our study. The AUROC values of liver stiffness measured
using SWE is slightly lower in the present study than previously reported. This could
be explained by the fact that a different hardware was used with the inclusion of
a heterogeneous cohort with different causes of CLD.
Future Directions
This is a pilot study with a small sample size and uture studies with a larger sample
size are required to establish the accuracy of USG elastography to find themost accurate
site of biopsy. With recent advancements, MR elastography–guided biopsy may produce
promising results.
Conclusion
Fibrosis is a heterogeneously distributed entity as concluded by the fact that the
SWE segmental mean velocity is different in different segments of the right lobe liver.
Therefore, elastography-guided liver biopsy helps in defining the accurate site for
biopsy and hence can improve the histopathological yield in detecting LF in patients
with CLD. This also helps in recording the baseline liver stiffness, which will be
helpful in follow-up. The diagnostic performance accuracy of SWE is comparable to
FibroScan.
