PET Imaging of Hypoxia
John Humm, PhD
Dept. of Medical Physics
MSKCC
AAPM
Session: Introduction to Cancer Biology
Tuesday 20th July
Many strong hypoxia radiotracer
imaging programs
ƒ
University of Washington Seattle have been
studying 18F-FMISO since 1996.
ƒ
Washington University in St. Louis developed
Cu-ATSM.
ƒ
The group at Edmonton developed a family of
hypoxia compounds including IAZA, FAZA and
IAZGP
ƒ
UPENN developed 18F-EF5
What do PET images of hypoxia
look like?
60Cu-ATSM
Chao et al., IJROBP 49;1171, 2001
18F-FMISO
Lee et al.,
IJROBP 40;2,
2008
What can be gained from
hypoxia imaging?
(i) Is this tumor hypoxic?
Identification of tumor hypoxia as a prognostic indicator of
treatment outcome (Höckel et al. Can Res 56: 4509, 1996)
(ii) Where is it hypoxic?
Determination of the spatial distribution of hypoxia for the
purpose of radiobiologically driven dose painting (Ling et al
Int.J.Radiat.Oncol.Biol.Phys.47: 551, 2000)
Methods to measure hypoxia I
pO2 probes
Eppendorf & OxyLite
Frequency %
50
8-10 mm FSa-II Tumors
Total 175 Readings in
5 Tumors
40
30
20
pO2 probe
-
MicroPET images of hypoxia tracers
0
2.5
7.5
12.5
17.5
22.5
27.5
32.5
37.5
42.5
47.5
52.5
57.5
62.5
67.5
72.5
77.5
10
pO2 in mmHg
2
Methods to measure hypoxia II
Resect Tumor, Freeze
& Section
MicroPET images of hypoxia tracers
Immunohistochemistry &
Autoradiography
2
Methods to measure hypoxia III
MRI Gd-DTPA blood flow
Compare to other
Imaging Modalities
c
MicroPET images of hypoxia tracers
Electron Spin Parametric Resonance
Courtesy Howard Halpern
University of Chicago
2
Figure 8 :
131
I-
Methods to measure hypoxia IV
Hypoxia Responsive
Element (HRE)
RX/HIF1α
TG
Transfect tumor cell lines with
a plasmid containing a reporter
gene under the regulation of
HIF – in this case with a TkGFP
reporter construct
EcoRI
Autoradiograph
of the Hypoxia
Tracer
Autoradiograph
of the reporter
construct
18F-FMISO
124I-FIAU
18F-FMISO
Scans of H&N Patients
The concept of a GTVh
CT
FDG
FMISO
Lee et al, Fluorine-18-Labeled Fluoro-Misonidazole Positron Emission and
Computed Tomography Guided Intensity-Modulated Radiation Therapy in the
Treatment of Head and Neck Cancer. IJROBP 2008;70(1):2-13.
How do we define the hypoxic volume on
an 18F-FMISO PET image?
• What criterion do we use to segment the
hypoxic tumor sub-volume?
• How reproducible is an 18F-FMISO scan?
Effect of segmentation threshold
FDG-FMISO
T/B > 1.0
FDG-FMISO
T/B > 1.2
FDG-FMISO
T/B > 1.4
PET
VOXELS
Hoechst 33342 - BLUE
Pimonidazole - GREEN
Why are the 18F-FMISO uptake
ratios so low
Comparison of 2 pre-treatment
18F-FMISO scans
FMISO1 vs. FMISO2
Day0
FMISO image
Day3
FMISO image
Plot registered voxel
intensities from 1st FMISO
image with the 2nd
18F-FMISO
Dynamic PET
A(t) [Bq/cc]
Analysis of
Hypoxia
criterion
TumorBlood
Ratio(T:B)
≥1.3
not reliable
t [min]
Kinetic analysis of Time-Activity
Curves (TAC) is necessary
Thorwarth et al,
BMC Cancer. 2005
Dec 1;5:152.
A compartmental model to mimic
FMISO metabolism.
Plasma
k1
Cp
k2
Free/NS Bound
Bound
k3
C1
C2
[O2]
H&N Patient Dynamic PET Images
Carotid artery
1 min
2 min
3 min
4 min
10 min
15 min
20 min
25 min
5 min
30 min
tumor
90 min
95 min
180 min
185 min
CT
17
Parametric Images of
Cp(t)
k1
C1(t)
k3
18F-FMISO
C2(t)
k2
46 voxels
Concordance
CT
3hr FMISO
k3 map
k1 map
Parametric Images of
18F-FMISO
Discordance
CT
3hr FMISO
k3 map
k1 map
51 voxels
Validation of Parametric Images
Image Guided pO2 Measurements
pO2 tracks overlaid on the late and
parametric image maps
Track 1
Track 2
Track 3
Track 2
Track 3
T:B
Late Image
Track 1
k3
Hypoxia
Image
Correlation of parametric rate
constants to pO2 track values
k3
[min-1]
k3
k1
k1
0.14
0.12
k3
0.030
0.745
0.025
[ml/min/g]
K1 [ml/min/g]
-1
k3 [min ]
0.10
0.08
0.06
0.04
0.015
0.010
0.005
0.02
0.00
0.1
0.020
1
10
pO2 [mmHg]
100
0.000
0.1
1
10
pO2 [mmHg]
100
Conclusions
• Radiobiological dose painting may be a possibility but there
are many obstacles.
• Validation steps is still ongoing, but appear promising in spite
of the difficulties.
• Ideally we would like to be able to image the patient at one
single time point and obtain a map of the hypoxia distribution,
but kinetic analysis may be a necessary.
• At a minimum. hopefully we will be able to strategy patients
by hypoxic volume.
Acknowledgements
Dept of Medical Physics
C. Clifton Ling Ph.D.
Pat Zanzonico, Ph.D.
Joseph O'Donoghue, Ph.D.
Sean Carlin, Ph.D.
James Russell, Ph.D.
Sadek Nehmeh, Ph.D.
Andrei Pugachev Ph.D.
Wenli Wang, Ph.D.
Rachel Bartlett, Ph.D.
Nuclear Medicine Service
Steven M Larson, MD
Heiko Schöder, MD
Cyclotron / Radiochemistry
Jason Lewis, Ph.D.
Shangde Cai, Ph.D.
Howard Sheh, M.S.
Peter Smith-Jones Ph.D.
Radiation Oncology
Nancy Lee, MD
Gloria Li, Ph.D.