Leader(s)

• 

  Garteiser Philippe

  • 01 57 27 74 55
  • philippe.garteiser@inserm.fr
• 

  Van Beers Bernard

  • 01 57 27 74 35
  • bernard.van-beers@inserm.fr

Presentation

Laboratory of imaging biomarkers – LBI

We perform translational research in medical imaging to develop and validate novel imaging biomarkers in inflammation, fibrosis and cancer. Our efforts range from fundamental research in imaging methods to applied preclinical and clinical research in hepatic and abdominal diseases.

We improve rapid, functional and quantitative methods at magnetic resonance imaging (MRI) (diffusion MRI, perfusion MRI, MR elastography, susceptibility and relaxometry) and ultrasound imaging. We also assess novel tracers for cellular imaging. These developments are first validated on thin tissue sections and in small animals. Once the proofs of concept are obtained, the methods are transferred to clinical imaging.

Our LBI laboratory participates in the imaging research network “France Life Imaging”, with Philippe Garteiser acting as coordinator of the Paris Centre hub. Our laboratory is also active within the Université de Paris “Imageries du Vivant” network. In addition, the LBI laboratory is one of the two imaging research teams in charge of the preclinical imaging platform “FRIM” (Inserm UMS34).

Main research projects

Imaging of hepatic inflammation in metabolic diseases

Non-invasive assessment of inflammation and hepatocyte injury is important to diagnose non-alcoholic steatohepatitis or NASH. The RHU “QUID-NASH” project that we pursue with other academic and industrial partners aims to assess new non-invasive imaging methods of NASH in murine models of steatohepatitis and in a cohort of 600 diabetic patients with biopsy-proven hepatic steatosis. We develop a virtual hepatic biopsy approach, integrating multi-omic data and quantitative parameters obtained with ultrafast ultrasound, multiparametric MRI and MR elastography. Recently, we have launched a study on the frequency dispersion of the diffusion coefficient in hepatic tissue (ANR funded “Stedi-NASH”). The proposed method enables to probe hepatic tissue at specific spatial scales. This novel method has the potential to inform hepatocyte size and cytoplasmic content.

Steatosis (left) and hepatic stiffness (right) maps in a patient graded S1A1F1 at histopathology (top row) and a patient graded S2A3F3 (bottom row).

Imaging of the forces in hepatic cancer 

To provide advanced characterization of tumor mechanical properties, the apparent elasticity at different levels of static load can be measured. We use this compression MR elastography method to assess tumor interstitial pressure and solid stress, two parameters playing a role in the efficiency of targeted treatments (H2020 European project “FORCE”).

Elasticity maps of patient-derived hepatocarcinoma xenografted in nude mouse, measured at different levels of static load, showing increase in apparent elasticity under load (Pagé G et al. J Magn Reson Imaging 2019, Pagé G et al. Cancers 2021)

Functional contrast-enhanced MRI of the liver

Pharmacokinetics of hepatospecific agents

Hepatocyte influx, biliary efflux and sinusoidal backflux can be assessed with pharmacokinetic modelling after dynamic liver MRI with hepatospecific contrast agents (Gd-EOB-DTPA, Primovist). Similarly to most drugs, those hepatospecific agents cross the hepatocyte membranes through specific transporters, including OATPs, MRP2 and MRP3, creating concentration gradients between the extracellular compartment, the hepatocytes and the biliary compartment.

The hepatocyte uptake of hepatospecific contrast agents (Gd-EOB-DTPA) is performed through OATP membrane transporters. MRP2 is responsible for biliary efflux, and MRP3 for sinusoidal back-flux. The amount and function of these transporters are modified in chronic liver diseases (Van Beers BE et al. J Hepatol 2012)

Liver perfusion and transport through the hepatocytes are important determinants of hepatic function. Using dynamic hepatobiliary contrast-enhanced MRI and developing pharmacokinetic modelling, we study the changes of the transport parameters in murine models of chronic liver disease as well as patients. Our aim is to use MRI pharmacokinetic parameters as biomarkers of liver function, since they reflect the alteration of quantity and functionality of hepatic transporters (OATPs, MRP2 and MRP3).

Parametric maps obtained with hepatobiliary contrast-enhanced MRI, showing modifications of hepatic transport parameters in a patient diagnosed with hepatic cirrhosis (bottom) compared to a patient without hepatic fibrosis (top) (Ronot M, Joly F, Van Beers BE in Seiberlich N et al. Quantitative magnetic resonance imaging, Elsevier 2020)

Macrophage-based contrast agents in fibrogenesis

The dynamics of fibrogenesis and fibrolysis has been seldom studied with biomedical imaging. Using dedicated MRI acquisition sequences and intravenous injections of ultrasmall iron oxide particles (USPIO), we observe the different macrophage populations involved in this dynamic process, according to their endocytic capacity.

Assessment of R2 (a), R2* (b), R2’ (c) relaxation rates and magnetic susceptibility (d) show that pro-inflammatory M1 macrophages exhibit higher MRI parameters than M0 and M2 macrophages, because of higher USPIO concentrations within the former. This increased uptake in M1 macrophages is confirmed at immunofluorescence (right panels) (Khaled W et al. J Magn Reson Imaging 2019)

MRI of liver cirrhosis and portal hypertension

Classically, the assessment of cirrhosis and portal hypertension severity relies on hepatic venous pressure gradient measurements. We have shown that MR elastography provides a non-invasive method for assessing portal hypertension severity (Ronot M et al. Eur Radiol 2014). We aim to improve this approach by integrating MR elastography data with radiomic information and biological data extracted from microcirculating vesicles (IFB project “Integrative bioinformatics”).

Parametric maps showing increase in hepatic (top) and splenic (bottom) shear modulus according to portal hypertension severity (A and D: hepatic venous pressure gradient below 10 mmHg; B and E: pressure gradient between 10 and 11 mmHg; C and F: pressure gradient above 11 mmHg) (Van Beers BE et al. Semin Liv Dis 2017)

Pancreatic quantitative MRI in obesity

We observed that pancreatic tissue in obesity could be characterized before and after bariatric surgery with quantitative MRI, including the measurements of the mechanical properties, fat fraction and R2* relaxation rate. We are currently developing a similar approach in the ANR project “PAIR-pancréas” to detect the pre-tumoral pancreatic lesions associated with the metabolic syndrome.

Elasticity (left), fat fraction (middle) and R2* relaxation rate (right) in the pancreas of control rats, obese rats and obese rats after bariatric surgery. Obesity induces an increase in MRI parameters, which is partially reversed after bariatric surgery. Histological analysis of pancreatic tissue confirms these observations (Rebours V, Garteiser P et al. Sci Rep 2018)

Quantitative MRI of intestinal inflammation and fibrosis

In a rat model of radiation-induced colitis, we observe that combining MRI parameters of signal intensity, perfusion, diffusion and magnetization transfer is useful for assessing inflammation and fibrosis, with potential therapeutic implications (Zappa M et al. NMR Biomed 2018).

Advanced methods for MRI data analysis

Our team also works on the extraction of quantitative data from MR images and parametric maps. We develop radiomic-based as well as unsupervised methods to extract those quantitative data. The integration of the ever-increasing volume of quantitative data generates new challenges for statistical analysis. Therefore, we develop innovative regularized multi bloc approaches for dimension reduction of highly multidimensional data.

If you want to join our team as post-doc, PhD or master student, please contact us: [mailto:bernard.van-beers@inserm.fr]