Intratumoral mRNA expression of genes involved in angiogenesis and HIF-1 pathway to predict outcome to
VEGFR tyrosine kinase inhibitor (TKI) in patients enrolled in CONFIRM1 and CONFIRM2
Peter M. Wilson, Ph.D.D. Yang, M. M. Shi, W. Zhang, C. Jacques, J. C. Barrett, K. Danenberg,
T. Trarbach, G. Folprecht, G. Meinhardt, H. J. Lenz Department of Pathology and Division of Medical Oncology
GI Oncology Program
The Sharon E Carpenter Laboratory
USC Norris Comprehensive Cancer Center
USC Keck School of Medicine
Los Angeles, CA.
Disclosures
Response Genetics
Novartis Pharmaceutical Corp
Bayer Schering Pharma
Background
Colorectal cancer is 3rd most common cause of cancer in the US.
In 2008 there will be an estimated†: 149,000 cases 50,000 deaths
Prognosis depends on: Stage Patient performance status Tumor differentiation Molecular Markers (MSI, 18q status)
† - American Cancer Society – Estimated New Cancer Cases and Deaths, US, 2008
CONFIRM Trials
PTK/ZK – Oral anti-angiogenic agent1
Competitive inhibitor at the ATP-binding site of VEGF receptors 1-3, platelet-derived growth factor and c-kit.
1. Wood JM, Bold G, Buchdunger E, et al. Cancer Res, 2000; 60:2178-89.
Randomized, double-blind, placebo-controlled, phase III trials in patients with metastatic adenocarcinoma of the colon or rectum.
VEGFR-1
/ Flt-1
VEGFR-2
/ KDR
VEGFR-3
/ Flt-4
• Angiogenesis• Lymphangiogenesis
• Metastasis
X X X X X XPTK/ZK PTK/ZK PTK/ZK
VEGF-AVEGF-B
PIGF
VEGF-AVEGF-CVEGF-D
VEGF-CVEGF-D
PTK/ZK Inhibits All Known VEGF Receptors
Extracellular
Intracellular
Inhibition of angiogenic signals1
Inhibition of metastasis
Sensitization to chemo-therapeutics2
2. Sini P, Samarzija I, Baffert F, et al. Cancer Res 2008;68:1581-1592.
Bevacizumab
1. Wood JM, Bold G, Buchdunger E, et al. Cancer Res, 2000; 60:2178-89.
CONFIRM Trials
Randomized
FOLFOX4/PTK 585 Patients
FOLFOX4/Placebo 583 PatientsCONFIRM 1
• 1ST Line• 1168 Patients
Randomized
FOLFOX4/PTK 429 Patients
FOLFOX4/Placebo 426 PatientsCONFIRM 2
• 2ND Line• 855 Patients
Progressed from irinotecan-based
therapy
Stratification Factors
PS 0, 1-2
LDH ≤, > 1.5 x ULN
ULN – Upper limit of normal; PS – Performance status; LDH – Lactate dehydrogenase
Response to PTK/ZK
Overall Response
Complete Response
Partial Response
Stable Disease
Progressive Disease
PTK/ZK Placebo
46 47 19 18
4 5 2 1
42 43 17 17
32 32 42 43
FOLFOX4 +
17 17 34 37
CONFIRM1 CONFIRM2
PTK/ZK Placebo
PFS and OS in CONFIRM1
Overall Survival High LDHProgression Free Survival High LDH
Hecht et al. European Cancer Organization,2007
No statistically significant increase in PFS or OS in total population with PTK/ZK treatment
PFS and OS in CONFIRM2
Overall Survival High LDH
Progression Free Survival Progression Free Survival High LDH
Kohne et al. Journal of Clinical Oncology, 2007 ASCO Annual Meeting Proceedings Part I. Vol 25, No. 18S, 2007: 4033
Overall Survival
M
Clinical Significance of LDH
Patients with high serum LDH appear to benefit most from PTK/ZK
Facilitates anaerobic glycolysis Marker for increased tumor burden and
aggressive disease Associated with poor prognosis
LDH may be a surrogate marker for HIF-1 signaling and proangiogenic propensity
M
M
M
LDHA Gene
M
LDH5
Enzyme
Koukourakis et al. Journal of Clinical Oncology, 2007 ASCO Annual Meeting Proceedings Part I. Vol 25, No. 18S (June 20 Supplement), 2007: 4107
Koukourakis et al. Journal of Clinical Oncology, 2006 Sep 10;24(26):4301-8.
Koukourakis et al. Clin Exp Metastasis. 2005;22(1):25-30.
LDH Correlates with HIF1 Signaling
Serum LDH level was associated with activation of HIF-related
intratumoral gene expression in 92 specimens from 36 mCRC patients
Azuma et al. Pharmacogenomics. 2007 Dec;8(12):1705-13.
VEGFR1
VEGFA
Serum LDHp=0.006
p=0.004
mCRC – metastatic colorectal cancer
HIF2α
VEGF
LDHA
p=0.001
HIF1α
Glut1
p=0.013p=0.044
p=0.007
Intratumoral
Objectives
Identify a panel of molecular markers predicting outcome to PTK/ZK in patients enrolled in CONFIRM1 and CONFIRM2
To measure intratumoral mRNA expression of genes involved in: Hypoxia Glycolysis Angiogenesis
HIF-1 Signaling
Gene Expression
Oncogene Signaling
PI3KPKCRASSRC
HIF-1α
HIF-1α
Hypoxia
Nucleus
VEGF (angiogenesis)
GLUT1 (glucose
transport)
LDH-A
(glycolysis)
EPO
(erythropoiesis)
VHL
NOS
ARNT
HRE
Normoxia
VHL
Methods I
191 formalin-fixed paraffin-embedded tumor samples were obtained from CONFIRM 1 and 2.
CONFIRM 1• 85 Patients
FOLFOX4/PTK 52 Patients
FOLFOX4/Placebo 54 Patients
CONFIRM 2• 106 Patients
FOLFOX4/PTK 43 Patients
FOLFOX4/Placebo 42 Patients
Specimens obtained at time of diagnosis.
Patient Characteristics
CONFIRM1 CONFIRM2
PTK/ZK Placebo PTK/ZK Placebo
Characteristics (n=43) (n=42) (n=52) (n=54)
Median age, y 63 62 64 62
Male/Female 31/12 25/17 34/18 40/14
(72%/28%) (60%/40%) (65%/35%) (74%/26%)
Low / High LDH 35/8 31/11 36/16 43/11
(81%/19%) (74%/26%) (69%/31%) (80%/20%)
WHO PS, 0/1 or 2 23/20 26/16 25/27 29/25
(53%/47%) (62%/38%) (48%/52%) (54%/46%)
Patient characteristics for CONFIRM 1 and CONFIRM 2 are well balanced between treatment groups, all p-values >0.05
Methods II
RNA Extracted
Reverse Transcription
PCR with TaqMan®
Data Analysis
RNA
cDNA
Laser Capture Micro-dissection
Candidate Genes
PTK/ZK
HIF-1α
Nucleus
ANGIOGENESIS• VEGF
• VEGFR 1/2
GLYCOLYSIS• Glut1
HYPOXIA• HIF1α• LDHA
Ability to predict outcome to PTK/ZK
Gene Expression Medians and Range
CONFIRM1 CONFIRM2
FOLFOX4+PTK/ZK FOLFOX4+Placebo FOLFOX4+PTK/ZK FOLFOX4+Placebo
Gene* N Median (range) N Median (range) N Median (range) N Median (range)
LDHA 42 1.11 (0.07-2.54) 41 1.09 (0.14-3.63) 52 0.91 (0.23-3.46) 50 1.04 (0.20-2.66)
Glut1 42 2.11 (0.26-23.86) 41 2.49 (0.36-38.59) 51 2.39 (0.45-13.37) 50 3.45 (0.50-35.48)
HIF1a 42 1.62 (0.59-3.68) 41 1.62 (0.31-4.98) 52 1.43 (0.60-6.45) 50 1.43 (0.55-3.65)
VEGF 42 6.21 (3.05-21.90) 41 5.79 (1.29-15.90) 52 5.71 (2.32-35.33) 50 7.57 (2.54-65.06)
VEGFR1 41 6.28 (0.49-23.25) 36 5.59 (0.56-28.06) 48 5.33 (1.53-20.62) 40 6.15 (2.54-17.83)
VEGFR2 42 1.91 (0.02-8.16) 41 1.64 (0.62-11.63) 50 2.11 (0.76-10.46) 50 1.95 (0.49-14.48)
No significant differences between trials or treatment groups; all p-values >0.05
Response to PTK/ZK
>LDHA 0.033
>Glut1 0.045
>VEGFR1 0.012
<HIF-1α 0.021
CONFIRM 1
Gene p-value Gene p-value
CONFIRM 2
Response to PTK/ZK
0%
10%
20%
30%
40%
50%
60%
70%
80%
<0.36 >0.36 <1.50 >1.50 <3.78 >3.78
Response Rate
N
Glut 1 VEGFR1 LDHA
6 35 11 30 10 30
P = 0.033 P = 0.045 P = 0.012
CONFIRM 1
Response to PTK/ZK
CONFIRM2
0%
10%
20%
30%
40%
50%
60%
<1.18 > 1.18
Response Rate
NHif1a
19 32
P = 0.021
Response to PTK/ZK
VEGFR1(n=42)
<3.85
Group 1
(10%)
≥ 3.85
10
1Group 3
(61%)32
20
Confirm 1
Hif1a(n=51)
≥1.21 <1.21
Group 2
(53%)19
10Group 1
(13%)32
4
Confirm 2
Response(n=93)
Multivariate Analysis:
- Serum LDH
- Age
- Gender
- Performance Status
PFS with PTK/ZK
>HIF-1α 9.4 v 3.5
>LDHA 11.3 v 7.6
>VEGFR2 8 v 4.1
CONFIRM 1
Gene Months Gene Months
CONFIRM 2
<HIF-1α 7.6 v 2.7
<LDHA 7.6 v 1.7
0.075
0.021 0.021
0.031
0.001
P Value
P Value
P Value
P Value
P Value
VEGFR2 CONFIRM1
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0 5 10 15 20 25 30 35 40
Months since randomization
Est
ima
ted
P
rob
ab
ility
o
f P
rog
ress
ion
-Fre
e S
urv
iva
lP for interaction between treatment and
VEGFR2 expression = 0.001
VEGFR2 <2.98 (n=34)
with PTK/ZK
VEGFR2 > 2.98 (n=8)with PTK/ZK
VEGFR2 <2.98 (n=34)w/o PTK/ZK
CONFIRM 1
VEGFR2 > 2.98 (n=7) w/o PTK/ZK
Confirm 2
Hif1a(n=52)
<0.85
Glut1(n=42)
≥3.25
≥0.85
Group 3HR=1.25(n=10)
Group 5HR=7.96(n=16)
<3.25
Group 4HR=3.02(n=26)
PFS with PTK/ZK
PFS(n=95)
LDHA(n=43)
≥ 0.92
Group 1HR=1(n=28)
< 0.92
Group 2HR=1.94(n=49)
Confirm 1
Multivariate Analysis:
- Serum LDH
- Age
- Gender
- Performance Status
Overall Survival with PTK/ZK
<VEGFR2 0.012
CONFIRM 1
Gene p-value Gene p-value
CONFIRM 2
<Glut1 0.021
Confirm 2
Glut1(n=52)
< 3.28
Glut1(n=36)
≥ 2.12 < 2.12
≥ 3.28
Group 5HR=12.9(n=16)
Group 3HR=2.00(n=17)
Group 4HR=4.32(n=19)
OS in CONFIRM 1 and 2
Overall Survival(n=95)
VEGFR2(n=43)
Confirm 1
< 1.78
Group 1HR=1(n=17)
≥ 1.78
Group 2HR=2.49(n=26)
Multivariate Analysis:
- Serum LDH
- Age
- Gender
- Performance Status
VEGFR2 Predicts OS in CONFIRM1
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0 6 12 18 24 30 36 42 48
Months since randomization
Est
imat
ed
Pro
bab
ilit
y o
f S
urv
ival
Adjusted P value = 0.012
VEGFR2 <1.76 (n=38)
VEGFR2 >1.76 (n=45)
CONFIRM 1
VEGFR2 >1.76
35.8 mo v 20 mo
Conclusions
Genes involved in HIF-1α and angiogenesis pathway are
associated with response and PFS with VEGFR TKI
treatment.
These molecular markers were independent of serum
LDH.
Genes in this pathway (VEGFR2) may have prognostic
value for overall survival.
Provides an initial panel of markers which warrant further
testing in clinical trials with future VEGFR TKIs.
Acknowledgements
The patients and their families and Investigators who participated in CONFIRM 1 and
2.
Response Genetics: Kathleen D. Danenberg.
USC Statistics: Dongyun Yang.
Novartis / Bayer Schering AG CONFIRM team:
Michael M. Shi, Christian Jacques, J. Carl Barrett, Bee Chen and Gerold Meinhardt.
Funding: Dhont Foundation and Carpenter Lab