Transient elastography and diffusion-weighted magnetic resonance imaging for assessment of liver fibrosis in children with chronic hepatitis C

Abstract

BACKGROUND

Chronic hepatitis C (CHC) is a health burden with consequent morbidity and mortality. Liver biopsy is the gold standard for evaluating fibrosis and assessing disease severity and prognostic purposes post-treatment. Noninvasive alternatives for liver biopsy such as transient elastography (TE) and diffusion-weighted magnetic resonance imaging (DW-MRI) are critical needs.

AIM

To evaluate TE and DW-MRI as noninvasive tools for predicting liver fibrosis in children with CHC.

METHODS

This prospective cross-sectional study initially recruited 100 children with CHC virus infection. Sixty-four children completed the full set of investigations including liver stiffness measurement (LSM) using TE and measurement of apparent diffusion coefficient (ADC) of the liver and spleen using DW-MRI. Liver biopsies were evaluated for fibrosis using Ishak scoring system. LSM and liver and spleen ADC were compared in different fibrosis stages and correlation analysis was performed with histopathological findings and other laboratory parameters.

RESULTS

Most patients had moderate fibrosis (73.5%) while 26.5% had mild fibrosis. None had severe fibrosis or cirrhosis. The majority (68.8%) had mild activity, while only 7.8% had moderate activity. Ishak scores had a significant direct correlation with LSM (P = 0.008) and were negatively correlated with both liver and spleen ADC but with no statistical significance (P = 0.086 and P = 0.145, respectively). Similarly, histopathological activity correlated significantly with LSM (P = 0.002) but not with liver or spleen ADC (P = 0.84 and 0.98 respectively). LSM and liver ADC were able to significantly discriminate F3 from lower fibrosis stages (area under the curve = 0.700 and 0.747, respectively) with a better performance of liver ADC.

CONCLUSION

TE and liver ADC were helpful in predicting significant fibrosis in children with chronic hepatitis C virus infection with a better performance of liver ADC.

Key Words: Apparent diffusion coefficient; Chronic hepatitis C; Diffusion-weighted magnetic resonance imaging; Liver fibrosis; Liver stiffness; Transient elastography

Core Tip: Although liver biopsy is not a necessity in the diagnosis of hepatitis C virus and is no longer a prerequisite for starting antiviral therapy, it remains a critical necessity to assess liver fibrosis for prognostic purposes. Noninvasive prediction of liver fibrosis is a challenging issue, especially in the pediatric population. Several studies have evaluated noninvasive serological and radiological tools for fibrosis prediction, among which are liver stiffness measurement using transient elastography (TE) and apparent diffusion coefficient using diffusion-weighted magnetic resonance imaging. The current study evaluated TE and diffusion-weighted magnetic resonance imaging compared to liver biopsy in assessing liver fibrosis.

INTRODUCTION

Hepatic fibrosis is the end result of liver injuries in most chronic liver diseases and can lead to cirrhosis, which is complicated by the development of portal hypertension and the majority of clinical complications if a person does not seek treatment[1]. One of the most substantial problems of public health concern is hepatitis C virus (HCV) infection as clinical liver disease is extremely rare in childhood. Infection discovery is almost incidental during routine work up for other reasons or if screened for in high-risk groups[2]. Undiagnosed HCV infection in children progresses to decompensated liver disease and hepatocellular carcinoma during adulthood[3].

World Health Organization estimated that in 2022, approximately 242000 people died of hepatitis C, mostly from cirrhosis and hepatocellular carcinoma. Globally, an estimated fifty million people have chronic HCV infection, with about one million new infections occurring every year. Direct-acting antivirals (DAAs) can cure more than 95% of patients with HCV infection, but access to diagnosis and treatment is low. There is currently no effective vaccine against hepatitis C[4]. The 69th World Health Assembly approved the Global Health Sector Strategy to eliminate HCV infection by 2030, which can become a reality with the launch of DAAs[5]. The advent of DAAs has revolutionized the natural history of HCV infection with a remarkable safety profile in all stages of chronic HCV[6].

The role of liver biopsy in pediatric patients with chronic viral hepatitis was questioned due to the development of noninvasive alternatives used for the assessment of the severity of liver fibrosis. However, none of these methods has been validated in children; therefore, liver biopsy remains the gold standard for the evaluation of liver disease progression in children with chronic viral hepatitis[7]. The main indication for the liver biopsy is prognostic purposes, evaluating disease severity and monitoring response to treatment[8].

Liver biopsy is invasive, and has several limitations such as physical and mental discomfort that may lead to a significant percentage of refusals, nonnegligible morbidity (1 in 1000) and severe intraperitoneal bleeding occurs at a frequency of 1:2500 to 1:10000[9]. Attempts have been made to substitute liver biopsies with noninvasive, low-cost, reproducible methods for the evaluation of chronic hepatitis C (CHC) fibrosis[10]. Some serum markers, such as aspartate aminotransferase-to-platelet ratio index (APRI) and fibrosis-4 index (FIB-4) have been described to correlate with liver fibrosis[11].

Radiological noninvasive method based on transient elastography (TE) is a technology that is based on liver stiffness measurement (LSM) evaluated by the propagation velocity of shear waves generated in liver tissue[12]. In addition, diffusion-weighted magnetic resonance imaging (DW-MRI) is a noninvasive method that enables the measurement of the microscopic motion of water in tissue which can be recorded in the liver within an apnea period. It measures the apparent diffusion coefficient (ADC) of water, a parameter dependent on the tissue structure. A decrease in the ADC was reported with an increase in liver fibrosis[13].

The aim of the current study was to evaluate TE and DW-MRI as noninvasive tools in evaluating liver fibrosis compared to liver biopsy in children with CHC.

MATERIALS AND METHODS

Study population

This prospective cross-sectional study initially recruited 100 children with chronic HCV infection attending the outpatient clinics of the Pediatric Hepatology, Gastroenterology, and Nutrition departments (National Liver Institute, Menoufia University, Al Minufiyah, Egypt) over a period of 3 years. The study protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki (6^th revision, 2008) and was approved by the institutional review board of the National Liver Institute (NLI-IRB 00003413 FWA0000227) of Menoufia University (Approval No. 0035). Written informed consent was obtained from the parents/legal guardians of the minors. All children were subjected to complete clinical examination and investigations to exclude any diseases other than chronic HCV. These included routine liver function tests, anti-HCV antibodies, quantification of HCV-RNA, other viral markers such as hepatitis B virus surface antigen, autoantibodies, protein electrophoresis, and ceruloplasmin level. Patients with other chronic liver disease associated with chronic HCV were excluded. Of the recruited children, 64 completed the study and 36 were excluded (20 children did not fulfill the complete set of investigations and 11 patients rejected liver biopsy after providing initial approval).

Serum markers of fibrosis

APRI was calculated according to the equation: APRI = [(aspartate aminotransferase (AST)/upper limit of normal)]/platelet count (10⁹/L) × 100]. In addition, FIB-4 was calculated according to the equation: FIB-4 = [(age × AST)/[(platelet × alanine aminotransferase (ALT)^½][11].

Liver biopsy

Liver biopsies were performed under sedation and local anesthesia using the Tru-Cut needle 14 G. Formalin-fixed, paraffin-embedded specimens were examined after staining with hematoxylin and eosin, Masson’s trichrome, reticulin, Perl’s stains, Prussian blue and picrosirius red[14]. Significant fibrosis was defined as Ishak score of F3 or more[15].

TE

TE was performed on the right lobe of the liver using the standard M probe through the intercostal space. LSM was performed using the FibroScan apparatus (Echosens, Paris, France), which consists of a 5-MHz ultrasound transducer probe mounted on the axis of a vibrator. Mild amplitude and low frequency vibrations (50 Hz) are transmitted to the liver tissue inducing an elastic shear wave that propagates through the underlying liver tissue. The velocity of the shear wave, as measured in kilopascals (kPa), reflects tissue stiffness. Faster shear wave propagation indicates stiffer tissue[16].

DW-MRI

MRI was performed as previously described[17]. Briefly, younger children had sedation after fasting for 4-6 h. Cloral hydrate was given at a dose of 70-80 mg/kg before MRI by 30 min for younger children. Older children had MRI without sedation. MRI was performed using T1.5 T MR unit (Ingenia, Philips Best, Netherlands) using bipolar diffusion encoding gradient. Routine axial T1- and T2-weighted images were obtained. DW-MRI was done using a single shot echo-planar imaging with automatic reconstruction of ADC map.

MRI analysis

MRI interpretation was performed by one radiologist expert in MRI for 25 years (AAA). Liver and spleen ADC was automatically calculated using three consecutive slices away from vascular and biliary elements. The final ADC was represented by the mean of the three readings. Length, width and thickness of liver and spleen were also measured[17]. Normalized liver ADC was calculated as the ration of liver ADC to spleen ADC[18].

Statistical analyses

Quantitative data are expressed as the mean ± standard deviation. The statistical significance was tested according to data normality by either the t-test or Mann-Whitney U test when comparing two groups, and by either analysis of variance or Kruskal-Wallis test when comparing multiple groups. For qualitative data, significance was tested by χ² test or Fisher’s exact test accordingly. Correlation was tested by Spearman test. Results were considered significant if P < 0.05. The sample size was calculated to be 60, using open epi (https://www.openepi.com/SampleSize/SSPropor.htm) with the prevalence of HCV in Egyptian children at less than 1%[19] and a confidence interval of 99.9. The cutoff values for optimal clinical performance of LSM and liver ADC were determined from the receiver-operating characteristic curve. The diagnostic performance was expressed as sensitivity and specificity percentages. The statistical review of the study was performed by a biomedical statistician. Statistical analysis was performed using SPSS, version 21 (IBM Corp., Armonk, NY, United States).

RESULTS

Study population characteristics

This study included 64 children with chronic HCV. Their age ranged between 4 years and 18 years with a mean of 12.01 ± 3.95 years and 65.4% were males. More than half of the patients (52%) had family history of HCV infection. Nearly 25% of them had hepatomegaly and/or splenomegaly. None was icteric. Most patients (82.8%) had Ishak fibrosis score of F1 or F2 while only 17.2% had F3. Other basic characteristics were as show in Table 1.

Table 1 Clinical, laboratory, and histopathological characteristics.

Parameter	Value
Hepatomegaly	24 (26.9)
Splenomegaly	19 (23.5)
Ascites	0.0
Aspartate transaminase (U/L)	49.72 ± 42.67
Alanine transaminase (U/L)	58.36 ± 56.40
Prothrombin time (second)	13.01 ± 1.16
Hemoglobin (g/dL)	12.42 ± 1.50
White blood cells (× 103/mm3)	7.04 ± 2.46
Platelets (× 103/mm3)	267.96 ± 102.50
PCR (U/mL	827319.64 ± 2415665.23
Fibrosis stage
F1	17 (26.5)
F2	36 (56.3)
F3	11 (17.2)
Activity grade
Absent: A0	1 (16.0)
Minimal: A1-A3	14 (21.9)
Mild: A4-A8	44 (68.8)
Moderate: A9-A12	5 (7.8)
Severe: A12-A18	0 (0)

Data are n (%) or mean ± standard deviation. PCR: Polymerase chain reaction.

Elevated vs normal transaminases according to the different fibrosis stages

Despite different fibrosis stages, more than half of the patients had normal transaminases while the others had elevated transaminases (Figure 1). Of those with F1, F2 and F3, normal AST was found in 7 (63.6%), 20 (55.5%), and 8 (72.7%) respectively while the remaining patients had elevated AST. Similarly, normal ALT was found in most patients as indicated in Figure 1.

Figure 1 Normal vs elevated transaminases according to different fibrosis stages. A: Aspartate transaminase (AST); B: Alanine transaminase (ALT). The dashed line represents the upper limit of normal (ULN).

Serological and radiological parameters in different fibrosis stages

LSM and liver ADC were the only parameters with significant difference among different fibrosis stages (P = 0.034 and P = 0.039 respectively), while spleen ADC, APRI, and FIB-4 were nonsignificant (Table 2). Comparing LSM and liver ADC between individual fibrosis stages showed that both LSM and liver ADC were nonsignificant when comparing F1 vs F2. On the other hand, LSM was significantly higher in F3 when compared to F 2 (P = 0.035) and liver ADC was significantly lower in F3 when compared to F2 (P = 0.017). In addition, normalized liver ADC showed significant difference between F1 and F2, and between F2 and F3 (Table 2).

Table 2 Liver stiffness, and liver and spleen apparent diffusion coefficient in different fibrosis stages.

Noninvasive parameter	F1, n = 17	F2, n = 36	F3, n = 11	aP value
LSM (kPa)	4.04 ± 1.41	4.98 ± 1.63	7.20 ± 5.21	0.034
	bP = 0.096	cP = 0.035
Liver ADC (mm2/s)	1.12 ± 0.24	1.15 ± 0.23	0.95 ± 0.22	0.039
	bP = 0.0787	cP = 0.017
Spleen ADC (mm2/s)	1.12 ± 0.23	1.01 ± 0.15	0.99 ± 0.26	0.271
Normalized liver ADC	1.01 ± 0.22	1.16 ± 0.23	0.99 ± 0.25	0.036
	bP = 0.033	cP = 0.016
APRI score	0.44 ± 0.42	0.48 ± 0.29	0.38 ± 0.37	0.332
FIB-4 score	0.34 ± 0.30	0.34 ± 0.15	0.22 ± 0.16	0.083

^aP: Significance among the three groups by Kruskal-Wallis test.

^bP: F1 vs F2.

^cP: F2 vs F3.

ADC: Apparent diffusion coefficient; APRI: Aspartate aminotransferase-to-platelet ratio; FIB-4: Fibrosis-4 index; LSM: Liver stiffness measurement.

Correlation of fibrosis and activity with the studied parameters

There was a significant correlation between LSM and both of fibrosis stage (P = 0.008) and activity grade (P = 0.002), while there was no significant correlation with any of the other studied serological parameters, and liver and spleen ADC (Table 3).

Table 3 Correlation of fibrosis and activity with different studied parameters.

Parameters	Fibrosis		Activity
	R value	P value	R value	P value
Age in yr	0.068	0.594	0.159	0.208
Aspartate transaminase (U/L)	-0.044	0.751	0.192	0.160
Alanine transaminase (U/L)	0.092	0.501	0.242	0.073
Prothrombin time (second)	-0.066	0.637	-0.184	0.188
Hemoglobin (g/dL)	0.041	0.764	-0.050	0.718
White blood cells as × 103/mm3	0.680	0.620	0.015	0.913
Platelets as × 103/mm3	0.047	0.731	-0.032	0.815
LSM (kPa)	0.338	0.008	0.385	0.002
Liver ADC (mm2/s)	-0.299	0.086	-0.281	0.840
Spleen ADC (mm2/s)	-0.212	0.145	-0.019	0.897
Normalized liver ADC	-0.064	0.653	-0.148	0.300
APRI score	-0.003	0.984	-0.088	0.549
FIB-4 score	-0.128	0.376	0.148	0.305

ADC: Apparent diffusion coefficient; APRI: Aspartate aminotransferase-to-platelet ratio; FIB-4: Fibrosis-4 index; LSM: Liver stiffness measurement.

Performance of LSM and ADC in discriminating significant fibrosis

LSM and liver ADC were able to discriminate significantly F3 from lower fibrosis stages with AUC of 0.700 and 0.747 respectively with a better performance of liver ADC. Sensitivity was equal for both parameters (63.3%) but with higher specificity of liver ADC (87.3%) compared to LSM (75.5%) (Figure 2).

Figure 2 Diagnostic performance of noninvasive parameters of fibrosis in discriminating significant fibrosis (≥ F3). A: Liver stiffness measurement; B: Liver apparent diffusion coefficient (ADC). AUC: Area under the curve.

DISCUSSION

The current study evaluated TE and DW-MRI as noninvasive tools in evaluating liver fibrosis compared to liver biopsy in children with CHC. Applying such a comparison in children needs to be addressed as it is very crucial in pediatric age group to find a less invasive alternative to detect and follow-up fibrosis[20]. In addition, guidelines on this topic are an unmet need in hepatology[21].

We found that APRI and FIB-4 did not significantly differ with different fibrosis stages or correlated with fibrosis. Contrary to our results, Güzelbulut et al[11] reported that APRI and FIB-4 were accurate noninvasive blood tests to predict the presence or absence of significant fibrosis and cirrhosis in adult patients with CHC. In addition, Barakat et al[22] studied 166 Egyptian children with CHC and found that APRI and FIB-4 could discriminate different stages of fibrosis.

In agreement with our results, other studies reported that APRI was not useful in predicting fibrosis in chronic viral hepatitis[23,24]. ElShahawy et al[20] reported a low specificity (57.1%) for APRI in predicting liver fibrosis in children with chronic HCV and a significant number of patients could not be correctly classified by this method[25].

The reason for the insignificance of APRI and FIB-4 in our study may be the presence of considerable number of patients (nearly 50%) with normal transaminases regardless the stage of fibrosis. Such markers are dependent on transaminases levels. In addition, previous studies included patients with higher fibrosis stages. Serum AST/ALT levels can fluctuate in those with chronic HCV and studies have demonstrated that pediatric patients can have normal AST/ALT despite histologic evidence[26]. El-Raziky et al[27] reported that ALT was high in 50% of patients and histopathological abnormalities were found in 75% of patients. This means that liver enzymes in chronic HCV infection do not reflect histopathological abnormalities in most cases and normal transaminases are frequently encountered in chronic HCV Egyptian patients[28].

Although TE measures the stiffness of a liver parenchymal volume 100 times bigger if compared to liver biopsy, it cannot enable segmental liver assessment and the presence of ascites or obesity represents an obstacle to its performance[29]. By contrast, DW-MRI evaluates the whole liver volume and enables segmental ADC measurements; thus, providing information about the most severely affected liver segment. In addition, it can be performed in obese patients or even in the presence of ascites[30]. ADC holds promise as a noninvasive imaging technique for assessing cirrhosis in patients with chronic viral hepatitis[31]. The performance is good when predicting severe fibrosis[32].

Our results revealed that there was no significant difference in liver ADC when comparing lower fibrosis stage (F1 vs F2), while the difference was significant when comparing higher fibrosis stage (F2 vs F3). Liver ADC did not correlate with fibrosis. Similar results was reported by Serag and Ragab[33]. The results of several studies have shown that the ADC values of cirrhotic patients are lower than those of noncirrhotic patients or of healthy volunteers[34,35], but the usefulness of the ADC in evaluating the intermediate fibrosis stages remains questionable.

Ozkurt et al[36] detected decreased ADC values in patients with hepatic fibrosis compared to patients with no clinical or biochemical findings of liver disease and there was a trend towards decrease in hepatic ADC values with an increasing degree of fibrosis. Taouli et al[37] assessed seven control subjects and 23 patients with hepatitis related liver disease. Although there was a significant difference in the ADC of the F0 and F1 groups compared with the ADC of the F2-F4 groups, there was much overlap in the ADC values of individual patients in each group. Boulanger et al[38] could not find a difference between the ADC values in 18 HCV patients with fibrosis due to chronic HCV and 10 healthy controls. Sandrasegaran et al[34] showed a significant difference in the ADC values of nonfibrotic (F0) and cirrhotic (F4) patients. However, it could not be used to reliably distinguish the intermediate stages of fibrosis.

Multiple studies have shown that normalized ADC, using the spleen as reference organ, improves the diagnostic performance in assessing liver fibrosis than using liver ADC alone. The spleen may be an ideal reference organ because it maintains a relatively stable ADC even in the setting of liver disease[39]. Our results showed that normalized liver ADC added no further significance over liver ADC.

The current study showed that LSM values had a significant correlation with liver fibrosis in the biopsy (P = 0.008), as well as with the activity grade (P = 0.002). There was no significant difference when comparing F1 vs F2, while the significance detected when comparing F2 vs F3 (P = 0.035).

Several studies have demonstrated the significant performance of LSM in predicting fibrosis[40,41]. Others have demonstrated discordance of LSM values with fibrosis stage and recommend that patients with a high LSM need proper attention for cirrhosis, even if liver biopsy does not reveal cirrhosis[42]. The usefulness of LSM appears not only in cross-sectional evaluation of fibrosis but also in longitudinal fibrosis follow up. Alswat et al[43] reported that the clearance of HCV with DAAs is associated with significant improvement in fibrosis as assessed by LSM, which supports the concept of post-treatment fibrosis regression.

The limitations in our study were the limited number of patients due to dropouts, the lack of follow-up of LSM and liver ADC after the antiviral therapy, and the absence of higher stages of fibrosis.

CONCLUSION

Our study demonstrated that LSM and liver ADC were able to discriminate significant fibrosis (F3) compared to lower fibrosis stages (F1 and F2) with a better performance of liver ADC. These parameters may be of help in evaluating disease severity and their use for monitoring fibrosis post-treatment is worthy. A future multicenter study including a larger population is strongly recommended.