Transcript Cavemen and Rocket Scientists: A Perspective on Pancreatic

Cavemen and Rocket Scientists:
A Perspective on Pancreatic
Cancer Researchers
Robert A. Wolff, M.D.
University of Texas, M.D. Anderson Cancer Center
ASCO Poster Discussion
Non-Colorectal GI Cancer Session
Sunday, May 31st, 2009
DISCLOSURE
I receive research funding
from:
Eli Lilly, and Company
Agenda
● Briefly discuss positive key elements of
these posters which support the notion of
rocket science
● Provide constructive criticism of our
collective Neandertal traits
● Attempt to conclude (without being shot!)
Table of Contents-1
● “Smart cytotoxics”
● Nanoparticle albumin-bound (nab) paclitaxel
(Von Hoff # 4525)
● Cationic liposomal paclitaxel (Loehr # 4526)
● A “clean” study regarding the role of radiation
in adjuvant therapy in resected pancreatic ca
● EORTC/FDCC/GERCOR: (Van Laethem # 4527)
● Single institution attempts to improve drug
development and therapy in pancreatic cancer
● Hidalgo # 4528 and Laheru # 4529
Table of Contents-2
● “Cooperative” group attempts to leverage
collected data from completed trials
● ECOG: DNA Repair Pathway SNPs (Yoon # 4530)
● CALGB: GWAS and a prognostic SNP (Innocenti # 4531)
● AViTA: Biomarker predicting rash to EGF-R inhibition and
rash in response to erlotinib as prognostic in patients with
metastatic pancreatic cancer (Verslype # 4532)
“Smart” Cytotoxics:
● nab-paclitaxel yields high response in SPARC + tumor cells!
● SPARC+ by P antibody: 80% of patients responded, PFS 14 M
● SPARC- by P antibody: 36% patients still respond to gem + nab
paclitaxel (higher than gem alone)
● Impressive PFS/OS in a pure metastatic population (especially those
with > 50% drop in CA 19-9)
● Phase III trial planned Gem +/- nab-paclitaxel
● Suggests SPARC P-antibody as a predictive biomarker
● Cationic liposomal paclitaxel and microenvironment.
● Randomized Phase II trial in locally advanced/metastatic pancreatic
cancer patients
● Median Overall Survival = 8.4-9.4 Months for Gem + Liposomal
Paclitaxel (dose dependent)
● Median Overall Survival = 7.2 Months for Gem alone
● Appears to be first “successful” attempt to target tumor angiogenesis
in pancreatic cancer.
Gemcitabine Dosing
● Gemcitabine is a minimally effective when dosed at 1000
mg/m2 over 30 minutes.
● In phase I, gemcitabine active at 180-525 mg/m2 over 30
minutes given weekly. No increase in intracellular levels of
gem-triphosphate were observed using higher doses.1
● 2 randomized trials demonstrate fixed dose rate
gemcitabine at or near MTD is better, but more toxic than
standard dose gemcitabine.2,3
● Individualized maximal repeatable doses of gem range of
from 300-700 mg/m2 weekly, closer to FDR gem.
● In combination with radiation, 400 mg/m2 x 7 weeks, led to
a median survival of 22 months in 86 patients with
resectable PC, 25% NOT RESECTED!5
1. Abbruzzese JL et al JCO, 1991
2. Tempero JCO, 2003
3. Poplin E, et al ASCO, 2006
4. Takahashi Y et al Pancreas, 2005
5. Evans DB, et al JCO, 2008
OR
?
● Developing sophisticated cytotoxics with
more specific targets in tumor cells or
microenvironment. (Laser guidance).
● Combining them with a club.
● We must be more intelligent about the
delivery of ALL our therapeutics to maximize
efficacy and minimize toxicity.
● Do you give IFL or FOLFIRI to your
patients?
Adjuvant Therapy:
EORTC/FDCC/GERCOR
● Clean study comparing 4 cycles of
gemcitabine alone vs gemcitabine for 2
cycles followed by gemcitabine/XRT.
Median Overall Survival
24 months
OR
?
● Adjuvant therapy requires patients to have
surgery first (a blunt weapon in pancreatic
cancer). No efforts to standardize this.
SURGERY
Adjuvant Rx
GITSG
Criteria for Resectability-No
EORTC
Standardized Surgery-No
Strict Margin Assessments-No ESPAC-1
RTOG 9704
CONKO 001
EORTC/FDCC
ESPAC-3
OR
?
● Adjuvant therapy requires patients to have
surgery first (a blunt weapon in pancreatic
cancer). No efforts to standardize this.
SURGERY
Criteria for Resectability-No
Standardized Surgery-No
Strict Margin Assessments-No
Adjuvant Rx
21.0
22.8
17.1
24.0
20.1
23.6
20.5
OR
?
● Adjuvant therapy requires patients to have
surgery first (a blunt weapon in pancreatic
cancer). No efforts to standardize this.
SURGERY
Criteria for Resectability-No
Standardized Surgery-No
Strict Margin Assessments-No
Adjuvant Rx
Survival Data Applies
only to the Numerator
Metastatic at
Restaging
12
T4 + Metastatic +
R1 + Islet Cell
558
Upfront
Surgery
1030
R0
Resections
472
The The Numerator
Eligible
Adjuvant
274
Therapy
58%
The Fit
Median
286
Survival
25.2 M
Inadequate
Recovery
Died
180
6
The Denominator at
the Mayo Clinic
1975-2005
Corsini M et al. JCO 2008
Metastatic at
Restaging
17
T4 or
Metastatic
16
Upfront
Surgery
889
R0 or R1
Resections
873
The The Numerator
Eligible
Adjuvant
468
Therapy
53%
The Fit
Median
485
Survival
21.2 M
Did not receive
adjuvant rx
Death
345
43
The Denominator at
the Johns Hopkins
1993-2005
Herman J et al. JCO 2008
Metastatic at
Restaging?
The
Fit
T4 or
Metastatic at
Surgery?
R0
Resections
The The Numerator
Eligible
Adjuvant
Therapy
90 patients
Median
Survival 24
months
Did not receive
adjuvant rx?
Died?
Upfront
Surgery
The Denominator at
EORTC Sites?
OR
?
● We need to select patients for surgery and for
adjuvant therapy.
● Insist up high-quality imaging.
● Use radiographic criteria for resectability.
● Standardize the operation (akin to TME).
● Strict margin assessments and restaging after
surgery.
Johns Hopkins and
Pancreatic Cancer Research:
Mice as Guinea Pigs
Important effort to truly personalize therapy.
Failure to engraft has prognostic significance.
?
Tumors grown in mice mimic sensitivity and
resistance to gemcitabine in the patients.
Very appealing model for evaluating cytotoxic
therapy.
Xenografts do not recapitulate the tumor
microenvironment.
Is it practical to perform tumor xenografts in
patients with advanced disease?
Johns Hopkins and
Pancreatic Cancer Research:
People as Guinea Pigs
Use small numbers of patients to maximize
information about specific therapies.
PK data regarding the cytotoxic (alone/with FTS)
and the molecular agent (alone/with cytotoxic).
Pre- and post-biopsy assessments of effects of
drug on the putative target (K-RAS).
?
Is a decrease in K Ras levels cause by FTS of
only 44% considered sufficient to alter biology?
Where are we going with FTS?
Exploiting ECOG 1201:
SNPs in DNA Repair Pathways predicting
Pathologic Complete Responses
Evaluated samples from patients with adenocarcinoma of
esophagus or GE junction enrolled on a prospective clinical
trial: 60 of 97 enrolled patients had adequate tissue for
analysis.
Exploiting ECOG 1201:
SNPs in DNA Repair Pathways predicting
Pathologic Complete Responses
Builds on previous retrospective analysis of
DNA repair gene SNPs in patients with
squamous cell or adenocarcinoma of E/GEJ1
Tissue analysis from a prospective randomized
trial in patients with adenocarcinoma.
XRCC1 Arg399Gln AA/AG vs GG had a 5-fold
increased risk of failing to achieve a path CR.
?
Prognostic, but not predictive. How do we use
this in the future? Exclude these patients?
1. Wu X, et al JCO, 2006
Leveraging CALGB 80803:
Gem/Placebo vs Gem/Bevacizumab:
Genome-wide Association Study (GWAS)
CALGB 80803 study of gemcitabine + placebo
versus gemcitabine + bevacizumab (no
difference in PFS or OS)
Robust GWAS of samples from 338 patients
testing >500,000 SNPs
Discovered a “bad” SNP IL17F H161R: 3.9%
frequency and associated with poor prognosis
IL17F H161R associated with up-regulation in
angiogenesis pathways and aggressive biology
?
Why no benefit with gem/bevacizumab in “good”
SNPs of IL17F?
Leveraging CALGB 80803:
Gem/Placebo vs Gem/Bevacizumab:
CALGB 80803 Association
study of gemcitabine
+ placebo
Genome-wide
Study
(GWAS)
This needs to be validated, but it appears
versus80803
gemcitabine
+ bevacizumab
CALGB
study of
gemcitabine +(no
placebo
in
PFS
or
OS)
todifference
be
prognostic,
not
predictive(noof
versus gemcitabine + bevacizumab
Robust
GWAS
samples
difference
in of
PFS
or OS)from 338 patients
response.
testingGWAS
>500,000
SNPs from 338 patients
Robust
of samples
a “bad”
SNP
IL17F
H161R:
3.9%
testing
>500,000
SNPs
WeDiscovered
are
smarter
now,
but
are
we wiser?
frequency aand
associated
withH161R:
poor prognosis
Discovered
“bad”
SNP IL17F
3.9%
Isfrequency
“Publish
orassociated
perish”
the
modern
IL17F
H161Rand
associated
with
up-regulation
in
with
poor prognosis
angiogenesis
pathways
and
aggressive
equivalent
of “fightwith
or up-regulation
flight”? biology
IL17F
H161R associated
in
angiogenesis pathways and aggressive biology
?
Why no benefit with gem/bevacizumab in “good”
SNPs of IL17F?
AViTA Trial: Rash as a Marker of
Efficacy to Erlotinib?
Previously
untreated
metastatic
pancreatic cancer
R
A
N
D
O
M
I
Z
A
T
I
O
N
GE-B
(n=306)
PD
GE-P
(n=301)
PD
1:1
G: 1,000mg/m2 on days 1, 8, 15, 22, 29, 36, 43 for first 8 weeks, days 1, 8, 15 in subsequent
4-week cycles; E 100mg/day; B 5mg/kg q2w
AViTA Trial: Rash is a Marker of
Good Prognosis!
Median Overall Survival
Treatment
No Rash
Any Rash
Gem/Erlotinib +
Placebo
Gem/Erlotinib +
Bevacizumab
4.3 M
8.1 M
5.0 M
7.9 M
All Patients
4.8 M
8.0 M
AViTA Trial: Rash is a Marker of
Good Prognosis!
Confirms results from PA.3 (Gem +/- Erlotinib).
This is an important observation
Multivariate analysis shows a positive
correlation between low CRP levels, no active
smoking, and increased chance of a rash
?
If lack of rash is bad, we can get patients to stop
erlotinib, or alternatively push up the dose of
erlotinib in an effort to generate a rash?
No evidence that development of rash is
predictive of erlotinib efficacy.
Retrospectively prognostic, not predictive!
In Conclusion
The brain trust in cancer research is getting
larger and more sophisticated.
Our researchers are doing A LOT of smart
things!
Not everything we do is smart and this impedes
our progress.
We owe it to our patients to do better!