Slouching towards Jerusalem and the passionate intensity of - - PowerPoint PPT Presentation

Slouching towards Jerusalem and the passionate intensity of personalization

Biomarker and surrogate skepticism along the way.

Elevation in BP as a biomarker of risk and reduction of BP as a surrogate of risk modification

• Ignored as a biomarker of risk as HDL was

preferred as a biomarker of benefit – torcetrapib

• Ignored as a biomarker of risk as benefits

predicted to outweigh risk – bevacizumab.

• Hailed as a “mechanism” of risk which with

treatment would attenuate risk – celecoxib and rofecoxib.

FitzGerald GA TiPS 2007.

Inhibition of prostacyclin synthesis by celecoxib and rofecoxib

1PGI-M = 2,3-dinor-6-keto-PGF1α ; † P<0.01 vs Placebo; *P<0.05 vs Placebo.

Catella-Lawson et al.

• JPET. 1999;289:735.

Indomethacin 50 mg tid Placebo Rofecoxib 50 mg qd

† †

40 80 120 160

n=12 n=12 n=10

PGI-M ± SE (pg/mg Creatinine)

Placebo Celecoxib 400 mg Ibuprofen 800 mg

40 80 120 160

†

*

n=7 n=7 n=7

McAdam et al.

• PNAS. 1999;96:272

Placebo Celecoxib 200 mg Rofecoxib 25 mg

n=50 n=50 n=50

40 80 120 160

†

† Fries et al Gastroenterology 130:55-64, 2006

Cheng et al., Science 296:593, 2002

Prostacyclin modulates the Bioactivity of Thromboxane

1.5 1.0 0.5 0.0 wild type knock-out IP TP IPTP intima area / media area

* *

* p<0.05 vs. wild type

7.5 5.0 2.5 0.0 IP TP IPTP Urinary 2,3 dinor TxB2 (Fold over basal)

* *

wild type knock-out

Vascular COX-2 restrains Thrombogenesis Yu et al unpublished

# # # #

* *

#, p<0.01, paired t test, *, p<0.05, unpaired t test, n=7-14

Deletion of Vascular COX-2 elevates Blood Pressure

100 110 120 130 140

VSM KO EC KO VSM/EC KO Flox Ctr

Normal diet High salt diet Systolic BP (mmHg)

13 11 7 14 Yu Y. et al unpublished

Vascular COX-2 contributes to Urinary PGI-M

*, p<0.05; ***, p<0.001, n=16-22

Flox Ctr EC KO VSM KO VSM/EC KO 1 2 3 4

PGI-M ng/mg Creatinine

* * *** *

P=.0.057

R2=0.2313 *, P=0.0105

PGI-BP

100 110 120 130 140 150 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5

BP PGI-M

Inverse correlation between BP and PGI2 in EC and VSMC COX-2 KOs

Yi et al 2008

Magic Markers?

• Depressed PGIM predicted CV hazard and all

elements of the human phenotype recapitulated in mice by disruption of the COX-2 dependent formation of PGI2

• Manifestation of hazard relates to drug exposure,

concomitant therapies and underlying CV risk

• Other biological systems promote and restrain

hemostasis, hypertension, arrhythmogenesis etc

• How big is the variability problem?

Interindividual variability in the pharmacological response to COX-2 inhibition

100 200 300 400 500 600 700

Serum TxB2

4 SS 0 SS 4

hours COX-1 activity ex vivo

10 20 30 40 50

Serum PGE2

4 SS 0 SS 4

Single dose steady state hours

Single dose steady state

COX-2 activity ex vivo

0.2 0.5 1 2 5 10 20

Cox-2 selectivity

(COX-2 inhibition / COX-1 inhibition)

Fries, Grosser, FitzGerald, Gastroenterology, 2006

Placebo n=50 celecoxib (200 mg) n=50 rofecoxib (25 mg) n=50

Attained COX-2 selectivity

4 8 12 24 36 1000 2000 3000

Celecoxib [ng/ml]

4 8 12 24 36 1000 2000 3000

Time after last dose (hrs) Celecoxib [ng/ml] Time after last dose (hrs)

CYP2C9*3 +/+ CYP2C9*3 +/-

Genetic contribution to interindividual variability

Fries, Grosser, FitzGerald, Gastroenterology, 2006

The Challenge of Personalization

• The contribution of genomics to variability in

drug response was ~ 30% in healthy male volunteers under controlled conditions.

• How do we predict analgesic efficacy and

cardiovascular risk?

• How do we detect emerging risk?
• How do we confer benefit that is incremental

to clinical practice?

The Personalized NSAID Therapeutics Consortium

PENTACON

• Harness heterogeneous quantitative data

from diverse model systems exposed to comparator NSAIDs at multiple doses

• Utilize systems approaches, progressively

populated with human data to develop models that result in novel predictive hypotheses

• Test these hypotheses prospectively at scale.

Personalization of NSAID Therapy

The Time 100. http://www.time.com, 2008

Can we develop algorithms which, when populated with an individual’s pre- and post test drug exposure can predict (i) whether they should take and NSAID and (ii) if so which drug, at what dose and for how long?

Zebrafish Models Mouse Models Human Biology (I, II, III) Population Pharmacology Curation Logistics Molecular Profiling High Throughput Sequencing Biomedical Imaging Systems (I), Modeling and Computation (II) Database and Integration Data Visualization, Sharing and Education Administrative Core Steering Committee External Advisory Committee Healthcare providers, research community, and the public

Integrated Data Management

Systems Pharmacology

Experimentation Data Analysis and Interpretation Technology

The Personalized NSAID Therapeutics Consortium

www.pentaconhq.org

Heterogeneity of nonsteroidal antiinflammatory drugs

Prostacyclin modulates the cadiovascular response to thromboxane in vivo

Cheng et al Science 296: 539 – 541, 2002

Genetic contribution to interindividual variability (ii)

Celecoxib Rofecoxib

Fries, Grosser, FitzGerald, Gastroenterology, 2006

IT’S NOT A SIMPLE “BALANCE”

Confirmed Thrombotic Endpoint

Kaplan-Meier Estimates (95% CI)

RR(95% CI): 1.96 (1.20, 3.19)*

* p<0.05 Patients at Risk Placebo Rofecoxib 25 mg 1299 1192 1148 1079 1039 1002 470 1287 1123 1050 986 935 898 411 6 12 18 24 30 36 Month 2 4 6 8 Cumulative Incidence(%) with 95% CI

Placebo Rofecoxib 25mg

• Non-steroidal antiinflammatory drugs (NSAIDs) are the most commonly

used pain killers in the world. Approximately 1000 NSAID containing drugs exist.

• 46 million U.S. adults with arthritis use an NSAID regularly or

intermittently for pain control

• Some patients will have severe complications on NSAIDs:

all NSAIDs can cause: stomach ulcers, gastrointestinal bleeds, hypertension some NSAIDs can cause: heart attack and stroke

NSAIDs: Non-Steroidal Antiinflammatory Drugs