VALidation of Imaging-based Liver Biomarkers in PEDiatric Patients

The purpose is to improve diagnostics and surveillance in children with liver disease and reduce the need for biopsies. The overall aim is to test the hypothesis that multiple novel imaging-based biomarkers can be used to diagnose and phenotype liver disease non-invasively. Specifically, we will evaluate if the imaging-based biomarkers shear wave elastography (SWE), shear wave dispersion (SWD) and attenuation coefficient image (ATI) can differentiate between various pathoanatomic phenotypes (fibrosis/inflammation/ steatosis) associated with liver disease and to determine if these imaging-based biomarkers can be used for detailed, non-invasive liver characterization in children.

Specific aims

1. To determine if SWE, SWD and ATI can be used as quantitative imaging-based biomarkers for fibrosis, inflammation and steatosis of the liver in children, and to establish cut-off values for grades of disease.
2. To identify interactions between imaging-based biomarkers and serology, age, sex, free-breathing, weight, BMI, cholestasis, necrosis, and congestion in children.
3. To evaluate the imaging-based biomarkers longitudinally to monitor treatment effects and determine if these biomarkers can improve surveillance of disease, compared to conventional ultrasound/clinical biomarkers, in children with liver transplants, cystic fibrosis, obesity and intestinal failure-associated liver disease (IFALD).
4. To compare the relationship between regional tissue characterization with ultrasound and global characterization with magnetic resonance imaging (MRI)

The project includes 3 studies, each planned to generate several manuscripts, planned to answer the 4 specific aims. The studies will be conducted according to the Declaration of Helsinki. Ethical approval has been obtained (Dnr;634-18, 2021-06760-02, 2022-01046-02, 2022-01661-02).

Study 1: Validation of multiple ultrasound-based liver biomarkers (SWE/SWD/ATI) Study 2: Surveillance of disease and treatment - longitudinal monitoring with SWE/SWD/ATI Study 3: MR elastography (MRE)

Study 1: Validation of multiple ultrasound-based liver biomarkers (SWE/SWD/ATI) The imaging-based biomarkers SWE, SWD, and ATI are being compared with histology (specific aims 1-2). During a 8-year period ( estimated 2019-2027), all patients scheduled for a clinically liver biopsy at the Queen Silvia Children's Hospital, are consecutively invited to participate in the study. With an estimated 60 liver biopsies/year, we aim to include 400-500 patients and 150 healthy children. Inclusion criteria are any form of suspected or confirmed liver disease, age between 0-18 years and consent by the child/parents to participate. The exclusion criteria are technical limitations regarding the measurements, according to the ultrasound consensus statement defined as interquartile range (IQR)/median kPa>30%. During anaesthesia (fasting>4h), as well as in awake state, and free-breathing, ultrasound measurements (SWE/SWD/ATI) are performed in the right liver lobe with an intercostal approach, applying minimal transducer (i8CX1) pressure.

Measurements are performed according consensus statement for ultrasound elastography, with sampling of the biomarkers: SWE (reflecting fibrosis), SWD (believed reflecting inflammation/oedema) and ATI (reflecting steatosis). Measurements are obtained 1.5-3 cm below the liver capsule as the median of 10 registrations with IQR/median<30%kPa. Following these measurements, 1-2 percutaneous 18 Gauge biopsies are obtained from the corresponding area. The measurements are performed by one of five physicians specialized within paediatric radiology. All are blinded to the histopathological results. The biomarkers (SWE/SWD/ATI) are correlated to histology, serological hepatic markers, age, sex and BMI. Two board-certified paediatric pathologist, who is blinded to elastography measures, scores all biopsies. Both fibrosis and inflammation severity are scored according to the Batts\&Ludwig classification (0-4). NAFLD score (0-3) is used to evaluate steatosis.

Approximately 150 children without liver disease, who are clinically indicated for a biopsy of the kidney, will be recruited as controls. The protocol for these controls is the same as for the patients (anaesthesia, fasting, etc.) but without biopsy of the liver.

Some diseases/conditions require monitoring with repeated liver biopsies as part of the treatment protocol. The study period allows for a natural longitudinal follow-up of patients who require repeated liver biopsies. Separate in-depth analysis will be performed for this sub-cohort.

Study 2: Surveillance of disease and treatment by longitudinal monitoring with SWE/SWD/ATI Several diseases are monitored with conventional ultrasound on a regular basis in children. Examples include cystic fibrosis, liver transplants, severe obesity and failure-associated liver disease (IFALD). We will determine the value of adding SWE/SWD/ATI to the conventional ultrasound examination longitudinally, to evaluate treatment effects and elucidate how biomarkers can improve surveillance of disease (specific aim 3). This study includes several sub-groups of children, which are described below.

Children with cystic fibrosis (CF) (study 2a) All children with CF are monitored regularly regarding liver function, both with serological liver tests as well as with conventional ultrasound. About 40% of these children have elevated transaminases and about 60% have steatosis. Moreover, an estimated 5% have significant liver disease that requires treatment in addition to standard care. In the long term, transplantation is necessary in those with the most severe liver disease. Therefore, it is desirable to identify individuals at high risk of liver disease and need for treatment as early as possible before damage to the liver becomes irreversible. Since SWE/SWD/ATI are promising non-invasive biomarkers, there could be benefits using them to screen for liver disease in children with CF. We will also evaluate the treatment effect of new CF transmembrane modulators (e.g. Kaftrio) using SWE/SWD/ATI. Prospectively, all children with CF treated at Queen Silvia Children's Hospital (estimated 70 children) will consecutively be included and examined with SWE/SWD/ATI at diagnosis (i.e. before treatment with Kaftrio) and at each yearly check-ups for 4-5 years. The examination procedure follows the same principles as in study 1 but does not include biopsies. SWE/SWD/ATI will be compared before and after treatment and followed over 4-5 years and further correlated to clinical status and serological liver markers (example; transaminases, albumin and gamma-glutamyl transferase [GGT]) longitudinally.

Children with liver transplants (study2b) Currently >60 patients are followed at the Queen Silvia Children's hospital for post-liver transplant surveillance with ultrasound examinations on a regular basis (twice-daily for the first week, once-weekly for 4-5 weeks as long as hospitalized, and thereafter 1, 3, 5 and 10 years postoperatively). Patients receive additional ultrasounds when clinically indicated, for example, if transplant rejection is suspected. All children with liver transplants will be consecutively included, and SWE/SWD/ATI will be added to the conventional examination to determine if surveillance can be improved. The biomarkers will be compared longitudinally over 4-5 years and correlated to clinical outcome and serological liver tests.

Children with intestinal failure-associated liver disease (IFALD) (study2c) A substantial proportion of infants and children, who depend upon parenteral nutrition because of intestinal failure or short bowel syndrome, develop cholestatic/fibrotic liver disease, known as IFALD. The diagnostic challenge is that liver tests can be normal despite severe liver disease. Thus, IFALD can be difficult to discover before the disease complications manifest. Early detection of IFALD is important since adjustment of parenteral nutrition can control the disease. Queen Silvia Children's Hospital is a national centre for children with intestinal failure, where around 40 children are monitored today. The hypothesis for this sub-study is that imaging-based biomarkers can identify liver damage early and thus enable a more proactive treatment. To test this hypothesis, and to elucidate how early liver damage manifests in children with intestinal failure, all currently-monitored children at the centre will be examined with ultrasound to measure SWE/SWD/ATI at their yearly check-up. The biomarkers will be compared longitudinally over 4-5 years and correlated to clinical estimates and liver blood tests. If clinically indicated (i.e. signs of severe liver disease), biopsy will be included.

Children with obesity (study2d) Children/adolescents with severe obesity and co-morbidities could also receive diagnostic benefit from the ultrasound-based liver biomarkers. Treatment with either bariatric surgery or novel appetite-regulating drugs (GLP1 analogues) is offered, at Queen Silvia Children's Hospital, to children with BMI ≥35 kg/m2 and a severe co-morbidity, and those with treatment-resistant (conventional treatment) obesity with BMI ≥40 kg/m2. Children scheduled for either bariatric surgery or GLP1 analogue treatment will be included over a 2-year time period and examined with ultrasound to obtain the biomarkers SWE/SWD/ATI pre- and post-intervention (after 1 and 2 year). The biomarkers will be compared with each other longitudinally and correlated to clinical outcome measures, weight, BMI and serological markers to determine if SWE/SWD/ATI can improve diagnostics and detect an early treatment effect. In total, about 50 children/adolescents will be included.

Study 3: MR elastography (MRE) (study 3) With conventional MRI, the gross morphology of the liver tissue can be estimated. With MRE, fibrosis and steatosis of the liver can be measured and quantified. This study evaluates the relationship between regional tissue characterization with ultrasound biomarkers SWE/SWD/ATI and global characterization with MRE in relation to biopsy in children (specific aim 4). This study cohort will include around 50 children who are participating in study 1 and 2, and in whom MRI is clinically indicated as a supplement to ultrasound examination - for example, children with obesity for whom ultrasound may be insufficient or children in whom a more global characterization of the liver is desired. The methodology is the same as in study 1 and 2, but MRE measurements are conducted to assess liver stiffness and steatosis. The ultrasound (including biopsy) and MRE are performed within the same week. Global measurements with MRE and regional ultrasound measurements are compared to determine the correlation between the ultrasound biomarkers SWE/SWD/ATI and MRE and linked to histology, serological hepatic markers and clinical outcome measures.

Timeplan (tidsplan) The feasibility has been tested in published pilot studies, performed between 2019-2021. To increase power study participants in the pilot studies will sometimes be included in the larger studies. Both study 1 and study 2 were initiated in 2021, when software for SWD/ATI was released on the market. These two studies will run in parallel. Inclusion of patients for study 1 with biopsy will be ongoing until approximately beginning of 2027 until we have included approximately 400-450 children in total of whom 300-350 have SWD/ATI sampled. Thereafter patients will be monitored longitudinally until we have 4-5 year data.

Data analysis and statistics will be performed using the anonymized data. Statistical methods have been chosen in collaboration with a statistician. Descriptive statistics will be presented. For pairwise comparison between groups of continuous variables, we plan to use Fisher´s non-parametric permutation test. For pairwise comparison between two groups of dichotomous variables, we plan to use Fisher's Exact test. To assess the relationship between SWE/SWD/ATI measures and continuous variables, Spearman rank correlation (rho) is planned to be used. Logistic regression analysis will be performed for each independent variable to predict the outcome. We will calculate area under ROC-curve (AUC-statistics) for the description of goodness of predictors. Intraclass correlation coefficient (ICC; Shrout-Fleiss reliability: single score) and Bland-Altman plot will be used to evaluate observer variation. Statistical methods can be altered depending on suggestion with statistical expertise. Power calculations have been performed with a simulation based approach, based on small pilot-studies.

Shear Wave Elastography (SWE):

A sample size of 400 individuals is required to distinguish between fibrosis grades 0-1 and grades 2-4 with 80% statistical power, assuming a target sensitivity of 95% and an optimism adjustment of 1.5% (i.e., expected real-world sensitivity approximately 93.5%).

Reduced SWE Sample Scenario:

With 300 individuals, the power remains at 80%, but the expected sensitivity decreases to approximately 85%, accounting for the same 1.5% optimism.

Shear Wave Dispersion (SWD):

A sample size of 300 individuals yields 78% power to differentiate between inflammation grade 0 and grades 1-3, assuming a sensitivity of 95% and an optimism adjustment of 2%.

Based on these calculations, a cohort of 300 SWD and 400 SWE participants is considered statistically robust and sufficient for the study's diagnostic validation objectives.

Regarding study 3, the lack of similar studies limits the estimation of sample size needed. Once more individuals are included, a power calculation will be performed using preliminary data. We will then adjust the number of individuals included in study 3.