Shear wave elastography and shear wave dispersion correlated to biopsy at the scheduled follow‐up of pediatric liver grafts

It is unknown how shear wave dispersion (SWD) is displayed in pediatric liver transplant recipients and not fully elucidated how ultrasound shear wave elastography (2D‐SWE) display within this cohort, which is important to determine to improve noninvasive surveillance of these patients. The study aimed to compare SWE and SWD values with histopathology in pediatric liver recipients.


| INTRODUC TI ON
3][4] Liver biopsy is an important tool in the diagnosis of various underlying causes of liver allograft damage, both in early and later in the follow-up.Liver transplant recipients are usually subjected to multiple biopsies over the course of their lives as part of routine follow-up. 5,6en though liver biopsy is considered the gold standard for liver tissue characterization, it is an invasive and costly procedure, often requiring anesthesia in the pediatric cohort. 710] Methods for noninvasive liver tissue characterization have improved in recent years.0][21] Even in adult liver recipients without signs of clinical complications, a recent study by Boeken et al., reported increased liver stiffness, speculating on multifactorial intra-and postprocedural changes as an underlying cause. 22wever, this study was limited by the lack of comparison to histopathology.Moreover, the range of SWE values in pediatric liver graft recipients without clinical complication has not been investigated.Shear wave dispersion (SWD), a novel ultrasound measure based on elastography technology [Canon Medical, Aplio i800], has been proposed to reflect edema and/or inflammation in liver tissue by measuring liver tissue viscosity. 23,24One pediatric feasibility study reported SWE and SWD as promising markers to rule out clinically significant fibrosis and inflammation, respectively, in a cohort of children with various liver diseases. 25Further, Sugimoto et al., reported SWD as a potential marker of allograft damage in adult liver recipients. 26

| Patient characteristics
This prospective study included 48 consecutive pediatric liver transplanted patients aged 0-18 years, who were scheduled for an elective, surveillance liver biopsy at Queen Silvia's Children's Hospital, Gothenburg, Sweden, from April 2019 through February 2023.Our hospital's surveillance program after pediatric liver transplantation includes a liver biopsy at 1, 5, 10 and 15 years after transplantation.The inclusion criteria were previous history of liver transplant surgery of any cause and consent to participate in the study provided by the child or parents/guardians.Twenty-one of the patients were part of the study investigating 2D-SWE as a marker for fibrosis in children with various chronic liver diseases 11 and 7 patients were part of a feasibility study of SWD. 25 This study was conducted according to the Declaration of Helsinki.Ethical approval (Dnr 634-18 -Dnr 2021-06760-02) was given by the regional ethics review board, and oral and written informed consent was obtained from all participants and/or their parents/guardians.All patients were part of the clinical longitudinal follow-up program at the study site, with assessments performed 6 months to 17 years after liver transplantation.

| Ultrasound-based biomarkers
In conjunction with the scheduled liver biopsy, a liver ultrasound examination with Doppler was performed.Two-dimensional SWE [Canon Medical, Aplio i800] was used to measure liver elasticity, which was performed during anesthesia (fasting >4 h) and free breathing by one of the five participating pediatric radiologists with elastography training.In cooperating patients, SWE was also performed in the awake state.
From January 2021, additional software was obtained by the radiology department, making it possible to measure hepatic SWD.SWD (m/s/kHz) was obtained simultaneously with SWE (kPa) sampling, and performed according to the latest liver elastography consensus statement by the Society of Radiologists in Ultrasound. 17tients were examined in a supine position during free breathing, with their right arm raised over their heads.Measurements were performed in the right liver lobe with an intercostal approach, applying minimal transducer (curved transducer i8CX1) pressure.In rare instances, whenever intercostal sampling was not possible, a subcostal approach was performed at a midline position.Liver measurements were performed 2.0-3.5 cm below the liver capsule using the continuous mode, where the median value of 10 registrations was recorded.

| Liver biopsy, clinical data and histopathological analysis
Immediately following the ultrasound examination, 1-2 percutaneous 18-gauge biopsies were obtained from the corresponding area of the liver.Exclusion criteria were median SWE values with IQR/ median > 30% kPa, which also indirectly resulted in the exclusion of corresponding SWD measurements.With few exceptions, all blood tests were collected on the same day as the biopsy or 1-2 days before.In the analyses, SWE and SWD values obtained during anesthesia were primarily used; however, in the few cases where these did not fulfill the inclusion criteria, measures in the awake state were used.Clinical data were also reviewed to see if the biopsy findings led to a change in each patient's treatment plan.

| Histopathology grading of fibrosis and inflammation
Two board-certified hepatopathologists (>10 years of clinical experience) scored the biopsies, blinded to the results of the ultrasound measurements.Fibrosis severity was scored according to the Batts & Ludwig classification as F0 (no fibrosis), F1 (mild fibrosis), F2 (moderate fibrosis), F3 (severe fibrosis) and F4 (cirrhosis). 27flammation was categorized according to the Batts & Ludwig classification in specimens with unspecified or chronic hepatitis as grade 0 (none), grade 1 (minimal), grade 2 (mild), grade 3 (moderate) and grade 4 (severe).The Banff scheme was used in cases of liver allograft rejection, and the Rejection Activity Index (RAI) for acute cellular rejection.For the analysis, the group of patients with F0 or F1 was classified as having low-grade fibrosis, and grade 0 or 1 inflammation as low-grade inflammation.Findings above F0-1 or grade 0-1 inflammation were classified as high-grade fibrosis or inflammation, respectively.

| Observer variation
Interobserver reliability of SWE measurements, and indirectly SWD measurements, for the physicians participating in the current study previously reported as excellent, as was intraobserver reliability measurements on previously obtained multimode cine-loops SWE sampling. 11

| Statistics
Descriptive statistics were used, and data were tested for normal distribution with n (%) presented for categorical variables and median (min; max) presented for continuous variables.For the pairwise comparison between groups of continuous variables, Fisher's nonparametric permutation test or Mann-Whitney U test was used.For the comparison between groups of dichotomous variables, Fisher's exact test was used.To test for the association between SWE and SWD values and grades of fibrosis, inflammation and liver function tests, Spearman rank correlation coefficient (rho) was used.Logistic regression analysis was performed for the variables SWE and SWD to predict the outcome.The area under the receiver operating characteristic (AUROC) curve was calculated to describe the goodness of predictors when applicable.The data were analyzed using the SAS System, version 9.4.

| RE SULTS
Of the 48 patients, two were excluded due to having SWE IQR/ median >30% kPa.During the inclusion period, three patients returned for a repeated liver biopsy and one patient returned twice.Thus, a total of 46 liver biopsies (26 from males, 56.5%) in 41 unique patients were included in the analysis.Out of 46 liver biopsies, 32 received liver segments 3 and 2 (16 were from living donors).Eight patients received a whole liver, four patients received segments 1, 4-8 and two patients received segments 2-4 (one from living donor).
The liver stiffness measurement SWE was performed in conjunction with all liver biopsies.When the ultrasound software SWD became available, it was additionally performed in the case of 21 patients with biopsy.Demographics and serological data of the patients are presented in Table 1 and the indications for liver transplant surgery are shown in Table 2.

| Shear wave elastography and fibrosis stage
Livers with high-grade fibrosis displayed a higher SWE value and greater variability, median (min; max) 6.0 kPa (4.5; 33.3), compared to those with low-grade fibrosis, median 5.7 kPa (3.2; 16.1); p = .05(Figure 1).There was a weak positive correlation between the grade of fibrosis and SWE values (r = .3;p = .05).The AUROC differentiating F0-1 from F2-4 was 0.62 (95% CI: 0.45-0.78).To minimize false-positive results, the cut-off value with highest sensitivity was chosen according to the J-max from Youden statistics.A cut-off SWE value of ≤4.7 kPa yielded 96% sensitivity and 73% specificity to rule out high-grade fibrosis.The stage of fibrosis and SWE values with respect to time after liver transplantation did not show any apparent relationship (Figure 2).Median time for biopsy after liver transplantation was 5 years and IQR 7.8 years.

| Shear wave dispersion and grade of inflammation
Only one patient had histologically high-grade inflammation (grade 2) whereas the remaining 45 had low-grade inflammation (grade 0-1).All patients examined with SWD showed histologically low-grade inflammation, with a median SWD (min; max) of 13.6 m/s kHz (10.7; 17.6) (Figure 3).No correlation was found between SWD and the grade of fibrosis (r = .25;p = .27).The relationship between SWE and SWD values is displayed in Figure 4.
Eleven patients had high SWD values (>13,6 m/s kHz), that is, values above or equal to the median value, despite low-grade inflammation.Medical records were examined to elucidate the presence of apparent confounding factors of liver viscosity.These results are displayed in Table 3.

| Impact of liver biopsy results on medical therapy
In 41 patients (89%), the liver biopsy results did not lead to any change in treatment per se.Three patients had biopsies indicative of a decreased stage of fibrosis compared to their previous biopsy, which resulted in a lowering of the dose of immunosuppression therapy.
Two patients had findings suspicious of low-grade rejection and had their immunosuppression therapy changed accordingly.One of these two patients had SWE measured 5.2 kPa (F1 and RAI score of 4/9), and SWD was not measured.The second patient had an SWE value of 9.8 kPa and an SWD value of 12.2 m/s kHz (F0 and RAI score 3/9).

| Liver function tests
Liver function tests are presented in Table 1.AST showed a moderate positive correlation to SWD, and platelets showed a poor negative correlation to SWE (Table 4).No further correlations were found between liver function tests and SWE or SWD.

| DISCUSS ION
In this explorative study, we compared SWE and SWD to histology findings in elective pediatric liver transplant recipients without clinical complications.Both SWE and SWD values were slightly higher TA B L E 2 Indications for liver transplant surgery.

Biliary atresia 12
Unknown liver failure 6 (10)   Hepatoblastoma 7 Progressive familial intrahepatic cholestasis whether SWD can be used as a measure of inflammation in grafted livers (Figure 3). 29r finding of increased SWE values in pediatric liver recipients without clinical complications is consistent with results in adult recipients recently published by Boeken et al., however, their study was limited by the lack of comparison to histology. 22Our study suggests that increased liver stiffness could partly be explained by the fact that liver fibrosis existed among these patients.It also showed that grafted livers with low-grade fibrosis (F0-1) had slightly increased SWE (median 5.7 kPa) compared to nongrafted livers with biopsyproven low-grade fibrosis, reported previously by others (median 5.0 kPa). 11The high frequency of confounding factors could explain the high SWE values in the absence of fibrosis in the grafted livers and the greater variability in SWE values across the pediatric cohort.
For instance, one patient with an SWE value of 16 kPa displayed only F1 on biopsy (Figure 4), but had underlying portal hypertension, which is a known confounding factor for increased liver stiffness.
Another confounder for increased SWE values is inflammation, for which SWD has been reported as a promising marker, 26,29  In our cohort, only one patient had histologically high-grade inflammation and, therefore, any correlation analysis between the grade of inflammation and SWD could not be made (Figure 3).However, this study shows that SWD can be increased, in comparison to healthy livers, in livers with a histologically verified absence of inflammation, or those with low-grade inflammation (Table 3).
SWD reflects tissue characteristics other than fibrosis, and multiple factors can alter tissue viscosity. 29,30It is important to elucidate what conditions affect liver viscosity in terms of increased SWD values.2][33] We sought to identify the presence of any such conditions in all patients with sampled SWD within the study cohort.Eleven patients had increased SWD values, that is, ≥ 13.6 m/s kHz (Table 3).Among these, five had portal hypertension and/or portal vein stenosis, two had chronic cholestasis and one displayed ischemic preservation injuries.Among the ten patients with SWD values below 13.6 m/s kHz no apparent confounders were found in eight of the patients while one patient (SWD 13.4 m/s kHz) had a this study size is too small to elaborate on the topic.In two adult studies, SWD combined with SWE showed better performance in the early diagnosis of biopsy-proven nonalcoholic steatohepatitis than SWE alone. 13On the other hand, in an adult cohort with various biopsy-proven liver diseases, the diagnostic performance in the detection of liver fibrosis, inflammation or steatosis was not improved by adding SWD to SWE. 29,37 Due to the lack of highgrade inflammation in the current cohort, it is still unclear whether SWD can be used as a measure of inflammation in grafted livers (Figure 3).Until large-scale studies are published, our findings suggest that an SWD value of 11 m/s kHz could be used as a value for when conditions affecting liver viscosity are unlikely to exist.Even though liver biopsy is considered the gold standard for tissue characterization, it is susceptible to sampling error and can result in misdiagnosis and staging inaccuracies due to the inhomogeneous distribution of disease. 38,39The majority (41 out of 46) of liver biopsies in our cohort did not impact medical therapy, which raises the question of whether elective liver biopsy can be omitted when Key limitations of this study are its exploratory pilot study design with a small sample size, which decreased statistical power, especially regarding associations with SWD, which was sampled in only 21 individuals.This likely contributed to the weak correlation seen between SWE values and grades of fibrosis.Our data were primarily collected from patients during anesthesia, which could be considered a limitation.However, it has previously been shown to have a weak impact at the group level. 25

| CON CLUS ION
The examination of uncomplicated liver grafts in a small pediatric cohort revealed slightly increased SWE and SWD values in comparison to previously reported values in healthy children.These increases seem to be related to fibrotic and inflammatory elements that can be detected histologically but also other confounders impacting the viscoelastic properties of the liver.Future studies are encouraged to investigate histopathological conditions that influence SWD in larger cohorts.Our results suggest that scheduled liver biopsies in well-functioning liver grafts may not be warranted if clinical and laboratory findings -along with low SWE and SWD measurements -do not raise suspicion of liver graft damage.

AUTH O R CO NTR I B UTI O N S
All authors have read and agreed to the publication of this manuscript.
At this time, the SWE and SWD value ranges have not been determined in pediatric liver graft recipients.If the range of SWE and SWD values in clinically uncomplicated grafted livers could be established, and related to reference biopsy, these techniques could improve the noninvasive surveillance of liver recipients.The current study aimed to prospectively measure and compare SWE and SWD values with histopathology findings in clinically uncomplicated pediatric liver graft recipients admitted for a scheduled follow-up biopsy.
SWE and SWD values were compared to histological findings, liver function markers (aspartate transaminase [AST], alanine transaminase [ALT], bilirubin and gamma-glutamyl transferase [gamma-GT]), white cell count, prothrombin time and patient age.

1 F I G U R E 1
Scatter plot of shear wave elastography and grade of fibrosis according to Batts & Ludwig classification.A ring represents patients with values of shear wave dispersion in addition to shear wave elastography.in this cohort of transplant recipients than corresponding values previously reported for healthy children. 11,25,28SWE correlated with fibrosis in the pediatric liver recipients, but the correlation was weak and there was a large variation in SWE values within the different grades of fibrosis.The study further showed that SWD reflects different tissue characteristics than fibrosis.Due to the lack of highgrade inflammation and rejection in the current cohort, it is unclear however, studies on this novel measure of viscosity, are very scarce, and to our knowledge lacking in pediatric liver recipients.The median SWD value of 13.6 m/s kHz in the current study, in which approximately 50% of the patients had no inflammation and 50% displayed grade 1 inflammation, is higher in comparison to the two studies previously published on healthy children, reporting SWD values of 11.4 m/s kHz and 11.7 m/s kHz.25,28Sugimoto et al., reported that in a cohort of adult liver recipients, SWD could aid in the detection of liver allograft damage since the degree of necroinflammatory activity in the liver was the only significant determinant factor for high SWD.26

F I G U R E 2 F I G U R E 3
Distribution of biopsies grouped according to the time after liver transplant (years) in relation to the degree of fibrosis according to Batts & Ludwig classification.Scatter plot of shear wave dispersion values and grade of inflammation in children with grafted livers.
SWE and SWD values, and other clinical and laboratory measures, remain stable.Our study suggests that it may be feasible to establish and thereafter monitor individual SWE and SWD values, curbing the routine use of biopsies at predetermined follow-ups for all patients.Cut-off values for SWE and SWD could provide a noninvasive means to identify any changes in liver stiffness and viscosity that warrant a biopsy.

kHz) SWE (kPa) Clinical findings in biopsy and medical charts
Scatter plot showing values of shear wave elastography and shear wave dispersion.
TA B L E 4