8/15/2011
IMAGING THE BIOLOGICAL
EFFECT OF DOSE IN THE
LIVER NORMAL TISSUE
Yue Cao, Ph.D.
Departments of Radiation Oncology
and Radiology
University of Michigan
Normal Liver Toxicity
RILD is a limiting factor for high dose RT of intrahepatic
cancers
Clinical complications usually present 2 -6 weeks after the
completion of therapy, with devastating outcomes
Radiation-induced tissue toxicity is a complex, dynamic and
progressive process
Individual sensitivity to radiation limits the utility of existing
NTCP models
Acknowledgements
• James
Balter, Ph.D.
• Joel F. Platt, MD
• Edgar Ben-Josef, MD
• Zhou Shen, MA
• Thomas L. Chenevert, Ph.D.
• Randall Ten Haken, Ph.D.
• Mary Feng, MD
• Hesheng Wang, Ph.D.
• Isaac R Francis, MD
• Database group
• Kirk Frey, MD, Ph.D.
• Tim Johnsoon, Ph.D.
• Ted S. Lawrence, MD,Ph.D.
• Dan Normolle, Ph.D.
• Charlie Pan, MD
NIH/NCI 3 P01 CA59827
NIH/NINDS/NCI RO1NS064973
NIH/NCI RO1 CA132834
Functional Imaging
A valuable tool for evaluation of normal tissue
toxicity
• Temporal changes in normal tissue from pre to during and
to post RT
• Spatially-resolved volumetric distribution of function of
subunits
• Individual sensitivity to dose
• pre-RT dysfunction of the live or functional reserve
Early assessment and prediction of potential
radiation risks and clinical outcomes
Determination of specific risks versus benefits of treatment
should be an integral part of clinical decision making for each
patient
Cao 2011
Re-optimization of individual patients’ treatment
plans and early intervention involving tissue/organ
protection
1
8/15/2011
Liver Functional Imaging
Liver Functional Imaging
Hepatic perfusion imaging (DCE MR and CT)
• Arterial and portal venous perfusion
• → Histopathology of RILD is venous occlusion
Hepatic perfusion imaging (MR and CT)
• Arterial and portal venous perfusion
1
• → Histopathology of RILD is venous occlusion
99mTc HIDA SPECT
250 ml/100g/min
45 ml/100g/min
250 ml/100g/min
• Hepatic extraction fraction and biliary function
Eovist MRI
• Hepatic extraction fraction and biliary function
99mTc galactosyl serum albumin (GSA) scintigraphy or
99mTc HIDA SPECT/CT
• Hepatic extraction fraction and biliary function
Eovist MRI
• Hepatic extraction fraction and biliary function
99mTc galactosyl serum albumin (GSA) scintigraphy or
SPECT
SPECT
• Hepatocyte binding via the asialoglycoprotein receptor
(Schneider, Surg clin N Am 2004)
0 ml/100g/min
Liver Functional Imaging
Hepatic perfusion imaging (MR and CT)
• Arterial and portal venous perfusion
Kubo, of
J RILD
Surg isRes
2002occlusion
• → Histopathology
venous
Regeneration
99mTc HIDA SPECT
• Hepatic extraction fraction and biliary function
Eovist MRI
• Hepatic extraction fraction and biliary function
99mTc galactosyl serum albumin (GSA) scintigraphy or
SPECT
Hepatocyte
receptor
GSA planar •image
prior binding
to rightvia the asialoglycoprotein
GSA planar image
2 weeks after
(Schneider, Surg clin N Am 2004)
portal vein embolization
right portal vein embolization
18F-Deoxy-Galactose (FDGal) PET
• High uptake in tumor cells
• Response in normal liver to dose (Hoyer, ESTRO 2011)
Cao 2011
• Hepatocyte binding via the asialoglycoprotein receptor
(Schneider, Surg clin N Am 2004)
0 ml/100g/min
18F-Deoxy-Galactose (FDGal) PET
Total perfusion
perfusion Portal vein perfusion
• High uptake inArterial
tumor cells
• Response in normal liver to dose (Hoyer, ESTRO 2011)
0 ml/100g/min
18F-Deoxy-GalactoseHepatic
(FDGal)Extraction
PET
CTin tumor cells
• High uptake
Fraction
• Response in normal liver to dose (Hoyer, ESTRO 2011)
Liver Functional Imaging
Hepatic perfusion imaging (MR and CT)
• Arterial and portal venous perfusion
• → Histopathology of RILD is venous occlusion
99mTc HIDA SPECT
• Hepatic extraction fraction and biliary function
Eovist MRI
• Hepatic extraction fraction and biliary function
99mTc galactosyl serum albumin (GSA) scintigraphy or
SPECT
• Hepatocyte binding via the asialoglycoprotein receptor
(Schneider, Surg clin N Am 2004)
18F-Deoxy-Galactose (FDGal) PET
• High uptake in tumor cells
• Dose response in normal liver (Hoyer, ESTRO 2011)
2
8/15/2011
Liver Functional Imaging at UM
Hypotheses
 Changes in portal venous perfusion and/or hepatobiliary
Prospective perfusion CT/MRI and HIDA SPECT
protocols
function during the early course of RT are potentially a
biomarker for liver dysfunction after irradiation
 Patients with intrahepatic cancers and treated with RT
 Patients at high risk for liver injury
 Large tumor volume
 Primary cancers, e.g., HCC with or without cirrhosis
 previous treatments, e.g., TACE, RFA, resection, or RT
 Perfusion and/or hepatobiliary function biomarkers may
allow us to select patients who are susceptible to liver
injury prior to clinical symptoms and therefore to modify
treatment
Assessment of Individual and spatial sensitivity to doses by
functional imaging during the course of therapy allow us to
adapt treatment plan to minimize local tissue damage and
thereby to prevent from organ injury
Venous Perfusion Dysfunction
DCE CT/MRI and HIDA SPECT
 Pre-treatment, mid-course (50-60%), and post RT (1 and 2
months after treatment completion)
 DCE MRI covers the whole liver
Overall liver function assessment
 Indocyanine green (ICG) clearance or retention: the best
established test for overall liver function
Dose Effect on Venous Perfusion
One month after RT
After 45 Gy (during RT)
30Gy
20Gy
10Gy
Fp after 30 Fx
(ml/100g/min)
40Gy
2.5 ml/100g/min per GY
1.6 ml/100g/min per GY
250
200
150
100
50
0
R = 0.47
y = -0.016x
+ 129.3
p<0.0001
0
2000
4000
6000
dose at the time of scan (cGy)
8000
8000
Fp 1 month after RT
(ml/100g/min)
170 ml/100g/min
250
R = 0.77
p<0.0001
200
150
100
50
0
0
2000
4000
6000
dose at the end of RT (cGy)
8000
0 ml/100g/min
Prior to RT
After 45 Gy
30 fx of 1.5 Gy/fx twice daily
Cao Y et al , Medical Physics 2007
Cao 2011
Note: (1) time dependent slopes – time dependent response
(2) Individual variation -- individual sensitivity
3
8/15/2011
Individual Sensitivity to
Radiation
Y-intercept or low
Individual Sensitivity to Radiation
X-intercept
Pts
Individual patient
Slope:
reduction in
perfusion caused by
unit dose
Fp 1 month after RT
160
120
100
60
40
y = -0.0423x + 229.7
R2 = 0.85
20
0
X-intercept:
1000
2000
3000
4000
5000
6000
dose at the end of RT (cGy)
Cao, et al, Int J Rad Onc Biol Phys, 2008
X-intercept
Slope
mL/(100 g min)
per Gy
Cao 2011
NA
NA
2
-2.6
60
3
-3.2
51
4
-2.2
63
5
-6.5
46
6
-4.2
60
7
-2.8
68
8
-1.1
74
9
-4.2
54
10
-1.3
81
11
-1.2
84
Overall Liver Function vs
Functional Subunits
Liver Functional Volume
Pts
1
80
0
critical dose resulting
in undetectable
venous perfusion
Dose (Gy)
Fp=0
140
Dose (Gy)
Fp=0
LV%
Fp=0
FLV%
Fp>0
0%
1
NA
NA
NA
2
-2.6
60
11
89%
3
-3.2
51
0
100%
4
-2.2
63
23
77%
5
-6.5
46
39
61%
6
-4.2
60
32
68%
7
-2.8
68
6
94%
8
-1.1
74
3
97%
9
-4.2
54
31
69%
10
-1.3
81
0
100%
11
-1.2
84
0
100%
Substantially Reduced
Venous Perfusion
1
Fit 
n
N
w
n 1
n
Pitn  1
n
N
w
n
(  it D int  a it   t )
n 1
130
Venous perfusion at a subunit of
functional liver (>critical value)
Overall liver function
(ICG clearance)
130
0
Functional volume
0
mean Fp post RT(Fp>20)
dose region:
reperfusion
Slope
mL/(100 g min)
per Gy
160
R2 = 0.89
140
120
P<0.001
100
80
60
40
20
0
2
4
6
8
10
12
T1/2 ICG post RT (min)
4
8/15/2011
mean D at the end of RT
Mean Liver Dose vs Liver
Functional Reserve
5000
R2 = 0.10
4500
NS
4000
Volumetric MRI Perfusion
17 patients in this analysis
 7 mets, 7 HCC, and 3 cholangio
 6 with previous treatments (TACE, RFA, SBRT or resection)
 Tx: 5 by SBRT, 12 by fractionated RT
Imaging and liver function tests
3500
3000
• Pre RT, 50%-60% planned doses, 1 and 2 months after RT
2500
Doses biologically corrected (LQ model w /=2.5) for
different fractional sizes
2000
2
4
6
8
10
12
T1/2 ICG post RT (min)
Mean of Portal Venous Perfusion of
Subunits and Overall Liver Function
Liver function pre RT assessed by ICG clearance
 5 pts: T1/2 > 10 min (high risk)
 10 pts: T1/2 ranged from 3.7 to 7 min (normal or close to normal)
Changes in Venous Perfusion
after RT
hyperperfusion
160
r=0.72, p<1.0x10-8
10 Gy
20 Gy
30 Gy
50 Gy
40 Gy
0
Pre RT
Cao 2011
One month post RT
5
8/15/2011
hyperperfusion
Cholangio treated
by Fx RT
met treated by Fx RT
Summary
Hypoperfusion pre RT and
Reperfusion after RT
Four patients:
Overall
Hypoperfusion
Mean perfusion in
the liver <
60 mL/(100g min)
Three patients:
Overall and
regional recovery
After RT, except in
the highest-dose
regions
venous perfusion across individual patients and over lobes,
segments or regions
0.3
One month
after RT
HCC treated with Fx RT
0.03
HEF
ImpResp
liver
VOI activity
There are large variations in dose responses of portal
Pre RT
99mTc-HIDA SPECT/CT in the
Liver
Venous perfusion imaging could be a biomarker for local
liver function
180
f it
B lood
0.2
Excretion rate
0.1
Activity
Individual Response Function
0.02
0.01
0
0
0
10
20
30
40
Tim e (m in)
50
60
0
10
20
30
40
50
60
Tim e (m in)
Significant hyperperfusion was observed. The potential
relevance for normal tissue sparing/regeneration is worthy of
further investigation
Cao 2011
99mTc-HIDA is an established agent for assessment of
hepatobiliary function using SPECT.
hepatic extraction fraction (HEF)
bile excretion rate
6
8/15/2011
50
50
effect on k1 (%)
1
Ef f ect o n H EF ( % )
Dose Effects on HEF and Bile
Excretion Rate
Hepatobiliary Function
40
30
20
Phys D o se
B io D o se
10
0
0
50 0 0
Do s e ( cGy)
Tx Planning
CT
Co-registered Co-registered HEF preRT
CT pre RT
CT 1m postRT
(SPECT/CT) (SPECT/CT)
HEF 1m
postRT
Summary
Functional and metabolic imaging is emerging as a promising
tool for risk management, particularly for patients at high-risk for
liver injury
Dose-response of liver measured by functional imaging may
depend upon treatment regimes
Functional/metabolic imaging agent uptake may depend upon
enzyme levels in the liver
10 0 0 0
40
30
20
Phys Dose
BioDose
10
0
0
5000
10000
Dose( cGy)
Every Gy causes a reduction:
0.87% in HEF when the dose is greater than ~15 Gy
0.76% in bile excretion rate
Summary
Our understanding of the histopathology and
biologic processes of radiation-induced
tissue/organ toxicity is quite limited
Functional and molecular imaging are potential
biomarkers for early assessment and prediction of
radiation-induced tissue/organ toxicity
Clinical trials with adequate endpoints are
warranted to establish the value of these methods
Whether other hepatic functional/metabolic imaging provides any
discriminatory information beyond portal venous perfusion has
yet to be demonstrated.
Cao 2011
7
8/15/2011
Hyperperfusion after liver irradiation →
Regeneration?
1
Kubo, J Surg Res 2002
GSA planar image prior to right
portal vein embolization
CT
Regeneration
GSA planar image 2 weeks after
right portal vein embolization
Hepatic Extraction
Fraction
Dose-Response of Portal
Venous Perfusion
Cao 2011
8