You are on page 1of 82

Elastography and noninvasive testing for

liver fibrosis
A Historical Review and a SOTA Presentation

Arizona GI Society Meetin


March 2015

Robert G Gish MD
Senior Medical Director
St Joseph Hospital and Medical
Center

Relevant Disclosures
None

Copy of these slides?:


rgish@robertgish.com

Who to treat:

a Focus on HCV

Fibrosis score
CRYOs
High HCC biomarkers
(baseline in all patients, then every 6 months in patients with F3 and F4)
AFPL3%, DCP, AFP (FDA approved as risk markers for HCC)
Levels : Correlate with advanced fibrosis

Liver Fibrosis

Start with a Physical Exam

Clinical Exam and Abdominal Ultrasound

Revista da Sociedade Brasileira de Medicina Tropical


Print version ISSN 0037-8682
Rev. Soc. Bras. Med. Trop. vol.43 no.2 Uberaba Mar./Apr. 2010

Choices
To Biopsy or not to Biopsy
that is the question!

Complications of Liver Biopsy


Complication

Percent (%)

Arterial hypotension

various

Pain at the right hypochondrium, shoulder

0.056-83%

Hemorrhagic complications

Subcapsular hematoma: 0.05%


Intrahepatic hematoma: 0.05%
Intraperitoneal bleeding: 0.03%
Hemobilia: 0.05%

Bacteremia

0.08%

Death

0.001-0.0001%

Bile peritonitis

0.03-0.22%

Pneumothorax, hemothorax

0.08-0.28%

Subcutaneous emphysema

0.014%

Break of the biopsy needle

0.02%

Biopsy of other organs

Lungs-0.001%
Bile-0.003%
Colon-0.003%
Kidneys-0.09%

Comparison of Liver Biopsy and Serum


Markers of Fibrosis
Factor

Liver biopsy

Serum markers

Elastography

Cost

2200$

Laboratory cost

Machine investment $130,000,


staff time

Risks

Significant

Minimal, phlebotomy

None

Contraindications

Multiple: bleeding diathesis,


morbid obesity, ascites,
extrahepatic biliary obstruction

Conditions with high rate of


false positivity

Patient needs be able to lay


still, ascites, volume overload

Accuracy

80%

60-80%

60-95%

System requirements

Operator, pathology laboratory,


pathologist

Clinical laboratory,
phlebotomy, materials

Machine, staff, time

Specimen adequacy

16 g needle
Length of liver fragment at
least 15mm with > 12 portal
tracts

Blood sample

Staff time

False positives

Interobserver variability

Sepsis, nonhepatic
inflammation, hemolysis,
trombocitopenia

Obesity, narrow ribs, ascites,


heart failure, volume overload

False negatives

Interobserver variability

Varies per test

various

Time for results

24-72 hours minimum

1-2 hours minimum lab,


1-2 weeks for results

15-45 min

Modified from: Crockett et all 2006

Liver Biopsy Stage of Fibrosis: 3 Samples


F4- bridging present
w/suggestion of nodule formation

F0 no fibrosis seen, no portal tracts


F3? Fibrosis present,
suggestion of a bridge

Accessed 9/21/14: http://www.meddean.luc.edu/lumen/MedEd/orfpath/cirhosis.htm

Liver Biopsy Stage of Fibrosis: One Patient

Accessed 9/21/14: http://www.meddean.luc.edu/lumen/MedEd/orfpath/cirhosis.htm

% Of Correctly Classified Biopsies With The Converted METAVIR Score Of


Fibrosis According To Length Of Biopsy Specimen All Stages Together

Message: obtain at least 2 cm of tissue


and use a 16 gauge needle
Bedossa P, et al. Hepatology 2003;38:14491457.

Quality of Liver Biopsy


300

Pitie experience 1773 biopsies:

225

16% 25mm

150
75
0

10

15

20

25

30

Biopsy Length (mm)

Poynard T, et al. Clin Chem. 2004;50:1344-1355.

35

40

45

50

Trends in Use of Liver Biopsy


HCV
HCV
Viral
Viral Serology
Serology

Interferon
Interferon
Liver
Liver Transplantation
Transplantation

Popularized

Imaging
Imaging
Ultrasound
Ultrasound
CT
CT

All
All oral
oral therapies
therapies
HCV
HCV

ERCP
ERCP
1940s-1950s

1960s

1970s

1980s

1990s

2000s

>2010

2015

Anesthesia for a patient undergoing a liver biopsy

Courtesy of Nid Afdhal

Liver Biopsy: Today

Iron work up
Autoimmune hepatitis
Primary biliary cholangitis
PSC
NASH
Cryptogenic
Infections: CMV, EBV, other
Liver masses not defined by CT/MR

Future: biopsy all hepatocellular carcinomas if patient has option for


systemic/targeted therapy

Just Draw Blood


Is this sufficient?

Fibrosure in the United States

FibroSure: A Continuous Variable (n=1,270)

Poynard, Clin Chem 2004; 50: 1344-55.

Diagnostic performance of selected biomarkers

Study
FibroTest
Fibrometer
ELF

Components
Bili GGT Hapto, A2m, ApoA1
PT, AST, A2M, HA, Urea
HA, PIIIP, TIMP1

FibroSpect

HA, A2M, TIMP1

0.83-0.90

APRI
FIB4
Hepascore

AST, Platelets
AST, ALT, Platelets
Bili, GGT, HA, A2M

0.69-0.88
0.74-0.85
0.74-0.86

Forns

CT,GGT, Platelets

0.75-0.91

Marcellin P, et al. Lancet 2012;6736: 61425-1.

AUROC F>2
0.74-0.89
0.78-0.89
0.77-0.87

AUROC F4
0.82-0.92
0.94
0.87-0.90

0.61-0.94
0.8-0.93
0.8-0.94

22

50% of Biopsies
May Be Avoidable with FibroSure

Imbert-Bismut. Lancet 2001; 357:1069-75.

Accuracy of APRI, FIB-4 and AST/ALT ratio

Holmberg CID 2013

Factors Associated With Significant Fibrosis

Age, yrs
BMI, kg/m
AST/ALT ratio
APRI
Liver stiffness, kPa

F0F1 Fibrosis

F2F3F4 Fibrosis

P Value

50
29
0.68 0.4
0.72 1.17
6.92

57
30
0.92 0.5
0.84 0.26
18.52

.002
.07
.005
.5
.0001

AST/ALT ratio of over 0.8 is over 90% predictive of >= F3

Just Use an Ultrasound


Portal Vein Diameter /Size
Spleen Size
Liver configuration
Small right lobe
Large caudate/left lobe
Heterogeneous

Varices

Clinical Exam and Abdominal Ultrasound

Revista da Sociedade Brasileira de Medicina Tropical


Print version ISSN 0037-8682
Rev. Soc. Bras. Med. Trop. vol.43 no.2 Uberaba Mar./Apr. 2010

How to do a Portal Vein Measurement


Broadest point just distal to the SPV/SMV confluence

Weinreb J et al. AJR 1982;139:497-499

Sensitivity, Specificity
PV size and cirrhosis

324 patients w/chronic viral hepatitis (HCV/HBV)


Liver biopsy and ultrasound
Tested ability of diagnosing cirrhosis
Spleen >12.1cm:
Sensitivity of 60%, specificity 75.3%

PV diameter>12mm
Sensitivity 76.7%, specificity 44.6%
Shen L et al. World J Gastroenterol. 2006 Feb 28;12(8):1292-5.

Portal Vein Size and Esophageal Varices


PV size 1.2cm on US could predict both
presence and risk of variceal bleeding
PV size on US is independently associated with
bleeding esophageal varices
PV size >14mm: at a great risk of bleeding from
esophageal varices
Devrajani BR et al. World J Med Sci 2009;4(1):50-53
Prihatini JLA et al. Acta Med Indones 2005;37(3):126-131
Plestina SR et al. Wien Klin Wochenschr 2005;20:711-717

OH: Just Look at the Platelet Count

Splenomegaly and Thrombocytopenia

Adinolfi, LE. Br J Haematol, 2001

Platelet Counts May Serve as a Marker of Progressive


Liver Disease Based on the HALT-C Trial Database
Cumulative Incidence of
Hepatic Decompensation

0.6

Baseline Platelets 150,000


Baseline Platelets >150,000

0.5
0.4

P<.0001

0.3
0.2
0.1
0

Years after Randomization


Analysis of baseline values from HALT-C trial database.
A model that included baseline platelet count and albumin as
well as severe worsening of AST/ALT ratio and albumin
was the best predictor of liver-related outcomes.
AST=aspartate aminotransferase.
Ghany MG, et al. Hepatology. 2011;54:1527-1537.

Predictors of Esophageal Varices in PBC

Courtesy of Keith Lindor


Levy C, Zein CO, Gomez J, et al. Prevelance and Predictors of Esophageal Varices in Patients with Primary Billary
Cirrhosis. Clin Gastro Hepatol 2007; 5(7):803-808.

Place a Probe
PaP testing!

Ultrasound-based Elastography

FibroScan

Transient elastography utilizing


single-frequency ultrasound
waves
No real-time US imaging or
doppler
130,000$
Has a direct
CPT code
Limited by obesity, ascites,
narrow ribs, NASH, volume
overload, flare of liver disease
(high ALT) and other technical
factors
Tissue volume imaged is the size
of a Bic Pen
Modifed by Rgish as
Accessed 9/21/14: http://www.fibroscan502touch.com/sites/default/files/fibroscanspecsheet.pdf

Shear elastic wave propagation

An ultrasound transducer is placed


between the vibrator and the tissues

Ultrasounds are used to measure the velocity


of the shear elastic wave
Vibrator position
Ultrasound
transducer

Time

Ultrasound acquisitions
Time

FibroScan
The probe induces an
elastic wave through
the liver

The velocity of the wave is evaluated


in a region located from 2.5 to 6.5
cm below the skin surface

2.5
cm

LB: 1/50,000 of the liver


FibroScan: 1/500 of the liver

Explored
volume
1 cm

4 cm

Diagnosis of Significant Fibrosis With FibroScan


Cutoff of FibroScan for the diagnosis of
significant fibrosis

8.75 kPa
Sensitivity: 81%; specificity: 78% Need
correction formula when using TE as
screening test for NASH
with > F2

Mean platelet counts (2A) and AST-to-platelet ratio index values (APRI: 2B) before HCV therapy (Pre-Tx) and at the time of
last follow-up evaluation in 100 patients with a sustained HCV virological response stratified by degree of hepatic fibrosis on
pre-treatment liver biopsy. Patients were categorized into three groups based upon pretreatment Ishak fibrosis scores of 0 to 2
(no fibrosis to portal fibrosis only), 3 to 4 (bridging hepatic fibrosis) and 5 to 6 (early and complete cirrhosis).

Koh, Aliment Pharm The


2013

Reproducibility of Transient Elastography

Fraquelli et al, Gut 2007

453 Subclinical Cirrhosis Diagnosed by Transient Elastography


Demonstrates Increased Risk of Severe Clinical Outcomes and HCC

Methods: Patients with chronic liver


disease (CLD) and a valid Fibroscan were divided into: 1)
SC (Fibroscan 13kPa and no thrombocytopenia, nor signs
of advanced liver disease on ultrasound or endoscopy); 2)
non-cirrhotic CLD (Fibroscan <13kPa).

Chen et al Montreal Canada

Table 4
Association of Initial Ishak Fibrosis Score from Initial, Pretreatment Liver Biopsy and
Transient Elastography Stiffness Score at the Time of Final Follow-up Evaluation
Analysis by McNemar's Test

Fibroscan on Follow-up

Baseline
Ishak
Score

Number
of
Patients

7.0

7.113.8

> 13.8

PValue

02

45

28 (62%)

16 (36%)

1 (2%)

0.0006

34

17

12 (71%)

4 (23%)

1 (6%)

56

1 (14%)

2 (29%)

Total

69

41 (59%)

22 (32%)

4 (57%)
6 Koh,
(9%)Aliment Pharm Ther
2013

Transient Elastography in CHB

Chan et al. Normal ALT

Chan et al. Elevated


ALT

F0
vs.
F1-4

Marcellin et al.

F0-2
vs.
F3-4

F0-3
vs.
F4

F0
vs.
F1-4

F0-2
vs.
F3-4

F0-3
vs.
F4

F0
vs.
F1-4

F0-2
vs.
F3-4

F0-3
vs.
F4

Cut-off

12

7.5

13.4

7.2

8.1

11

AUROC

0.88

0.90

0.96

0.76

0.87

0.94

0.81

0.93

0.93

Sensitivity

91

71

60

92

96

75

70

86

93

Specificity

75

100

95

17

59

93

83

85

87

PPV

98

100

82

94

68

78

80

66

38

NPV

38

79

87

13

94

92

73

95

99

Accuracy

90

86

86

88

77

89

76

85

87

Chan HL, et al. J Viral Hepat 2009. Marcellin P, et al. Liver Int 2009.

57

FibroScan and APRI for


Fibrosis in CHB

N=175 Dutch cohort CHB (n=93), CHC (n=82)


Original study cutoffs (F2=7.1 and F3=9.5kPa)
For CHB pts. AUROC F2-4 (FS=0.84, APRI=0.73); APRI + FS=0.85
Reduced need for liver biopsy by 48% in CHB, 38% for CHC
Proposed Algorithm
APRI
<0.5

0.5-1.5

>1.5

FibroScan
<7.5

7.5

Liver BX

F0-F1

F2
Bergmann J, et al. EASL 2008

FibroScan: Advantages
Fibroscan has the most extensive data on
pubmed in HCV and HBV
The CPT code was written specifying the
technology (VCTE) that Fibroscan uses
Source: personal communication RGish

Examples: of narrow picture


Elastograms
in liver in vivo
10

10

0
40
50

30
0
40
50

20
40
Time (ms)

60 %

-5

60

VS = F0
1.0 m/s
E = 3.0 kPa
F0

30
0
40
50

20
40
Time (ms)

60 %

-5

60

VS =F2
1.6 m/s
E = 7.7 kPa
F1

20
Depth (mm)

30

60

10

20
Depth (mm)

Depth (mm)

20

F2

20
40
Time (ms)

60 %

VS =F4
3.0 m/s
E = 27.0 kPa
F3

F4

-5

High Failure Rate Related to High BMI

604 subjects, free medical checkup, asymptomatic, no cause of


CLD, normal LTs
Overall failure rate of LS
measurement:
25% for BMI 30-35 kg/m2
41% for BMI 35-40 kg/m2
88% for BMI > 40 kg/m2

Fibrostic (largest prospective,


multicenter trial of TE with FS):
29% of FS measurements not
interpretable due to technical
pitfalls
Degos, J Hepatol 2011
Roulot, J Hepatol 2008

False positive during FLARE of HBV


FibroScan
Asian CHB n=38, ALT > x10ULN, FS at
baseline during flare and at 3-6 months
Baseline stiffness=18.5 kPa (6.9-73.5),
peak ALT 1543
3-6 months=8.4 kPa (4.4-25.1); median
ALT 31
Good correlation FS and ALT r=0.64
Fung J et al. EASL 2008

3rd Generation: Ultrasound-based Elastography


Aixplorer

Supersonic shear imaging (ShearWave)


Multiple wave fronts with frequencies ranging from 60-600 Hz
Real-time imaging available to target area of interest
Built in Doppler US
130,000$
Does the CPT code apply ?

Muller et al. UMB 2009; 35: 219-29


Bavu et al. UMB 2011;37: 1361-73

Advantages of the SW with US and Doppler

The technician can guide the device to a


larger sampling area to include regions of
the liver directly avoiding the central
vessels, gallbladder, kidneys, surroungdng
blood vessels, lung, ribs and liver defects

Source: personal communication RGish

Ultrasound-based Elastography

Cassinotto C, et al. J Hepatol. 2014;61:550-7.

Let us Contrast
Shearwave (Supersonic) and Transient Elastography (Fibroscan)

Why would you want to do sequential fibrosis exams


(biopsy vs elastography?)

Parenchymal extinction nodule PEN


Hepatology.1995 May;21(5):1238-47.
Hepatic and portal vein thrombosis in cirrhosis:
possible role in development of parenchymal extinction and porta
Wanless IR1,Wong F,Blendis LM,Greig P,Heathcote EJ,Levy G.

Just Add the Elastography Test to Your Magnetic


Resonance Exam
Benefit: you can also calculate fat %
Iron
Device advise? (Rgish)
GE best
Siemens second best
Phillips has not matured their technology
New 2nd or 3rd Gen: 1.5t scanner probably best
New 3t scanner second best

1st Gen 1.5t should not be used


Elastograhy Device cost is $60,000 + software cost

Elastography Studies
Magnetic-resonance elastography (MRE)
Continuous longitudinal vibrations at 60 Hz via the
driver
2D gradient-ECHO MRE sequence acquires images

Talwalker JA, et al. AJR. 2009;193:122-7.

Magnetic Resonance Elastography

Speakers notes: looks like something from the 1960s Jefferson Airplane concert?

Talwalker JA, et al. AJR. 2009;193:122-7.

Abstract#211
The Severity of Steatosis Influences Liver Stiffness
Measurement in Patients with Nonalcoholic Fatty Liver Disease

Petta, M Maida, FS Macaluso, V Di Marco, C Camm, D Cabibi, A Crax.


Sezione di Gastroenterologia, Di.Bi.M.I.S.,
University of Palermo, Italy
salvatore.petta@unipa.it
Slides Courtesy of S Petta

Noureddin

Two Patients with Chronic Liver Disease:


Is Hepatic Fibrosis Present?

2.2 kPa

M, 59 yrs
4

5.9 kPa

Shear Stiffness (kPa)

Noureddin

M, 64 yrs
4

Shear Stiffness (kPa)

Importance and Implications


MRE has performed better in detecting
fibrosis than fibroscan.
Further studies are needed.

Noureddin

chronicliver diseaseinUnitedStates1
Spectrumof diseases (steatosis,steatohepatitis,brosis,HCC)
Goldstandarddiagnosis liver biopsy
Risks: infection,bleeding,perforation
Cons: samplingerror,subjectiveinterpretation
Utilityof noninvasiveimaging/diagnosticsemerging
Conventional ultrasound,CT,advancedMRI
Newquantitativeultrasoundtechnique(QUS) tissueparameter
backscatter coefcient (BSC)2
Several animal models, fewlargehumanstudies

Rawultrasonicsignalsfromliver andcalibrationphantom
capturedusingcurvedtransducer
Signals analyzedofineto calculateBSC(numeric
parameter representingenergyof ultrasonicsignals
scatteredbyliver)
Sametransducer placedonreferencephantom(Figure1)
Phantomsignalsusedtoeliminateoperator- and
instrument-dependent factorsincalculations
Makes BSCindependent of machinefactors
andthuspotentiallydeployabletoany
clinical USscanner
Datafor aeldof interest,or FOI (Figure2),is recorded

Abstract # 1084 AASLD 2014


Noninvasive Diagnosis of Nonalcoholic Fatty Liver Disease and
Quantification of Liver Fat by New Quantitative Ultrasound

40
35

Assess accuracyof QUSBSCtodiagnosehepaticsteatosis


Quantifysteatosis withQUSBSCusingMRI-PDFF as reference

Figure1.Transducercapturingdataon
aparticipantandareference
phantom.
Transducer rst placedover the
participantsliver (left); datacaptured
for aeldof interest (FOI) fromwhich
BSCiscalculated.Transducer isthen
placedonphantom(right) tocorrect
for operator/machine-dependent
factors.

METHODS
IRB-approved,HIPAAcompliant,cross-sectional analysis
(participantsderivedconsecutivelyfromprospectivecohort
Inclusion: >18yearsold,informedconsent
Exclusion: pregnant/nursing,CIs toMRI,reasonsfor liver disease
alcohol,meds,infections,AI,metabolic,HCC
Participants screenedwithhistory/physicals, biochemical testing
Eligibleparticipants - samedayMRI/QUSof liver
Trainingor validationgroup(stratiedrandomization)
Spearmanscorrelation(PDFF vs QUSBSC)
AUCsand95%DeLongcondenceintervals
Optimal QUSBSCcut-offs obtained; secondaryAUC
analysis conductedat other MRI-PDFF thresholds

BSCcut-offs de
B riv
S C ed
a tfrom
3

FOI

Participants
Liver

Reference
Phantom

T r a in in g S e t
V a lid a tio n S e t

30

M R I P D F F (% )

AIMS

RESULTS

25

0 .0 0 3 8 1 /c
C u t-o ff fo r

20
15
10
5
0
0 .0 0 0 0 1

0 .0 0 0 1

Figure4.CorrelationbetweenBS
optimal BSCcut-off.
WhenBSCiscorrelatedwithMRIgroupsof 0.82and0.79,P <.0001
theoptimal BSCcut-offwas 0.003

MRI Protocol
Correlation between BSC
and
MRIProtondensity
fat fraction
(PDFF,standardizedMR-based
quantitativeimagingbiomarker of steatosis) of liver estimatedat 3T
PDFF in training/validation Ugroups
sedadvancedmulti-echoMRI techniquethat
minimizesT1bias,corrects for T2*decay
(n = 204), at optimal BSC cut-off.
Addressesspectral complexityof protons infats
MRI-PDFF maps(reconstructedofine) usedto draw1cmradius
circular ROIsfor four right liveNoureddin
r lobesegments (5-8) to obtainPDFF
values,whicharethenaveraged

Figure2.QUSBSCimageswithcorrespondingMRI-PDFFsegmentationmaps.
Aeldof interest (FOI) isselectedwithinahomogenouspart of theliver (top).
Thisareaissubdividedintosub-regionof interests(sub-ROIs) for whichBSCis

Table1.Secondaryanalysisofvario
andvalidationgroups,withoptimal
95%condenceintervals inparenthe

Liver Multi-Scan
The next wave in imaging

Reference: Seattle Radiology PDF

Reference: Seattle Radiology PDF

Perspectum System

Pavlides et al; J Hepatology 2014

Predicting outcomes from LIF


(Liver Imaging Fibrosis) score

T2*

Reference: Seattle Radiology PDF

Case
Super morbidly obese woman with
pre-and post-bariatric surgery scan

CPT Code for Elastography?

Yes for MRE since 2009

0346T
$628 (CPT code. 74183)

Yes for TE 2013


Specifies the methodology for Fibroscan
(0346T, 91299)
The FibroScan 502 Touch device utilizes Vibration Controlled Transient
Elastography (VCTE) to aid in the clinical management of chronic liver
disease.

Payment:
CMS $54 outpatient, $154 hospital seeting

In Summary
Use multiple tests (7 or more) to stage your patients
PE, labs, US imaging, blood tests, endoscopy,
elastography, APRI, Fib-4

If you are using these tests to obtain HCV approval:


use the best test to prove advanced fibrosis
Do not let the insurance companies bully you into
doing a liver biopsy

For the Obese Patient Where


TE or SSI and US are Not Usable

Do MRE and fat % analysis

What Do I Do?

Always:

look at the platelet count


AST to ALT ratio
Spleen Size, PV diameter, liver texture, liver configuration
APRI score
FIB-4 score

HCV patients: where available


Elastography,
Phoenix, San Jose GI and Stanford practices

Complex patients: MR with elastography, fat and iron assessment

When to Biopsy
The obese patient who will not lose weight
Autoimmune hepatitis: all
Iron overload: meeting AASLD criteria with elevated
ferritin, and AST
Cryptogenic Liver Disease
FHF: to make a diagnosis or stage/grade disease
Other: granulomatous disease, infections, DILI,

Listservs:
rgish@robertgish.com
For weekly updates
Robertgish.com
For monthly newsletters HepaHealth

CLDF website
University of Washington HCV Project

Thank You To
Arizona GI Society

You might also like