Transcript pptx

Hypoxia-activated
prodrugs for cancer:
conception to clinic
Bill Denny
Auckland Cancer Society Research Centre
• Cancer is primarily a genetic disease; mutations in
our DNA result in mutations in key proteins that
control cell division
• Every cancer’s genome has different combinations of
DNA mutations, and thus a different suite of altered
proteins
Signalling networks in a cell
But at the level
of tissue
physiology,
virtually every
cancer is
different from
all normal
tissue – it’s
hypoxic.
Can we target
this?
Hypoxia in solid tumours
•
Functioning (cancer) cells need oxygen, so tumours need to
develop their own blood supply to grow beyond a few
millimetres in diameter
•
•
•
But they are very bad at it
Thus tumour cells remote from blood vessels are hypoxic
Long since recognised as a limitation of radiotherapy
Necrotic
Hypoxic
Aerobic
O2
Normal
Tumour
In 1986 we had an idea……………
Journal of Medicinal Chemistry
©Copyright 1986 by the American Chemical Society
June 1986
Volume 29, Number 6
Perspective
Considerations for the Design of Nitrophenyl Mustards as Agents with Selective
Toxicity for Hypoxic Tumor Cells
William A. Denny* and William R. Wilson*
Cancer Research Laboratory and Section of Oncology, Department of Pathology, University of Auckland, School of Medicine,
Private Bag, Auckland, New Zealand. Received. September 9, 1985
Windansea
Beach, La
Jolla, CA
Bill Wilson
Prodrugs with oxygen sensors
•
•
A prodrug is defined as an inactive (non-toxic) compound,
converted in the body to an active drug
Activate by electron addition thru cellular reductase enzymes.
Hypoxia
prodrug
design
Hypoxia
prodrug
mechanism
Importance of prodrug diffusion and
bystander effect (cartoon)
Non-toxic
prodrug
Activated
drug
Blood vessels
Necrosis
Hypoxic regions
Cross-section through a human head-and-neck tumour
What oxygen sensor?
Oxygen sensor needs to:
• be rapidly reduced by cellular reductases to a 1-electron adduct
• adduct needs to be stable enough so that in oxygenated
(normal) cells if can be reoxidised to the original prodrug
• adduct needs to undergo further changes in hypoxic cells the
that activate the toxin
NITROBENZENES
What toxin?
Toxin needs to:
•
•
•
•
be deactivated/activated by the oxygen sensor
have a well-understood mechanism
be a potent and ubiquitous cell killer (kill cells in a
variety of proliferative states and pH environments)
have good bystander effects
NITROGEN MUSTARDS
8
Nitroaniline mustards as hypoxia prodrugs?
•
•
Potency of “mustards” depends almost entirely on the
electron density at the mustard nitrogen
Nitro to amino change gives a large increase in electron
density, resulting in an 10,000-fold increase in potency
towards cells in culture
R
N
Cl
Cl
10,000fold
Easy; is this our lead compound?
1985
1993
2004
Optimising biological properties: diffusion
• 3D cell cell culture in a controlled atmosphere
• Physics and mathematical modelling of diffusion kinetics
• Can the prodrug reach remote hypoxic tumour cells?
• Need fast diffusion and slow metabolism
• Less water-soluble prodrugs diffuse more rapidly
100
Surviving Fraction
10-1
10-2
10-3
10-4
TPZ
10-5
10-6
10-7
10-8
5x increase
in diffusion coefficient
10-9
10-10
0.01
0.1
1
10
Oxygen concentration (µM)
A “pre-prodrug” concept to solve a dilemma
•
•
•
•
•
Dilemma: need lipophilic drug for good diffusion, but a watersoluble drug for iv injection.
Increase aqueous solubility by converting alcohol to phosphate
(PR-104)
Rapid non-specific cleavage of this by phosphatase enzymes
in serum to lipophilic alcohol PR-104A; rapid diffusion to
hypoxic regions
Activation by reductases to active amine PR-104M
This stable enough to have a good bystander effect
PR-104
PR-104A
PR-104M
Selection of lead candidate (2004)
• Selected
candidate PR104 from 6 final
compounds (allday discussion)
• Compromise
between activity,
potency, hypoxic
selectivity,
toxicity, ease of
synthesis (cost
of goods)
PR-104 does preferentially kill hypoxic
HT29
human
colon
cancer
cells
in model
tumours
HT29 human HT29
colonhuman
carcinoma
colon
xenografts
carcinoma
xenografts
Xenografts:
single
drug
dose
Single
prodrug
Single
dose
prodrug
at MTD
dose
at MTD
.999
99
•
99.9
99
90
Cell killing in human cervical
cancer xenografts in
immune-deficient mice
An initial radiation dose kills
all the oxygenated cells
(about 95% of the tumour)
•
PR-104 kills up to 3 logs
(99.9%) of the remaining
radiotherapy
radiotherapy
Radiotherapy
Radiotherapy
Radiotherapy
Radiotherapy
hypoxic tumour cells
+ PR-104
29
TP
Z
90
99.99
(20TP
GyZ
)
RASN
D 292
+T 4
PZ4
RA
DRA
+D
SN(
220
9G
24y
4)
RA
D
+T
PZ
RA
D
+S
N
29
24
4
9.9
99.999
24
4
9.99
Tumor cell kill (%)
•
(oxic cells)
D
RA
SN
(kills oxic (kills
cells)oxic plus
cells)PR-104
plus PR-104
(hypoxic
cells)
•
In vivo proof-of-concept a
major requirement
So, are we done…….?
To clinical trial: start-up
company Proacta Inc
•
•
•
•
Much of the 15 years of work (1986-2001), to this point was done
under grants from the HRC of NZ, CR UK, and the US NCI:
resulted in enough IP to set up a startup company
In 1998, with UK ICR, set up EPTCO; could not raise UK funding
In 2001, UniServices transferred assets to Proacta Therapeutics
NZ; still took 3 years to raise $US 12M from Aust, NZ, Switzerland
PR-104 first went to clinical trial in 2006 (Waikato, Auckland, US);
then raised another US $35M from US; Proacta Inc in San Diego
Trevor Twose
CEO 1998-2000
Aki von Roy
CEO 2000-2004
Paul Cossum
CEO 2004-2006
John Gutheil
CEO 2006-2012
PR-104: Discovery/development timeline
date
task
1986-2001 Development of concept and methods
discovery
2003
Nine advanced phosphate pre-prodrug
candidates
2001-2004 Set-up and funding of Proacta
development
2004
PR-104 selected by Proacta to develop
2005
GMP scale-up, drug formulation decided,
animal safety toxicology completed, IND
submitted to FDA and approved
Jan 2006
First patient treated at Waikato Hospital
May 2007
Phase I completed
2008
Phase II trial in non-small-cell lung cancer
But then, the unexpected (which you should always expect!).
Activity in Phase II in NSCLC and leukemia, but variable toxicity
Also see aerobic activation of PR-104 in
some human cancer cell xenografts
70
H460 NSCLC
Low % of
hypoxic cells
A1
Percentage hypoxic* cells
60
50
40
% of hypoxic cells in
various cancer
xenografts
30
E2
20
B3
10
0
H460
C33A
SiHa
HT29
22Rv1
A549 HCT116 A2780
250
In vitro PR-104A aerobic reduction
200
Extent of aerobic
metabolism of
PR-104
pmol PR-104H+M /106 cells
100
80
60
40
20
Control
C33A
A2780
HCT116
H1299
MDA231
MiaPaCa
PC3
22RV1
HT29
SiHa
H460
A549
0
H1299
A new aerobic nitroreductase (AKR1C3)
AKR1
cluster
• Microarray profiles of genes in 23 human tumour cell
lines show high aerobic toxicity to PR-104 correlates
with expression of AKR1C3, a member of the aldoketo reductase (AKR) gene family (not known as a
nitroreductase).
• Only AKR1C3 reduces PR-104 under aerobic
conditions.
• AKR1C3 reduces only PR-104 under aerobic
conditions (not even the class A and D series)
B
A 160
40
20
V5 TAG
NQO1
AKR1B10
AKR1B1
AKR1C3
AKR1C2
AKR1C1
WT
0
C33A
A2780
H1299
HCT116
MiaPaca 2
MDA231
8
22Rv1
PC3
60
10
SiHa
HT29
80
12
H460
100
14
HCT116 AKR1C3 #1
HCT116 AKR1C3 #3
A549
120
IC50 ratio (WT/AKR1C3)
16
HCT116 AKR1C3
PR-104M
PR-104H
140
No V5 induction
control
pmol PR-104A metabolites
formed per 106 cells
18
AKR1C3
6
Actin
4
2
0
4
9
e
9
A
C
N
in
le
le
zo azo -106 195 crin -104 Q4 cin myc EO
a
A
a
y
B
d
d
r
U C
t
ro
ni
ni
PR
Ni
tom rfi
RS
so tro
Mi Po
Mi Me
AKR1C3
18
C
Where now for PR-104?
• A Phase II trial in leukemia showed about
30% responses, but still unpredictable
aerobic activation so Proacta terminated
their development of it in 2012.
• Plans are underway for a new (noncommercial) trial by the US Southwest
Oncology Group in relapsed T-cell
lymphoblastic leukemia, which overexpresses AKR1C3 (exploiting this
weakness). We are supplying the clinically
formulated drug
Hypoxia-activated prodrugs that didn’t make it
Despite a compelling biochemical rationale and much work by many,
no hypoxia-activated prodrug has yet been FDA- or EMA-approved
Where now for hypoxia-activated prodrugs?
Apart from PR-104 we also have:
− Tarloxotinib, a completely
different class, now in
Phase II trials in NCSLC
in the US
− an A-series AKR1C3proof analogue of PR-104
about to be trialled in
Europe
− a lead candidate in a third
new class; looking for a
commercial partner
Conclusions
•
•
•
No hypoxia-activated prodrugs yet registered; a quixotic quest?
No, this is where academic involvement in drug discovery should be;
longer time horizon; can be more innovative (note half of all classes
of current cancer drug had their origins in academic groups)
This project could not have been started in a large pharma:
–
–
•
•
•
concept not “validated”, thus much too risky
PR-104 has nitro groups, a mesylate and a halide; all on the “toxic” list
of no-no substituents)
Drug development in a new field can take a long time; need to
develop analytical methods to measure (e.g.) drug diffusion through
tissue, reductive enzymology, hypoxia biomarkers
In hindsight, we made the wrong initial choice of sub-class to
develop with PR-104 (AKR1C3 susceptibility of class B) but could
not know that
Tarloxotinib may be the “breakthrough” compound in the field?
ASCRC staff
Commercial and
grant sources