Post on 05-Jan-2016
transcript
The Biological Basis for the Cardiovascular Consequences of COX-2 Inhibition
Tilo Grosser
Institute for Translational Medicine and TherapeuticsUniversity of Pennsylvania
Membrane phospholipids
Differential Inhibition of COX isoforms
Grosser et al., JCI, 2006;116(1):4-15
Selectivity for COX-2 is a continuous, rather than a discrete variable
FitzGerald and Patrono NEJM 2002
COX-1
CO
X-2
COX-1 selectivity
COX-2 selectivity
A large fraction of human prostacyclin biosynthesis is COX-2 dependent
1PGI-M = 2,3-dinor-6-keto-PGF1 ; † P<0.01 vs Placebo; *P<0.05 vs Placebo.
Catella-Lawson et al.JPET. 1999;289:735.
Indomethacin50 mg tid
Placebo Rofecoxib50 mg qd
†
†
0
40
80
120
160
n=12 n=12 n=10
PG
I-M
1 ±
SE
(pg/
mg
Cre
atin
in)
Placebo Celecoxib 400 mg
Ibuprofen 800 mg
0
40
80
120
160
†
*
n=7 n=7 n=7
McAdam et al. PNAS. 1999;96:272
Placebo Celecoxib 200 mg
Rofecoxib25 mg
n=50 n=50 n=500
40
80
120
160
†
†
Fries, Grosser, FitzGerald, Gastroenterology 2006; 130:55
pg/mg creatinine pg/mg creatinine
2,3-Dinor-6-Keto-PGF12,3-Dinor-TXB2
Controls
Non-cardiac chestpain
Unstable angina
Myocardial infarction
Fitzgerald, D.J., NEJM 1986
Urinary Thromboxane and Prostacyclin metabolites are increased in acute coronary syndromes
[pg/ml] [pg/ml]
3000
2000
1000
0N = 4 6 16 14
3000
2000
1000
0N = 4 6 16 14
TxA2
COX-1
Thrombosis
Platelet activation and aggregation
COX-1 COX-2
PGI2
Thromboxane A2 amplifies platelet activation and recruits additional platelets to the site of clot formation
??
Cheng et al., Science 296:593, 2002
Prostacyclin modulates the bioactivity of Thromboxane
* p<0.05 vs. wild type
Common carotid artery injury
flexible wire
7.5
5.0
2.5
0.0IP TP IPTP
Urin
ary
2,3
dino
r T
xB2
(Fol
d ov
er b
asal
)
*
*
wild typeknock-out
COX-2 disruption and thrombosis
(12.5 25
0
100
200
300 WTCOX-2 KO
COX-2Y385F
Collagen(g)
* *Pla
tele
t co
un
ts (
x103
/ l)
**
Platelet depletion
(14)
(12)(10)
* *
0
20
40
60
80
100
TxA
2 a
go
nis
t-in
du
ced
th
rom
bo
sis
(p
erce
nt
mo
rtal
ity)
WT COX-2 KO COX-2Y385F
Mortality
Cheng Y, et al. J Clin Invest. 2006;116:1391-1399.
Prostacyclin modulates thrombosis dose dependently
+/+
P<0.01
P<0.05
Co
mp
lete
occ
lusi
on
tim
e (m
in)
+/- -/-
IP
0
20
40
60
80
100
120
COX-1 KD
vehicle DFU
P<0.05P<0.01
P<0.05
vehicle DFU
wildtype
Cheng Y, et al. J Clin Invest. 2006;116:1391-1399.
TxA2
ADP
Thrombin
Thrombosis
COX-2 derived prostacyclin acts as a constraint on all thrombotic stimuli
COX-2 inhibition augments thrombosis
• COX-2 disruption or inhibition augments the thrombotic response to various stimuli in vivo by depression of prostacyclin biosynthesis.
• All coxibs depress prostacyclin biosynthesis = class effect
• Suppression of COX-2 dependent prostacyclin does not cause spontaneous thrombosis, but augments the response to thrombogenic stimuli
• Hazard from coxibs particularly in those otherwise predisposed to thrombosis ?
Are there differences in hazard between distinct compounds?
- Selectivity for COX-2
- COX-2 unrelated effects (“off-target effects”)?
Differential recovery from steady-state inhibition of platelet COX-1 by low-dose Aspirin and Naproxen
Capone et al., Circulation. 2004; 109(12):1468-71
P<.01
Naproxen (N = 9)
Aspirin (N = 8)
P<.01P=.074
1 3 12 24
0
25
50
75
100
% T
hro
mb
oxa
ne
inh
ibit
ion
hours after last dose
Opposing roles of COX-2 and COX-1 products
in blood pressure regulation and atherogenesis
80
90
100
110
120
130
140
150
** **
Sy
sto
lic
BP
(m
mH
g)
WT
COX-2 KOCOX-2Y385F
WT/Celecoxib
##
COX-1 KD/Celecoxib
Cheng et al., J Clin Invest 116; 1391, 2006
* p<0.05 vs WT; ** p<0.01 vs WT; ## p<0.01 vs WT/Celecoxib
Kobayashi et al, JCI 114: 784-94, 2004
+/+
0
5
10
15P<0.05
P<0.01
% le
sio
n a
rea
TP -/- IP -/-
APOE -/-APOE -/-
Heterogeneity of response to traditional NSAIDs
Garcia Rodriguez et al 2005
• Opposing roles of COX-1 and -2 products in thrombosis, blood pressure regulation and atherogenesis
• Non-isoform selective NSAIDs (e.g. naproxen) may afford prolonged platelet inhibition throughout the dosing intervals in some individuals
• COX-2 critical in maintaining renal medullary blood flow under conditions of increased vasoconstrictor tone; Hypertension on NSAIDs relates to inhibition of COX-2
• Predisposition to atherosclerosis and hypertension attenuated by coincident inhibition of COX-1
• Differences in the degree of COX-2 selectivity are likely to affect cardiovascular risk
COX-2 selective NSAIDs vs isoform non-selective NSAIDs
Are there differences in hazard between distinct compounds?
- Selectivity for COX-2
- COX-2 unrelated effects (“off-target effects”)?
0.001 0.01 0.1 1 10
Plasma concentration (M)Celecoxib
0.001 0.01 0.1 1 10
Plasma concentration (M)Rofecoxib
COX-1
COX-2
COX-1
COX-2
Selection of the drug concentration in hASMC
hASMC
In vivo
Veh Cel Rof Veh Cel Rof Veh Cel Rof Veh Cel RofBas
-2 hrs 0 hrs 2 hrs 8 hrs 24 hrs
interleukin 1
All regulated Genes
(FDR <0.1)
relative expression level
low
high
High-thruput screen for off-target genomic effects
0 1 3 4Differentially regulated Genes (FDR <0.1)
3
Expression profiles of genes differentially regulated by rofecoxib and celecoxib
Affymetrix qRT-PCR Affymetrix qRT-PCR
0 2 8 24 hrs
2
3
4
*
2059
26_a
t (l
og 2
exp
ress
ion)
IL27RA
0 2 8 24 hr
0.0
2.5
5.0
RofecoxibControl
Celecoxib
IL27
R m
RN
A x
107/
18s
rRN
A
0 2 8 24 hrs
2.5
3.0
3.5
4.0
*
2049
46_s
_at
(log
2 e
xpre
ssio
n)
TOP3A
0 hr 2 hr 8 hr 24 hr
12
16
TOP3
A m
RN
A x
107/
18s
rRN
A
PTGIS
0 2 8 24 hrs
13
14
15 *
2081
31_s
_at
(log
2 e
xpre
ssio
n)
0 hr 2 hr 8 hr 24 hr
0
400
800
*
PTG
IS m
RN
A x
107/
18s
rRN
A
0 2 8 24 hrs
7.5
10.0
12.5
*
2187
29_a
t (lo
g 2
expr
essi
on)
LXN
0 hr 2 hr 8 hr 24 hr
10
20
LTX
mR
NA
x10
7/18
s rR
NA
*
0 2 8 24 hrs
2
3
4
*
2184
79_s
_at (
log
2 ex
pres
sion
)
XPO4
0 hr 2 hr 8 hr 24 hr2.5
5.0
7.5
10.0
XPO4
mRN
A x1
07/18
s rRN
A
0 2 8 24 hrs
5
10
15
*
2024
81_a
t (l
og 2
exp
ress
ion)
DHRS3
0 hr 2 hr 8 hr 24 hr
100
200
300
DH
R3
mR
NA
x10
7/18
s rR
NA
*
0 2 8 24 hrs
2.50
2.75
3.00
3.25
IL-1
*
2130
54_a
t (l
og 2
exp
ress
ion)
KIAA0841
0 hr 2 hr 8 hr 24 hr
1
2
3
4
5
6
7
KIA
A08
41 m
RN
A x
107/
18s
rRN
A
IL-1
0 2 8 24 hrs
10.0
12.5
15.0
*20
1010
_s_a
t (lo
g 2
expr
essi
on)
TXNIP
0 hr 2 hr 8 hr 24 hr
1000
2000
TXNI
P m
RNA
x107
/18s
rRNA *
0 2 8 24 hrs
1.75
2.25
2.75
*
2130
54_a
t (lo
g 2
expr
essi
on)
PIK3C2B
0 hr 2 hr 8 hr 24 hr
0.0
2.5
5.0
PIK
3C2B
mR
NA
x10
7/18
s rR
NA
GALNT12
0 2 8 24 hrs
7
8
9
10
IL-1
2188
85_s
_at (
log
2 ex
pres
sion
) *
0 hr 2 hr 8 hr 24 hr
5
10
15
20
25
IL-1
GA
LNT1
2 m
RN
A x
107/
18s
rRN
A
*
0 2 8 24 hrs
3
4
5
6*
2071
43_a
t (l
og 2
exp
ress
ion)
CDK6
0 hr 2 hr 8 hr 24 hr
100
200
CD
K6
mR
NA
x10
7/18
s rR
NA
Placebo Celecoxib Rofecoxib 200 mg 25 mg
post dose / pre dose ratio
low(decrease)
high (increase)
compounds
known: 92unknown: 137
median post / pre dose ration = 50 healthy subjects
Metabolomic plasma profiles of celecoxib, rofecoxib are more similar than any drug to placebo
Metabolon Inc.Robert Mohney Felice de Jong
Can individuals at risk for CV complications be identified early during treatment?
Towards an individualization of NSAID therapy
Interindividual variability in the pharmacological response to COX-2 inhibition
0.2
0.5
1
2
5
10
20
Co
x-2
sele
ctiv
ity
(CO
X-2
in
hib
itio
n /
CO
X-1
in
hib
itio
n)
Fries, Grosser, FitzGerald, Gastroenterology, 2006
Placebo
n=50
celecoxib (200 mg)
n=50
rofecoxib(25 mg)
n=50
Attained COX-2 selectivityAttained COX-2 selectivity
100
200
300
400500600700
Se
rum
Tx
B2
0 4 SS 0 SS 4 hours
COX-1 activity ex vivo
10
20
30
4050
Se
rum
PG
E2
0 4 SS 0 SS 4
Single dose steady statehours
Single dose steady state
COX-2 activity ex vivo
0 4 8 12 24 36
0
1000
2000
3000
Ce
lec
ox
ib [
ng
/ml]
0 4 8 12 24 36
0
1000
2000
3000
Time after last dose (hrs)
Ce
lec
ox
ib [
ng
/ml]
Time after last dose (hrs)
CYP2C9*3 +/+
CYP2C9*3 +/-
Genetic contribution to interindividual variability (i)
Fries, Grosser, FitzGerald, Gastroenterology, 2006
Cardiovascular complications from inhibition of COX-2 relate most plausibly to Inhibition of COX-2 derived prostacyclin formation.
Prostacyclin acts as a constraint on all thrombotic stimuli.
The cardiovascular hazard pertains to all coxibs and likely to traditional NSAIDs with high COX-2 selectivity.
Hazard would be expected to relate to - baseline cardiovascular risk- attained drug selectivity in vivo- dose and duration of exposure- interindividual differences in drug response.
High interindividual variability in the pharmacological response to coxibs.Genetic variability in PK and PD contributes to variance.
Garret A. FitzGeraldSusanne Fries Yan ChengYing YuDairong WangEmanuela Ricciotti