Liver imaging the revolution

Transcription

1 Liver imaging the revolution Valérie Vilgrain Hôpital Beaujon, Paris, France PHC

2 At the Beginning of the story

3 Radiology in the 1970s US Garrett Radiology 1976 abscess Taylor Radiology 1976 CT liver specimen liver mets Philips Radiology 1975 hepatic cyst Bryan Radiology 1977

4 Radiology in the 1980s MRI Radiology 1984 Liver mets

5 Chronic Chronic liver liver disease disease Tumor Tumor detection detection Tumor Tumor characterization characterization Tumor Tumor response response Morphological imaging

6 Chronic liver disease Cirrhosis Non cirrhotic: Portal cavernoma Budd-Chiari Biliary

7 Chronic liver disease Cirrhosis Non cirrhotic: Portal cavernoma Budd-Chiari Biliary

8 Chronic liver disease Cirrhosis Non cirrhotic: Portal cavernoma Budd-Chiari Biliary

9 Chronic liver disease Cirrhosis Non cirrhotic: Portal cavernoma Budd-Chiari Biliary

10 Tumor screening What are the clinical needs? Impact on patient management Impact on patient survival 17mm HCC Which patients High risk patients: chronic liver disease (HCC) Which imaging technique? Widely performed US Non invasive Wu, Gut 2015

11 Liver tumor staging What are the clinical needs? Impact on patient management Impact on patient survival Which patients 2010 Oncologic patients Surgical resection (colorectal liver metastases) Which imaging technique?

12 Liver tumor staging liver metastases Conventional Conventional DW-MR DW-MR HB HB contrast contrast agents agents T1, Restricted T1, T2 T2 Restricted Lesion Lesion enhancement enhancement diffusion diffusion Hypointensity Hypointensity on on HB HB phase phase

13 Liver tumor staging liver metastases Meta-analysis on liver mets 39 articles (1989 patients, 3854 liver metastases) MR Sensitivity (%) diffusion HB phase using Gadoxetic acid both 87.1% 90.6% 95.5% Higher Highersensitivity sensitivitywith withcombined combineddw DWand andhb HBphase phase Only Onlycolorectal colorectalliver livermets mets Liver Livermets mets<<11cm cm Neoadjuvant Neoadjuvantchemotherapy chemotherapy Histopathology Histopathologyalone aloneas asreference referencemethod method Translation in clinical practice? Vilgrain, Europ Radiol 2016

14 Liver tumor staging liver metastases VALUE study: 360 patients with suspected colorectal cancer liver metastases who had either gadoxetic acid enhanced MRI MRI with extracellular contrast medium or contrast-enhanced CT as a first-line imaging method in patients Gadoxetic A MR Regular MR CT Further imaging 0% 17% 39.3% Diagnostic confidence 98.3% 85.7% 65.2% Surgical changes 28% 32% 47% Zech, Br J Surgery 2014

15 Tumor characterization What are the clinical needs? Impact on patient management in all patients Incidental Incidental lesions lesions Mostly Mostly benign benign f/u? f/u? Rarely Rarely malignant malignant Hemangioma FNH Adenoma Oncologic Oncologic pts pts Mostly Mostly liver livermets mets some some benign benign 50% 50% <2cm <2cm Chronic Chronic liver liver diseases diseases HCC HCC Benign Benign nodules nodules

16 Tumor characterization HCC HCC Lesion enhancement FNH FNH CT CT MRI MRI CEUS CEUS hemangioma hemangioma

17 Tumor characterization Fat Hemorrhage Iron Highly cellular Hepatocellular

18 Tumor characterization Fat Hemorrhage Iron Highly cellular Hepatocellular

19 Tumor characterization Fat Hemorrhage Iron Highly cellular Hepatocellular

20 Tumor characterization Fat Hemorrhage Iron Highly cellular Hepatocellular

21 Tumor characterization Fat Hemorrhage Iron Highly cellular Hepatocellular MRI MRI

22 Tumor response RECIST criteria Complete response Partial response Progression

23 Can we push the morphological imaging forward? Probably marginally The revolution is elsewhere Quantitative Quantitativeimaging imaging Multiparametric Multiparametricimaging imaging Multimodal Multimodalapproach approach 3D 3Dassessment assessment

24

25 quantitative imaging

26 Quantitative imaging the big four Liver volume Fat quantification Murray Djokovic Iron quantification Nadal Federer Widely Widelyused used Highly Highlyreproducible reproducible Liver stiffness Validated Validated Clinical Clinicalimpact impact

27 Quantitative imaging liver volume Entire liver Prognostic factor in fulminant hepatic failure Future remnant liver critical predictor of perioperative outcomes after major liver resection Indicates portal vein embolization 1 mo Post-PVE pre-pve CC CC Tumor volume After Rt. hepatectomy CC CC CC CC Yamagishi, J Gastroenterol Hepatol 2005

28 Quantitative imaging fat NAFLD is the most frequent cause for referral for chronic liver disease and an increasing cause of HCC In patients undergoing liver resection, NAFLD increases the risk of postop complications and death In liver transplantation, presence of steatosis in the graft increases the risk of graft failure Berzigotti, J Hepatol 2014

29 Quantitative imaging fat with MRI MR spectroscopy direct measure of the chemical composition of tissue chemical-shift techniques (T1 GE) in-phase/out-of-phase (Dixon in 1984) IP OP Current Dixon-type sequences: Decomposition of water and fat with echo asymetry OP IP water fat

30 Fat-water Fat-waterseparation separationsequences sequences for forquantification quantification Steatosis Mild iron overload PDFF 29% T2*: 12 ms 25μmol/g No steatosis Mild iron overload PDFF 7% T2*: 5 ms 52μmol/g

31 Quantitative imaging liver stiffness Increased Increasedstiffness stiffnesscorrelates correlateswith withthe theamount amountof offibrosis fibrosis Treatment planning in chronic viral hepatitis depends on the degree of fibrosis In patients undergoing liver resection, marked fibrosis and cirrhosis increase the risk of postop complications and death Liver and spleen stiffness are correlated to the

32 Quantitative imaging liver stiffness Ultrasound Transient Elastography (TE) Integrated onto conventional US Acoustic Radiation Force Impulse (ARFI) ShearWave Elastography (SWE)

33 Fibroscan Ultrasound MRI availability dimension 1D 2D 3D performance Other data steatosis All US All MRI limitations Ascites, obesity few few

34 Quantitative imaging May Maybe beused usedcurrently currently poorly poorlyreproducible reproducible other Diffusion-weighted MRI b = 10 ADC, IVIM technique Insufficiently Insufficientlyvalidated validated D f D* Perfusion parameters Blood flow, blood volume Fingerprinting Simultaneous measure of T1 and T2 Liver function: acid Verloh, Europgadoxetic Radiol 2014 Saito, JMRI 2012 MELD 10 MELD >10

35 Conclusion Liver imaging is everyday morphological and multimodal The role of imaging biomarkers is increasing but few are validated Tomorrow imaging will be quantitative and multiscale: Radiomics Artificial Intelligence in Imaging is the next revolution and will change the practice