Shear Wave Elastography in the Assessment of Liver Changes in Children with Cystic Fibrosis

Abstract

A standard ultrasound examination of the liver was performed in 232 children. It was supplemented by a two-dimensional shear wave elastography. There were 200 healthy children aged 3 to 18 years (control group) and 32 patients with cystic fibrosis aged 2 to 17 years (study group) among them. The procedure was carried out by means of Aixplorer device (Supersonic Imagine, France) using a convex sensor operating in the 1–6 MHz frequency range. Ten measurements of Young modulus values were carried out in different segments of the right lobe of the liver followed by data averaging. In patients with cystic fibrosis, the values of Young modulus were significantly higher than in healthy children (Emean median: 6.50 and 5.00 kPa, interquartile range: 5.62–7.52 and 4.70–5.38 kPa, respectively (p < 0.001). In patients with severe cystic fibrosis, the values of Young modulus were significantly higher compared to patients suffering from moderate disease (Emean median: 7.30 and 5.90 kPa, interquartile range: 6.20–10.70 and 5.20–6.75 kPa, respectively (p < 0.002). Shear wave elastography is a non-invasive technique that can be successfully used in a comprehensive ultrasound assessment of the liver in children with cystic fibrosis to facilitate the diagnosis and monitoring of fibrous changes.

1. Introduction

Recently, ultrasound elastography has become perhaps the most important achievement in the evolution of non-invasive techniques, particularly ultrasound examination, to assess the condition of the liver in general. This is a method of qualitative and quantitative analysis of elastic properties of tissues, which makes it possible to evaluate the elastic properties of tissues during a conventional ultrasound examination by measuring the values of shear wave velocity (m/s) or Young modulus (kPa) in the organs and tissues of interest [1].

As a rule, the stiffness in pathological tissues is more pronounced compared to the adjacent healthy tissues, and this fact is registered by ultrasound elastography of different types. According to the international guidelines, ultrasound elastography methods are divided into compressive elastography (SE) and shear wave elastography (SWE) [2, 3, 4, 5]. SWE methods measure the velocity of shear waves generated in the tissues by an external mechanical shock (transient elastography, TE) or an electronic impulse (ARFI). The advantages of ARFI-based elastography methods (point elastography, pSWE and two-dimensional elastography, 2D-SWE) are that they are fast and integrated into the ultrasound diagnostic system, which enables to perform grayscale navigation. When conducting 2D-SWE, we get not only quantitative data in the form of digital values of the shear wave velocity but also qualitative information, since the areas with different values of the Young modulus are mapped in different colours. It is the digital values of indicated parameters that determine the colour in the area of interest [1, 6].

Most guidelines and recommendations for the clinical use of elastography (EFSUMB, WFUMB) focus on the assessment of diffuse liver disease in adults [2, 3, 4, 5, 6, 7, 8, 9]. However, recommendations for adult patients cannot be immediately used in paediatric practice, taking into account the peculiarities of paediatric patients, i.e., restless behaviour during the examination of young children, small intervals between meals in infants, difficulties with breath-holding. These factors can affect the reproducibility of measurements and the accuracy of diagnostics in paediatrics. Psychological, anatomical and morphological features of children make the technique of elastography even more complex than a conventional ultrasound examination. Therefore, ultrasound elastography is less studied in children than in adults. Nevertheless, there are more and more reports from different research groups about the use of elastography to assess liver stiffness in healthy children. Special attention is paid to the age and gender characteristics of stiffness, the dependence of values on body mass index, use of sedatives and food intake. The technique of the procedure is discussed concerning the position of the patient during the examination, the choice of the sensor, the zone and the number of measurements, the ambient conditions during the procedure. There are also works where elastography is used in the assessment of the spleen, thyroid gland, renal parenchyma, intestines and muscles [6, 10, 11, 12, 13].

Numerous studies in adult patients have demonstrated that ultrasound elastography is a useful non-invasive method for diagnosing liver fibrosis. Preliminary findings using TE, pSWE and 2D-SWE have also shown that they are all feasible and can be used to assess liver fibrosis of various aetiology in children [6, 10, 11, 13]. Various infectious agents can act as an etiological factor causing fibrosis and cirrhosis of the liver in children, such as hepatitis B, C, D, G viruses, cytomegalovirus, Epstein-Barr virus, as well as autoimmune liver diseases, cystic fibrosis, metabolic diseases and others. Regardless of the aetiology, cirrhosis results in lethal outcome of patients due to the development of complications, i.e., bleeding from the oesophageal varices, ascites, encephalopathy, hemorrhagic syndrome.

Liver biopsy, being an invasive procedure, is less acceptable for children due to the need for general anaesthesia, as well as physical and emotional impact on the child. This procedure can cause a number of complications, such as pain syndrome, profuse bleeding, formation of subcapsular hematomas of the liver, development of biliary peritonitis, etc. [14, 15]. Non-invasive methods are extremely important in paediatrics, especially when repeated examinations are necessary, for example, during follow-up of patients with chronic liver diseases, when elastography can be used additionally to ultrasound examination and laboratory data facilitating observation of children with chronic liver diseases [10, 11, 13].

2. Pathophysiology of fibrotic changes in the liver in cystic fibrosis

Cystic fibrosis is one of the most common monogenic diseases with an autosomal recessive type of inheritance and multiple organ manifestations. In most countries of Europe and North America, cystic fibrosis prevalence ranges from 1:2,000 to 1:4,000 newborns, in Russia, it is 1:8000–1:10,000 newborns. The disease is caused by the gene mutation of the transmembrane regulator of cystic fibrosis and is characterised by damage to the external secretion glands and severe respiratory disorders. However, due to the increased life expectancy of patients with cystic fibrosis, liver damage becomes an important clinical manifestation that determines the prognosis and quality of life [16, 17, 18].

Cystic fibrosis is no longer considered exclusively a childhood disease since the current life expectancy of such patients is more than 35 years [19]. According to different authors, the incidence of clinically apparent liver lesions in cystic fibrosis (cystic fibrosis-associated liver disease – CFLD) varies from 27 to 35% [18]. Symptoms of liver fibrosis of varying degrees are revealed in almost all patients with cystic fibrosis, and 5–10% of patients develop biliary cirrhosis of the liver with portal hypertension syndrome requiring surgical intervention [20, 21]. In the general list of causes of death in cystic fibrosis, liver cirrhosis is in the second place after bronchopulmonary complications and is 2.3–2.5% [20, 22].

The hepatobiliary system failure in cystic fibrosis is a direct consequence of a basic genetic defect. The CFTR protein is responsible for the pathogenesis functions as a channel of chloride ions [20, 23]. Insufficiency of the function of the channel of chlorine ions of cells lining the intrahepatic and extrahepatic bile ducts and gallbladder results in dehydration of hepatic secretions, i.e., they become adherent and poorly soluble [24]. Consequently, hepatocellular and canalicular cholestasis develops, which leads to a number of undesirable reactions, namely, delay of hepatotoxic bile acids, production of inflammatory mediators, cytokines and free radicals, increased lipid peroxidation and damage to cellular membranes, an excessive inflow of bile into the blood and tissues [25]. According to the clinical and morphological principle, liver cirrhosis in cystic fibrosis refers to biliary cirrhosis with obstruction of the intrahepatic biliary tract; microscopically – to multilobular cirrhosis; according to etiological characteristic – to cirrhosis caused by genetic metabolic disorders. Liver damage is characterized by chronic inflammatory cell infiltration and bile duct proliferation. These alterations are initially of a limited focal nature, but then they progress and lead to multilobular cirrhosis and portal hypertension. The process can slowly develop without pronounced clinical and biochemical manifestations, but it is irreversibly progressing [21, 22, 23, 26].

To date, there is no consensus on the risk factors and the rate of development of fibrotic liver changes in cystic fibrosis. Prematurity, low birth weight and prolonged parenteral nutrition are indicated as the causes of cystic fibrosis progression. Also, the severity of liver damage is associated with recurrent sepsis (including catheter sepsis) and bacterial load. Although it is believed that liver damage is more common in patients with severe mutations pertaining to classes I–III, genotype-phenotype correlation, which predicts the effect of mutation on the clinical expression of CFLD, is not possible at this stage. So, the clinical course in patients with diagnosed cystic fibrosis and the same mutation of the CFTR gene may be different [27]. It is still unclear why only a small number of patients with the same severe mutations develop CFLD symptoms. Some authors consider age at the time of diagnosis, male gender, intestinal obstruction of meconium in the anamnesis, external secretory insufficiency of the pancreas to be important factors [28, 29].

Early diagnosis of liver damage is complicated by a prolonged subclinical phase and lack of a reliable diagnostic technique: biochemical indices of liver failure (increased bilirubin level, decreased albumin concentration and increased prothrombin time) appear late, when severe liver failure has developed. Therefore, all patients with cystic fibrosis should be carefully monitored for the occurrence of this complication in the first decade of their life [18, 25, 26]. Regular examination of patients, biochemical tests and imaging methods of examination are of utmost importance. At the same time, a normal ultrasound picture cannot exclude the presence of fibrosis [30].

Despite the high informativeness of morphological methods, diagnostic liver biopsy for all patients with cystic fibrosis, and especially performed repeatedly, cannot be justified, primarily because of its traumatic nature. In addition, due to inhomogeneous liver damage, a biopsy may underestimate the severity of the lesion. The procedure is indicated when the diagnosis is doubtful or to confirm the findings before liver transplantation [17, 20].

That is why the efforts of researchers are aimed at finding such diagnostic non-invasive methods (especially important in paediatric practice), which will be informative, accessible, capable of detecting liver changes and carrying out dynamic monitoring of the fibrous process. The use of shear wave elastography in the early diagnosis of liver diseases in patients with cystic fibrosis becomes particularly relevant since a number of authors point to the reversible nature of such conditions as fatty hepatosis and cholestasis following treatment [17, 25]. The aim of our investigation was to study the stiffness of the liver using shear wave elastography in children suffering from cystic fibrosis.

3. Materials and methods

The study was approved by the Ethics Committee of ‘The Russian Medical Academy of Continuing Professional Education’ of the Health Ministry of the Russian Federation and the Ethics Committee of the Chelyabinsk Regional Children Clinical Hospital. Written informed consent of legal representatives was obtained for all patients. Liver biopsy was not included in the algorithm of examination of patients with cystic fibrosis. Ultrasound examination was performed by an Aixplorer device (Supersonic Imagine, France) with a broadband convex sensor operating in the frequency range of 1–6 MHz. A standard ultrasound examination of the liver, supplemented by two-dimensional shear wave elastography, was performed in 232 children. There were 200 healthy children aged 3 to 18 years (control group) and 32 patients with cystic fibrosis aged 2 to 17 years (study group) among them.

The control group included healthy children. The following criteria were taken into account:

• height and weight of each child in the range from the 5th to the 95th percentile of the age norm [31];

• absence of liver disease and (or) congestive heart failure in the history;

• absence of inflammatory changes according to general and biochemical blood tests (signs of cholestasis, cytolysis);

• absence of pathology of the liver, biliary tract, pancreas and spleen according to ultrasound in grayscale and Doppler exam (colour Doppler mapping, pulse-wave Doppler) modes;

• calm behaviour of a child during the study.

The diagnosis of cystic fibrosis was established on the basis of a comprehensive clinical and laboratory examination with history data analysis, DNA diagnostics with genotype specification, and it was confirmed by a positive sweat test. The course of the disease in 17 (53.1%) children was regarded as moderate (subgroup I), in 15 (46.9%) children – as severe (subgroup II). The severity of the disease course was assessed by the Schwachman-Brasfield scale, modified by S.V. Rachinsky and N.I. Kapranov. This scale takes into account the general activity of the patient, state of his nutrition and physical development, clinical manifestations of the disease, as well as the results of X-ray examination [32].

After a standard ultrasound examination of the abdominal organs in the grayscale mode, the stiffness of different segments of the right liver lobe was measured in the areas free of vascular structures at a depth of 3–5 cm from the capsule. The study was performed fasting, the patients breathing calmly, the older children holding breath for no more than 10 seconds or during shallow inhalation, in a supine position. The sensor was positioned perpendicular to the body surface with minimal manual pressure, using subcostal, intercostal and epigastric approaches. The area of interest (colour window) was selected, the image stabilization was set, the measurement was considered successful if more than 90% of the colour window was filled with colour. Ten measurements of the average value of the Young modulus (Emean) (kPa) were performed, and according to the results, the arithmetic mean value of Emean was calculated. Examples of liver stiffness assessment in patients of both groups are shown in Figures 1–4.

Figure 1.

An example of stiffness assessment of unchanged liver parenchyma in a healthy child: B-mode and two-dimensional shear wave elastography mode. The results of one of 10 measurements. Emean = 4.4 kPa. The child is 10 years old.

Figure 2.

An example of liver stiffness assessment in a child with cystic fibrosis: B-mode and two-dimensional shear wave elastography mode. The results of one of 10 measurements. Emean = 5.6 kPa. The child is 13 years old.

Figure 3.

An example of liver stiffness assessment in a child with cystic fibrosis: B-mode and two-dimensional shear wave elastography mode. Emean = 11.2 kPa. The child is 11 years old.

Figure 4.

An example of liver stiffness assessment in a child with cystic fibrosis: B-mode and two-dimensional shear wave elastography mode. The results of one of 10 measurements. Emean = 27.4 kPa. The child is 12 years old.

4. Statistical analysis

Statistical data processing was performed by the IBM SPSS Statistics 19 pack. Most quantitative values were not within the normal distribution, so nonparametric statistic methods were applied. All quantitative data were presented as M (mean), m (standard error of the mean), σ (standard deviation), median (50th percentile), 25th–75th percentiles and minimum and maximum values. Quantitative parameters were compared using Mann–Whitney criterion, qualitative ones were compared using Fisher criterion of accuracy. The differences were considered significant at P ≤ 0.05.

5. Results

Two-dimensional shear wave elastography was performed in 200 (control group) conditionally healthy children aged 3 to 18 years who underwent ultrasound examination for reasons unrelated to hepatobiliary diseases. Routine ultrasound examination did not reveal any changes in the liver, spleen and gallbladder in children of the control group. The technique and results of two-dimensional liver elastography performed in healthy children of different age groups had been published by us earlier [33, 34]. Statistical processing of the results of the previous study allowed to establish the mean value of Young modulus in the group of healthy children, i.e., 5.01 ± 0.03 kPa, the median of the average Emean value was 5.00 kPa (4.70–5.38). We did not find statistically significant gender differences in liver stiffness in children of the control group [33, 34].

In the group of children with cystic fibrosis, hepatomegaly was revealed in the vast majority of children – 31 patients (96.9%), splenomegaly – in 10 patients (31.3%). In 13 children (40.6%), changes in the gallbladder were detected in the form of a wall thickening of more than 4 mm, or its decreased size. Six children had signs of portal hypertension (18.8%), which was manifested by hepatosplenomegaly, portal vein dilation with decreased linear blood flow velocity and oesophageal varices according to gastroduodenoscopy. The ultrasound picture of the liver was presented as unchanged parenchyma. Also, a diffuse or inhomogeneous increase of echogenicity, heterogeneity of the parenchyma with a pronounced vascular pattern, cirrhotic nodes with depletion and deformation of a typical vascular tree, pronounced periportal fibrosis were determined. The elastography picture of the liver in children with cystic fibrosis was various. It was represented both by homogeneous colouring of the colour window in dark blue or blue tones, with the absence of areas of increased stiffness, in that case, the qualitative characteristics did not differ from the control group of healthy children (Figures 1 and 2), and also, blue-green with yellow areas, as well as red-orange colouring of the area of interest in children with pronounced ultrasound signs of cirrhotic liver changes (Figures 3–5).

Figure 5.

An example of liver stiffness assessment in a child with cystic fibrosis: B-mode and two-dimensional shear wave elastography mode. The results of 4 measurements. Emean = 38.0 kPa. Median – 36.8 kPa. The child is 14 years old.

The values of the liver parenchyma stiffness in the studied groups and subgroups are shown in Tables 1 and 2.

Groups	M ± m	σ	Median	25th–75th percentile	Minimum – maximum values
Control group (n = 200)	5.01 ± 0.03	0.49	5.00	4.70–5.38	3.00–6.30
Study group (n = 32)	7.11 ± 0.44	2.48	6.50	5.62–7.52	4.30–16.90

Table 1.

Stiffness (Emean, kPa) of liver parenchyma in children of the studied groups: healthy children and children with cystic fibrosis.

Note: Comparison of two groups, p < 0.001.

Subgroups	M ± m	σ	Median	25th–75th percentile	Minimum – maximum values
Moderate course (n = 17)	5.98 ± 0.23	0.94	5.90	5.20–6.75	4.30–7.90
Severe course (n = 15)	8.39 ± .79	3.05	7.30	6.20–10.70	5.60–16.90

Table 2.

Stiffness (Emean, kPa) of liver parenchyma in children with cystic fibrosis of various severity (n = 32).

Note: Comparison of two subgroups, p < 0.002.

Significant differences in the values of Young modulus (Emean) were obtained when comparing the values of the study and control groups: median Emean – 6.50 and 5.00 kPa, interquartile range – 5.62–7.52 and 4.70–5.38 kPa, respectively (P< 0.001). Most children suffering from cystic fibrosis had some degree of change in liver stiffness. In the group of patients with cystic fibrosis, only 9 (28.1%) children had Young modulus values that did not exceed those in healthy children. Analysis of the data obtained showed that the values of Young modulus in the group of patients with a severe course of the disease was significantly higher than in patients with a moderate course of cystic fibrosis: median Emean – 7.30 and 5.90 kPa, interquartile range – 6.20–10.70 and 5.20–6.75 kPa, respectively (P< 0.002).

6. Discussion

In the available publications, we were able to find only a few works devoted to the shear wave elastography assessment of liver stiffness in a group of children suffering from cystic fibrosis. It should be noted that the paediatric age group is the most interesting for assessing liver changes since cystic fibrosis-associated liver diseases develop in early childhood, and new cases are rare after the age of 20. In 2009, P. Witters et al. [35] used transient elastography to study liver stiffness in 66 patients with cystic fibrosis [35]. The obtained elastometry findings were compared with those of a control group consisting of 59 people (98th percentile (or M + 2σ), Young modulus for children under 12 years of age was 5.63 kPa, 12 years and older – 6.50 kPa). In our study, the maximum value of 6.30 kPa (with a median of 5.00 kPa) in the control group was used as a threshold. Such clinical manifestations as hepatosplenomegaly and changes in biochemical parameters were taken into consideration in the patients of the study group. Only one patient underwent a liver biopsy. The study showed an increase in stiffness in patients with clinical manifestations, such as palpable hepatosplenomegaly (11.07 ± 5.51 kPa (n = 6) vs. 5.08 ± 3.45 kPa (n = 60), P < 0.0001), biochemical (7.40 ± 3.10 kPa (n = 7) vs. 5.42 ± 4.08 kPa (n = 59), P = 0.013) and ultrasound (8.19 ± 5.96 kPa (n = 23) vs. 4.27 ± 0.94 kPa (n = 41), P = 0.0001) signs of liver damage. The mean value of liver stiffness in children with cystic fibrosis was 5.63 ± 4.02 kPa [35]. In our study, the median of Young modulus was 6.50 kPa in patients with cystic fibrosis. We got fairly close values, although the two studies were conducted on different devices. The stage of liver fibrosis during puncture biopsy was not evaluated in both studies.

In 2012, the research group of L. Monti et al. [36] evaluated liver stiffness using shear wave point elastography in the group of 75 children with cystic fibrosis. Measurements of the shear wave velocity were carried out in the right liver lobe, followed by averaging of 10 indices. The patients underwent ultrasound examination, gastroscopy and analysis of biochemical parameters. Liver biopsy was not performed. The study found that the median shear wave velocity was significantly higher in patients with clinical, biochemical and ultrasound signs of liver damage, than in patients with their absence. The velocity increased as the signs of decompensation of portal hypertension increased, that is, with the progression of fibrous changes. The median in patients with portal hypertension, splenomegaly and oesophageal varices was 1.30, 1.54 and 1.63 m/s, respectively (P < 0.001) [36].

A study by T. Canas et al. conducted in 2015 [37] included 72 patients with cystic fibrosis aged 9 months to 18 years. The stiffness of the liver and spleen was assessed using shear wave point elastography after the routine ultrasound examination. The biopsy was not included in the protocol. As in our study, a convex sensor was used, five measurements of the shear wave velocity were carried out in the right and left liver lobes (intercostal and subcostal access) and in the spleen. The obtained data were compared with the results of the control group (n = 60). The shear wave point elastography revealed a significant increase in the shear wave velocity in patients with cystic fibrosis-associated liver disease compared with healthy children and patients with cystic fibrosis without liver involvement (P = 0.003). The diagnosis of ‘cystic fibrosis-associated liver disease’ was made using non-invasive Colombo criteria [18]. The threshold value of the shear wave velocity for the diagnosis of cystic fibrosis-associated liver disease was 1.27 cm/s (measurement in the right liver lobe) with 56.5 % sensitivity, 90.5% specificity, AUC 0.746 [37].

Reports on the study of liver stiffness by two-dimensional shear wave elastography using the Aixplorer device (Supersonic Imagine, France) in children suffering from cystic fibrosis could not be found. We carried out measurements on segments of the right liver lobe using a convex sensor to standardise the study protocol as the linear sensor measurements were not always considered technically possible in older children. The analysis of the data obtained by us, as well as other research groups, revealed that the indices of hepatic tissue stiffness in patients with cystic fibrosis had significantly higher values than in healthy children. Liver stiffness in children with a severe course of the disease was more apparent than in children with a moderately severe course. As the clinical, biochemical and ultrasound signs of liver damage increased, the stiffness indices accordingly increased. A clear insight into the degree of damage, stage and rate of fibrosis progression in chronic liver disease in children is important when making a comprehensive decision when to resort to surgical methods of treatment and to liver transplantation. Taking into consideration the above mentioned and based on the results of our study, we consider it appropriate to monitor the indices of Young modulus in patients with cystic fibrosis, as it enables to identify a group of children with high rates of fibrosis.

7. Conclusions

Shear wave elastography can be successfully used in the comprehensive assessment of liver damage in children with cystic fibrosis to facilitate diagnosis and dynamic monitoring of the severity of fibrous degenerations in the parenchyma and sampling of patients in need of liver transplantation.