QUANTITATIVE IMAGING OF HUMAN LIVER IRON CONCENTRATIONS IN VIVO

1
QUANTITATIVE IMAGING OF HUMAN LIVER IRON
CONCENTRATIONS IN VIVO
Tim St Pierre1, A Fleming1, W Chua-anusorn1, P
Clark1, E Rossi, G Jeffrey2,3, J Olynyk2
2
Declaration of competing interests

• Three of the investigators on this project are
  involved with a commercial venture based on
  non-invasive measurement of liver iron
  concentrations
• Tim St Pierre
• Paul Clark
• Wanida Chua-anusorn

3
Standard methods of assessingthe body iron burden

• Ferritin serum assay
• Needle biopsy of the liver

4
Liver biopsy
5
Non-invasive measurement of liver iron
concentration
6
The magnetic resonance imager (MRI)
7
A liver full of magnets!
8
Proton transverse relaxation rate (R2) image and
distribution
Clark et al, Mag Res Imaging 18 (2000) 431-438.
9
Liver R2 images and distributions
healthy volunteer 3 iron loaded subjects with
sequentially increasing liver iron concentrations
10
R2 image analysis of human liver samples
Iron loaded liver tissue dissection outline
11
R2 image analysis of human liver samples
Clark et al, Magn Reson Med 49 (2003) 572-575
12
R2 image analysis of human liver samples
Clark et al, Magn Reson Med 49 (2003) 572-575
Mean R2 vs iron concentration for 32 cubes of
liver dissected from a single iron loaded liver
post mortem.
13
Measurement of liver R2 in vivo
Needle biopsy samples a few milligrams of tissue
from the right hand side of the liver
14
Relationship between ltR2gt in right hand side of
liver and needle biopsy iron concentration (dry
wt)
St Pierre et al (2005) Blood 105, 855-861
15
Sensitivity and specificity of R2-LIC
measurements to biopsy LIC measurement
LIC Threshold mg Fe/g dry Clinical Relevance Sensitivity Specificity
1.8 Upper 95 of normal 94 (86-97) 100 (88-100)
3.2 Suggested lower limit of optimal range for liver iron concentrations for chelation therapy in transfusional Fe overload 94 (85-98) 100 (91-100)
7.0 Suggested upper limit of optimal range for liver iron concentrations for transfusional Fe overload and threshold for increased risk of iron induced complications 89 (79-95) 96 (86-99)
15.0 Threshold for greatly increased risk for cardiac disease and early death in patients with transfusional iron overload 85 (70-94) 92 (83-96)
Olivieri NF, Brittenham GM. Blood. 1997 89
739-761
16
Measurement of R2 standards
17
Measurement of R2 standards
St Pierre et al (2005) Blood 105, 855-861
R2 vs paramagnetic Mn2 concentration for the
same series of phantoms measured on 13 different
1.5T MRI scanners
18
Reproducibility of liver R2 measurements on 2 MR
scanners
Random error 7.7
Systematic error 1.2
St Pierre et al, NMR in Biomed 2004 17, 446-458
19
Assessment of liver damage risk from iron loading
20
Using age and LIC at diagnosis to predict
fibrosis grade in HH
Olynyk et al (2005) Am. J. Gastroenterol 100, 837
21
Using age and LIC at diagnosis to predict
fibrosis grade in HH
Olynyk et al (2005) Am. J. Gastroenterol 100, 837
22
Relationship of R2 distribution with liver
histology
250 mm
Perls Stain
250 mm
Reticulin Stain
Clark et al, Magn Reson Med 49, (2003) 572-575
23
Relationship of R2 distribution with liver
histology
250 mm
Perls Stain
250 mm
Reticulin Stain
Clark et al, Magn Reson Med 49, (2003) 572-575
24
Relationship of R2 distribution with liver
histology
250 mm
Perls Stain
250 mm
Reticulin Stain
Clark et al, Magn Reson Med 49, (2003) 572-575
25
R2 distribution and liver biopsy histology
Non-cirrhosis LIC 10.2 mg/g DW R2 99 20
Cirrhosis LIC 12.6 mg/g DW R2 132 50
26
R2 distribution and liver biopsy histology
Mild Fibrosis LIC 1.1 mg/g DW R2 30.9 7.4
Cirrhosis LIC 1.8 mg/g DW R2 29.9 13.4
27
Relationship between LIC and total body iron
stores in hereditary hemochromatosis
28
Serum ferritin and body iron stores in hereditary
hemochromatosis
Olynyk et al (1998) Am. J. Gastroenterol. 93, 346
29
LIC by biopsy vs total body iron stores by
quantitative venesection
Hereditary Hemochromatosis
Olynyk et al (1998) Am. J. Gastroenterol. 93, 346
Summers et al (1990) Hepatology 12, 20
30
Body iron store distribution
Total body iron store liver iron store
extrahepatic iron store
Liver iron store liver iron concentration x
liver volume
31
Measuring total body iron stores (TBIS) and
liver iron stores

• Measure LIC 3 times during venesection
• Measure liver volume
• Weighted fit through data
• Extrapolate LIC to zero to obtain TBIS
• Extrapolate LIC to 1 mg Fe/g dw for comparison of
  TBIS with other studies

32
Subjects methods

• Newly diagnosed hereditary hemochromatosis
  subjects were recruited (n19) (male 12, female
  7)
• 17 C282Y homozygotes, 1 C282Y/H63D, 1
  Wild-type/H63D
• LIC measured with R2-MRI 6
• Liver volume measured with MRI simultaneously
• Subjects recalled for follow-up LIC measurements
  at estimated half way point in venesection
  schedule and near end of schedule

33
Non-invasive monitoring of LIC during venesection
Subject 1
Subject 2
Subject 3
Subject 4
LIC measured using R2-MRI. Dashed line is upper
limit normal
Subject 5
Subject 6
Subject 7
Solid lines are weighted fits to the data
yielding estimate for initial LIC and total body
iron store (from mL of blood to reach zero LIC).
34
LIC vs TBIS by quantitative venesection in HH
Biopsy Measurements
Olynyk et al (1998) Am. J. Gastroenterol. 93, 346
Summers et al (1990) Hepatology 12, 20
35
LIC vs TBIS by quantitative venesection
Biopsy Measurements (thalassemia)
Angelucci et al (2000) N Eng J Med 343, 327
36
LIC vs TBIS by quantitative venesection
Biopsy Measurements (thalassemia)
R2-MRI Measurements (hereditary hemochromatosis)
Angelucci et al (2000) N Eng J Med 343, 327
37
Body iron store distribution
Total body iron store liver iron store
extrahepatic iron store
Liver iron store liver iron concentration x
liver volume
38
a is fraction of total body iron store in the
liver
Total body iron store liver iron store / a
39
Fraction (a) of iron store in liver

• Mean fraction of iron stores in liver is 45
• Range 26 - 80

40
New model for iron distribution in HH

• Assume fraction, ? , of total body iron store in
  the liver varies linearly with rate of liver iron
  loading (LIC/Age)

41
New model for predicting iron stores in HH
95 limits of agreement
New Model 40
Using LIC 72
Using Ftn 102
42
Summary (1)

• R2 imaging
• can be used to measure non-invasively liver iron
  concentrations (LIC) with known accuracy and
  precision
• has good sensitivity and specificity for
  measurement of LIC at both low and high LIC
  ranges
• has dynamic range of measurement from normal LIC
  to the very highest concentrations encountered in
  clinical practice
• works on most 1.5 T MRI units
• has health regulatory authority clearance for LIC
  measurement in USA (FDA), Europe (CE Mark), and
  Australia (TGA)
• may have the potential to detect liver
  cirrhosis/fibrosis

43
Summary (2)

• Non-invasive measurement of LIC
• Enables serial monitoring of patients on blood
  transfusion and chelation therapy to aid in
  chelation dose determination
• Aids in the identification of newly diagnosed
  hemochromatosis patients who are at risk of iron
  induced liver damage

44
Summary (3)

• Non-invasive measurement of LIC and liver volume
• Has demonstrated a correlation between fraction
  of TBIS in the liver and rate of iron loading in
  HH
• Enables a more accurate prediction of venesection
  requirements for HH subjects compared with LIC or
  serum ferritin measurements

45
Co-workers

• School of Physics, University of Western
  Australia
• Paul Clark
• Wanida Chua-anusorn
• Adam Fleming
• School of Medicine, University of Western
  Australia
• Gary Jeffrey
• John Olynyk
• Ric Rossi
• Thalassemia Research Center, Mahidol University
• Pensri Pootrakul
• Department of Haematology, Prince of Wales
  Hospital
• Rob Lindeman
• SKG Radiology
• Erin Robins