Post on 21-May-2015
Meet qPharmetra, LLC
a pharmacometric consulting company
SMB / Health Valley event Sept 25, 2014
“from molecule to business”
Lars Lindbom, PhDAnders Viberg, PhDKlas Petersson, PhD
Anja Henningson, PhDEva Hanze, MSc
Jacob Brogren, PhD
Klaas Prins, PhDMarita Prohn, MScJan Huisman, BEng
Kevin Dykstra, PhDLee Hodge, MBA
Eric Burroughs, MScJason Chittenden, MSc
qPharmetra LLC
• Founded in 2010 by 4 company owners: US (2), NL, SE
• 13 seasoned scientists with background in mainly pharmacy & engineering
• Serving ~25 innovative pharma companies (small biotech – large cap)
• Working as home- or office-based consultants
US-based, international, pharmacometric consulting company
"from Molecule to Business" 25 Sept 2014
Pharmacometrics
Pharmacometrics
Branch of science concerned with mathematical models of biology, pharmacology, disease, and physiology used to describe and quantify interactions between xenobiotics and patients, including beneficial effects and side effects resultant from such interfaces.
Analogy: think of it as the pharmaceutical version of econometrics
Pharmacometricians quantify in silico any measured biological relationship arising from administering drugs to humans (and animal species)
Note: QSAR – quantitative structure activity relationships could fall under pharmacometrics, but as it comes often before study in any animal species, leave humans, it is considered a separate field.
What is that?
"from Molecule to Business" 25 Sept 2014
Pharmacometrics
Pharmacokinetics (PK)
What the body does to the drug
Pharmacodynamics (PD)
What the drug does to the body
Population pharmacokinetics (popPK)
The study of the sources and correlates of variability in drug concentrations among individuals who are the target patient population receiving clinically relevant doses of a drug of interest .
Population pharmacokinetics (popPK-PD)
The study of the sources and correlates of variability in drug exposure –response relationships among individuals who are the target patient population receiving clinically relevant doses of a drug of interest .
Further General Concepts
"from Molecule to Business" 25 Sept 2014
What’s so special about pharmacometricians?
"from Molecule to Business" 25 Sept 2014
these nerds talk the language of the statistician …
"from Molecule to Business" 25 Sept 2014
… and that of the MD …
"from Molecule to Business" 25 Sept 2014
Shared interests, different language
Different means to the same end
"from Molecule to Business" 25 Sept 2014
Gauss
Our expertise needs to be pretty broad
Data Manager
Preclinical pharmacologist
Statistician
Clinical pharmacologist
Formulation expert
Member of Data Monitoring Board/Committee
Pharmacokineticist
Disease Expert
Development Team member / lead
Etc…
"from Molecule to Business" 25 Sept 2014
Without being The Expert in one field we have sufficient expertise in all
Pharmacometric analyses contributions
"from Molecule to Business" 25 Sept 2014
drug exposure effect filing market
across entire (pre) clinical drug development phase
What formulation? Plasma exposure?Drug Accumulation?Drug-drug interactions?Impact of renal impairment?…
Desired efficacy vs.Adverse events
Pharmacokinetic and pharmacometric sections mandatoryWhat the minimum effective dose?
Pharmacometric can aid line extensions Post –marketing clinical studies
We model the (measured) past to project out to the future
Patient C
Patient B
qPharmetra Services
"from Molecule to Business" 25 Sept 2014
We use integrated pharmacometric methods to help companies make the best drug development decisions
Decision
Mentoring / Partnering
StakeholdersManagement, External
Decision Makers, Project Team, other R&D
FunctionsEf
fica
cy
Patient A
Exp
osu
re
Time
Effi
cacy
TimeOur Drug’s Best Dose
Competitors
Clin
ical
Uti
lity
Dose
Tole
rab
ility
Dose
P(S
ucc
ess
)
Trial ScenarioA B C
ScenarioB
success(60%)
failure
ScenarioA
success(20%)
failure
Clinical Utility
Efficacy 1
Ease of Use
Tolerability
Efficacy 2
Big trial, slow to market
Small trial, fast to market
$$$
$$
$
Scenario B
$ Net Present Value
A
B
PopPK PK/PDMeta-
AnalysisClinical Utility
Decision Analysis
Virtual Trials
"from Molecule to Business" 25 Sept 2014
Case Study
Predicting Survival as a function of Tumor Growth Inhibition in Oncology
The oncology model framework
"from Molecule to Business" 25 Sept 2014
Client question: what dose do I need to take forward into the next trial?
Dose Exposure PFS models and simulations
PK Model
Exp
osu
re
Time
Tumor Growth Model
Tum
or
Exposure
Survival Model
Time
Surv
ival
𝑆𝑡,𝐷𝑜𝑠𝑒 = 𝑓 𝑇𝐺𝐼𝑡,𝐷𝑜𝑠𝑒
𝑇𝐺𝐼𝑡,𝐷𝑜𝑠𝑒 = 𝑓 𝐶𝑡
𝐶𝑡 = 𝑓 𝑡, 𝑋
Pharmacodynamics
Pharmacokinetics
Tumor Growth Inhibition after Novanib administration
"from Molecule to Business" 25 Sept 2014
Integrating individualized exposure as driver of tumor shrinkage
model:
Mean+/- 95%CI and mean model prediction
𝑑𝐴1𝑑𝑡 = 𝐾𝐿 ∙ 𝐴1 − 𝐾𝐷 ∙ 𝑒−𝜆𝑡 ∙
𝐶𝑠𝑠
𝐶𝑠𝑠
∙ 𝐴1
Tumor1
KLKD∙e-λt∙exposure2
43
1
1 2
3 4We established a significant
relationship between exposure and tumor shrinkage
Progression-Free Survival Advantage vs. Exposure
"from Molecule to Business" 25 Sept 2014
Increased exposure to drug increases probability to survive
Increasing drug exposure in plasma
Concentrationquartiles
Among novanib patients, there is a clear exposure-response relationship with PFS
Trend with increasing AUCSS, with q4 clearly superior to q1
coxph(formula = Surv(time = pfs, event = cens) ~
aucSS.q4, data = pfsData)
coef exp(coef) se(coef) z Pr(>|z|)
aucSS.q4(1.56,2.06] -0.3431 0.7096 0.2217 -1.548 0.121741
aucSS.q4(2.06,2.69] -0.4061 0.6662 0.2175 -1.867 0.061841 .
aucSS.q4(2.69,6.51] -0.7894 0.4541 0.2397 -3.294 0.000988 ***
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
Treating AUCSS as continuous:coxph(formula = Surv(time = pfs, event = cens) ~
aucSS, data = pfsData)
coef exp(coef) se(coef) z Pr(>|z|)
aucSS -0.0003256 0.9996744 0.0001101 -2.958 0.00309 **
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
Similar relationship with Cavg
"from Molecule to Business" 25 Sept 2014
Furthermore, predicted tumor shrinkage is a predictor of PFS
"from Molecule to Business" 25 Sept 2014
Increased exposure leads to tumor shrinkage which increases Pr(survival)
0 20 40 60 80 100
0.0
0.2
0.4
0.6
0.8
1.0
Progression Free Survival by Quartiles of Predicted Tumor Inhibition
Time Since First Dose (w)
Fra
ctio
n o
f P
atie
nts
with
PF
S
TGI,cfb Q4
TGI,cfb Q3
TGI,cfb Q2
TGI,cfb Q1
Prediction of PFS as a function of novanib-induced TGI
"from Molecule to Business" 25 Sept 2014
Using the model to predict different scenarios – an example: doubling the dose
20 mg
10 mg
0 20 40 60 80 100
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Time (d)F
ractio
n P
atie
nts
Su
rviv
ing
Tumor Shrinkage (% CFB)
tixladone 10 mg
tixladone 20 mg
-80 -60 -40 -20 0 20
novanib 20 mg
novanib 10 mg
95% CI
Prediction of PFS as a function of Novanib-induced TGI
"from Molecule to Business" 25 Sept 2014
Zooming in on 1 year survival cut the deal for taking 20 mg into phase III
tixlatinib 10 mg
Fraction Patients Surviving
De
nsity
0.3 0.4 0.5 0.6 0.7 0.8 0.9
01
23
45
6
tixlatinib 20 mg
Fraction Patients Surviving
De
nsity
0.3 0.4 0.5 0.6 0.7 0.8 0.9
02
46
8
Vertical line indicates standard of care (SOC) 1-yr survival
The model allowed to evaluation of different dose levels and regimens in in-silico
Conclusion: phase III dose (10 mg) might have been too low for optimal efficacy.
tixlatinib 10 mg
Fraction Patients Surviving
De
nsity
0.3 0.4 0.5 0.6 0.7 0.8 0.9
01
23
45
6
tixlatinib 20 mg
Fraction Patients Surviving
De
nsity
0.3 0.4 0.5 0.6 0.7 0.8 0.9
02
46
8
novanib 10 mg
novanib 20 mg
SoC
SoC
We recommend to study 20 mg vs SoC in the next trial
(note: a separate adverse event analysis that was an integral part of this recommendation supported this)
How do we turn our work into business
• In a landscape of other providers branding is essential
• Give clients a reason to go to you specifically
• For qPharmetra this branding theme is reproducible quality
• We believe that delivery of top-end quality products has led and will lead to repeat and new business
• How? SOPs, Automation, QC & QA on products delivered
• Our market is global with many companies US-based
• Flyering in central Nijmegen not helpful
• In EU: UK, Germany, Switzerland
• The NL – Germany area is increasingly vibrant
• Here Novio Tech Campus / SMB could play a role for us
"from Molecule to Business" 25 Sept 2014
“wie goed doet, goed ontmoet”
"from Molecule to Business" 25 Sept 2014
In Novio Tech Campus through SMB since Sept 1st 2014
Thank you !
Data Exploration
"from Molecule to Business" 25 Sept 2014
Challenge: Graphically explore data, uncovering the interrelationships between variables and covariates.
The qP Solution
With standardized datasets in hand, we are able to efficiently construct attractive and informative graphics of endpoints vs. exposure and other covariates. Having a standardized toolbox of graphing programs available means we can spend more time in the creative aspects of exploring these visualizations for insights.
Analysis-ReadyPAT DOSE TIME OBS AGE SEX
Effi
cacy
Tole
rab
ility
Dose
Effi
cacy
Covariate
Model-Building
"from Molecule to Business" 25 Sept 2014
Challenge: Develop mathematical framework quantifying the strength of and uncertainty in the relationships among endpoints and covariates
The qP Solution
For model-building, we don’t always have a standard, one-size fits all solution. Using our experience, we often define a structural model that describes the relationships among key variables and gives an appropriate distribution of random effects. We work to find models that are adequate for the task at hand, mechanistically appropriate, and capable of producing robust predictions.
Effi
cacy
Tole
rab
ility
Dose
Effi
cacy
Covariate
Effi
cacy
Tole
rab
ility
Dose
Ob
serv
atio
n
Prediction