Role of imaging in oncology
UNIVERSITY OF WISCONSIN
SCHOOL OF MEDICINE
AND PUBLIC HEALTH
ANATOMICAL
IGRT
MOLECULAR
Positron Emission Tomography
for Treatment Assessment
BEFORE
THERAPY
AFTER
| | | | | …... | |
DIAGNOSIS
STAGING
Robert Jeraj
UW Paul P. Carbone Comprehensive Cancer Center,
Departments of Medical Physics, Human Oncology and Biomedical Engineering,
University of Wisconsin, Madison, Wisconsin, USA
Jozef Stefan Institute, Ljubljana, Slovenia
Why treatment assessment?
Based on early treatment assessment one could
modify treatment:
– If likely not successful:
• Escalate therapy
• Change therapy
• Selectively add dose (RT)
– If likely successful:
• Descalate therapy
• Stop therapy early
Enormous benefits for the patient - improved tumor
control, reduced side effects, and costs
TARGET DEF.
RESPONSE EVALUATION
RT PLANNING
TUMOR RESP., EARLY TOX
days, weeks
LOCAL FAILURE, LATE TOXICITY
months, years
What is used for
treatment assessment now?
Histopathological evaluation:
– Typically by regression score(viable tumor vs.
fibrosis), e.g., Salzer-Kuntschik for osteosarcomas
– Limitations: need complete resection, problem with
point biopsies because of tumor heterogeneities
Radiological evaluation:
– Defined by the therapy-induced reduction of
tumor size: WHO, RECIST
– Limitations: numerous
1
Anatomic response criteria
WHO (Miller, Cancer, 207, 1981):
– the size of a tumor should be estimated based on
two perpendicular diameters
– positive tumor response to therapy should be
defined as a reduction of at least 50% in the
product of these two diameters
RECIST (Response Evaluation In Solid Tumors)
(Therasse, JNCI, 205, 2000):
– The size of a tumor is estimated based on
unidimensional
– Positive tumor response to therapy is at least 30%
decrease in the largest dimension of the tumor
Ideal…
RECIST
Complete Response (CR): Disappearance of all disease compared
with the baseline examination, whether it be measurable or not
Partial Response (PR): A reduction in the sum of the greatest
lengths of individual tumors by at least 30% compared with the
baseline measurement
Stable Disease (SD): This lies between the definitions for PR
and PD
Progressive Disease (PD): An increase in the total length of all
measurable lesions of more than 20% compared with the
smallest sum of lesion sizes (not the baseline values), or the
appearance of unequivocal new disease
… and reality
Inherent inconsistencies in
expert observers: 15-40%
Jaffe, J Clin Oncol, 24, 3245 (2006)
Jaffe, J Clin Oncol, 24, 3245 (2006)
2
Other issues…
Visual example
Tumor shrinkage is only the final step in a complex
cascade of cellular and subcellular changes after
treatment
Pre-treatment
3 months post treatment
Several cycles of therapy (radiotherapy fractions,
chemotherapy) are needed before treatment
response can be assessed by anatomic imaging
Residual mass is often present after treatment – it is
hard to differentiate between viable tumor
posttreatment changes, such as scarring and fibrosis.
CT scan and comparable FDG-PET scan in a patient with gastro intestinal
stromal tumor (GIST) with hepatic metastatic lesions
Gwyther, Eur J Nucl Med Mol Imaging, 33, S11 (2006)
Prognostic relevance of FDG-PET
AFTER therapy
Tumor
Ref
No
pat
Head neck
Kunkel 2003
35
Esophagus
Lung
Flamen 2002
Survival
respond (mo)
36
>60
>34
Survival
non-resp (mo)
18
7
Swisher 2004
70
>24
14
Post-treatment
metabolic
response
is
MacManus 2003
73
>36
<12
highly predictive of overall survival
Hellwig 2004
47
56
19
Cervix
Grigsby 2004
152
>45
<20
Lymphoma
Several 2001-04
>200
(>30,>60)
3-30
All chemoradiotherapy except lymphoma (chemo only)
Prognostic relevance of FDG-PET
DURING therapy
P
Tumor
Ref
No
pat
Crit
Head neck
Brun 2002
47
50%
>120
40
0.004
Esophagus
Weber 2001
37
35%
>48
20
0.04
0.002
0.005
0.01
0.001
<0.001
<0.001
Lung
Survival
respond (mo)
Survival
non-resp (mo)
Mid-treatment
metabolic
response
is
Wieder 2004
20 30%
>38
18
highly
predictive
of
overall
survival
Weber 2003
57 20%
9
5
P
0.01
0.005
Stomach
Ott 2003
35
35%
>48
17
<0.001
Lymphoma
Kostakoglu 2002
30
Vis
>24
5
<0.001
All chemotherapy
3
FDG-PET and radiation therapy
FDG-PET and radiation therapy
16
16
14
12
12
10
10
SUVmax
SUVmax
NSCLC
14
Average
8
6
NSCLC
Metabolic non-responders
Average
8
6
Metabolic responders
4
4
2
2
RT
0
RT
0
0
10
20
30
40
50
60
70
0
Time [days]
30
40
50
60
70
Baardwijk, Radiother Oncol, 82, 145 (2007)
Molecular imaging targets in oncology
FLT
Rapid cellular proliferation
Subverted cellular regulation:
regulation:
-Intracellular signaling
-CellCell-toto-cell signaling
-Extracellular matrix
signaling
20
Time [days]
Baardwijk, Radiother Oncol, 82, 145 (2007)
FRGD
10
Response to chemotherapy in AML
Antracycline, cytarabine and/or etoposide
FDG
Increased cellular metabolism:
metabolism:
-Increased glycolysis
-Increased aminoamino-acid
metabolism
Evading cellular death
Altered tumor microenvironment:
-Hypoxia
-Changes in perfusion
-Changes in diffusion
FAnnexin V
CuATSM
FLT-PET/CT
Responder
FLT-PET/CT
Non-responder
4
Response to targeted therapy in HNSCC
Avastin (bevacizumab) therapy
Pre-Avastin
Tumor response to radiation therapy
can be very dynamic
SUV
7
Day 3
Pre
Day 6
SUVmax = 4.2
0
CuATSM-PET/CT
Post-Avastin
FLT-PET/CT
SUVmax = 7
CuATSM-PET/CT
Spatial response assessment
Pre-RT FLT-PET
Treatment plan
Spatial response assessment
Pre-RT FLT-PET
Mid-RT FLT-PET
5
Anatomical change more important then
biological change
Other challenges…
Pre-RT
cm
1. 6
3
0
0
cm
Mid-RT CT
Pre-RT CT
The problem of partial volume effects
Pre-RT
SUV
3
SUV
0.4
Post-RT
Partial volume effects are complicated
SUV
1.05
0.4
1.00
0.95
0.3
0.3
Recovery Coefficient
2. 3
Post-RT
SUV
0.90
0.85
0.80
0.75
IRCTAC
OSEM2
OSEM2 nofilter
Ramp
0.70
0.65
0.60
0.55
0.50
8
10
12
14
16
18
20
22
24
26
Sphere Diameter
6
Pre-treatment
Treatment response can reveal
unwanted results
Mid-treatment
(1 wk of XRT)
Mid-treatment
(1 wk of XRT)
Pre-treatment
Treatment response can be
extremely powerful
Assessment of normal tissue damage
Pre-RT
Post-RT
Assessment of normal tissue damage
Pre-RT
Post-RT
7
Conclusions
Early treatment assessment has enormous
potential to change patient care
Anatomical imaging used for treatment
assessment has many limitations
Functional/molecular imaging is more powerful
and versatile in assessment of treatment
response
Tumor response kinetics can be very complex
Need for clinical trials with extensive
functional/molecular imaging component
Thanks to:
Image-guided therapy group:
– Vikram Adhikarla
– Dave Barbee
– Steve Bowen
– Ryan Flynn
– Joseph Grudzinski
– Keisha McCall
– Matt Nyflott
– Manuel Rodriguez
– Urban Simoncic
– Benny Titz
– Matt Vanderhoek
Radiology
– Scott Perlman
– Jamey Weichert
– Chris Jaskowiak
Medicine/UWCCC
– George Wilding
– Glenn Liu
– Dawn Church
Medical Physics
– Jerry Nickles
– Onofre DeJesus
Human Oncology
– Søren Bentzen
– Paul Harari
– Mark Ritter
– Minesh Mehta
– Wendy Walker
Veterinary School
– Lisa Forrest
– David Vail
Hematology
– Mark Juckett
– Brad Kahl
– Nancy Turman
NIH, NSF, Susan Komen
Foundation, UWCCC
8