52
Impact at a glance
Sekar, UppoorDrug-drug interactions in label
Risperidal
Sun, Doddapaneni
Gobburu,
Marroum
Fadiran et al
Reviewer(s)
Supportive evidencePain drug
Pediatric dosing
based on biomarkers
Sotalol
Risk assessment to support approval
Minoxidil
ImpactDrug
Impact at a glance
Booth, Rahman
Ramchandani, Booth, Rahman
Beasley, Marroum
Sekar, Duan, Uppoor
Reviewer(s)
Pediatric dosingBusulfan
Dose adjustment in renal impaired
Zometa
“Approvable” due to sub-optimal dosing
regimen
CCB
Monotherapy in pediatrics without controlled clinical
trials
Trileptal
ImpactDrug
Impact at a glance
Yasuda, Uppoor
Wang, Beasley, Marroum
Beasley, Bhattaram, Marroum
Zheng, Srikanth, Doddapaneni
Reviewer(s)
Confirmatory evidence
CNS drug
Cause of trials failure; alleviation of false +ve QT signal
PAH drug
supportive evidence; Contraindication of hepatic impaired
Ranolazine
Dose-response to support lower doses in labeling
Varenicline
ImpactDrug
Non-exclusive list
Acknowledgements
• All OCP reviewers, past and present, who contributed to the many pharmacometrics reviews and without whose excellent work I will not be here
– Special thanks to our Pharmacometrics team!
• OCPB Team Leaders who contributed to the survey and reviews
1992-2000 2005-20102000-2004
Evolution of Pharmacometrics at FDA: More IND/NDA Opportunities and Greater Demand from Medical
Decisions
Label
Disease
Models
Policy
Knowledge
Design
Impact of Pharmcometrics on Drug Approval and Labeling Decisions: A Survey of 42 NDAs
Venkatesh A. Bhattaram, Brian P. Booth, Roshni P. Ramchandani, B. Nhi Beasley,
Yaning Wang, Veneeta Tandon, John Z. Duan, Raman K. Baweja, Patrick J. Marroum,
Ramana S. Uppoor, Nam Atiqur Rahman, Chandrahas G. Sahajwalla, J. Robert Powell,
Mehul U. Mehta, Jogarao V. S. Gobburu
AAPS Journal. 2005; 7(3): Article 51. DOI: 10.1208/aapsj070351
FDA Pharmacometrics reviews pivotally impacted approval and labeling
0
10
20
30
40
50
60
% N
DA
s (
To
tal=
42)
Approval Labeling
Pivotal
Supportive
No Contribution
Case Study#1
FDA’s proactive model-based analysis
alleviated the need to conduct additional
clinical trial for the approval of Trileptal
monotherapy in pediatrics
Reviewers
Drs. Sekar, Duan, Uppoor, Gobburu
Regulatory Issue
FDA/Sponsor pursued approaches to best
utilize knowledge from the positive trials to
assess if monotherapy in pediatrics can
be approved without new controlled trials
Adjunctive Monotherapy
Adults Clinical trials Clinical trials
Children (4-16 years ofage)
Clinical trial “Model Based Bridging”approach proposed byFDA
Motivation for PM Analysis
• Monotherapy of anti-epileptics is
important
– Better safety, Ease of Rx mgmt
– Avoid unnecessary costs
• Monotherapy trials are challenging
• Reasonable ER knowledge available
– Integration of knowledge across trials and populations is needed
• Law supports model based thinking
Is the exposure in pediatrics predictable from that in adults?
Population PK analysis suggested that differences in
PK can be explained using body size
0
1
2
3
4
5
0 1 2 3
Body Surface Area, m2
CL
, L
/h
Simulated curve using
the final PPK model
Is the placebo effect in pediatrics comparable to that in adults?
• Suggests similar
response toadjunct therapyKS Test
p=0.609
Change in Seizure Frequency (%)
Is the exposure-response in pediatrics comparable to that in adults?
• Significant Cmin (trough) - seizure reduction relationships exist (adjunct therapy)
• Exposure-response for adults and pediatrics are reasonably similar
Population N β0 (s.e.) β1 (s.e.)
Adults 480 4.55 (0.04) -0.010 (0.0011)Peds 230 4.54 (0.06) -0.0072 (0.0015)
β0 placebo-effect
β1 Cmin-Seizure
reduction slope
Value of Pharmacometrics
• Modeling and simulation aided in utilizing all
previous data to justify approval without
additional controlled clinical trials
• Allowed selection of dosing guidelines in
pediatrics
• The presented approach has a greater global
impact
– Precedent was set
• Sponsor’s perspective
– Economic implications
Application of Quantitative Tools to Efficient Decision
Making: Where is FDA Going?
Joga Gobburu
Pharmacometrics
Office Clinical Pharmacology, Office of Translational
Science (OTS), CDER, FDA
High attrition rate even in late development
Kola I, Landis J.Can the pharmaceutical industry reduce attrition rates?
Nat.Rev.Drug.Disc. Aug 2004.
Pharmacometrics (or Quantitative Experimental Medicine?)
• Science that deals with quantifying disease
and pharmacology
– Single individual or diverse group?
• Clinical pharmacologists, Pharmacometricians, Clinicians, Statisticians, Bioengineers
Office of the CenterDirector
Office of ClinicalPharmacologyDr. Larry Lesko
Office of BiostatisticsDr. Bob O’Neill
Critical Path Initiatives
IntramuralResearch
RCC, RIHSC, RSR
Office of TranslationalSciences
Dr. Shirley Murphy
OCP Larry Lesko, Director
Office of Clinical Pharmacology
Pharmacometrics Staff
Bob Powell, Director
DCP1
Mehul Mehta, Director
DCP2
Hank Malinowski, Director (Acting)
DCP3
John Hunt, Director (Acting)
DCP4
John Lazor, Director
DCP5
Shiew-Mei Huang, Director (Acting)
Psychiatry
Ray Baweja, Team Leader
Neurology
Ramana Uppoor, Team Leader
Cardio-Renal
Patrick Marroum, Team Leader
Metab-Endo
Hae Young Ahn, Team Leader
Pulmon-Allergy
Tayo Fadiran, Team Leader
Anesth-Crit Care-Addiction-Rheum
Suresh Doddapaneni, Team Leader
Gastro
Dennis Bashaw, Team Leader
Derm-Dental
Dennis Bashaw, Team Leader
Repro-Uro
Ameeta Parekh, Team Leader
Spec Path
Phil Colangelo, Team Leader
Anti-inf
Venkat Jarugula, Team Leader
Oncology
Brian Booth, Team Leader
Med Imag-Hema
Young Moon Choi, Team Leader
Biologics Hong Zhao,
Team Leader (Acting)
Chandra Sahajwalla,
Associate Director for Operations
Shiew-Mei Huang,
Deputy Director for Science
Felix Frueh,
Associate Director for Pharmacogenomics
Pharmacometrics
Services Joga Gobburu, Team Leader
Pharmacometrics
Tools & Methods Peter Lee,
Team Leader
Antiviral
Kellie Reynolds, Team Leader
OCPB Pharmacometrics OrganizationOCPB Pharmacometrics Organization
Director
Bob Powell
Review Services Joga Gobburu
Tools & Methods Peter Lee
Software programmer(s)Atul Bhattaram
Yaning Wang
Christoffer Tornoe
Pravin Jadhav
Christine Garnett
Raj Madabushi
Bayesian Statistician
~10 more primary clinical pharmacology reviewers havePharmacometrics skills.
PKPD Expert
Pharmacometrics Mission• To improve the public health by increasing the efficiency
and quality of clinical drug development with the application of model-based drug development.
• To develop quantitative model based tools to improve key drug development decisions (e.g., trial strategy & design, regulatory drug & label approval).
• To train and develop scientists who will perform this work at the FDA and elsewhere.
• To work collaboratively across therapeutic areas and disciplines to accomplish this mission.
• To both create disease models predicting patient outcome that can be shared inside and outside the FDA and establishing disease data library that can be used inside and outside FDA to promote understanding the disease process and how to measure improvement or worsening.
Regulatory Expectations and Opportunities for Sponsors
• Continued NDA Pharmacometrics reviews
– Impact on approval/labeling decisions
• Advice on drug development strategy, trial
design
– Enhance FDA-Sponsor interactions
– Involve quantitative thinking early-on
• Disease models to optimize development plans
– Critical Path Initiatives
• Sponsors responding to FDA’s call
Advice on drug development strategy and trial design:
Prerequisites
• Leverage prior knowledge
– Disease models
• Efficient tools
• Build inter-disciplinary expertise
• Integrate pharmacometrics into decision
making
Surrogate
Rela
tive
Ris
kMorbidity#1
Morbidity#2
Mortality
PLACEBO/DISEASE MODEL
Su
rro
ga
te
TIMEDose
Exp
os
ure
Su
rro
ga
te
Exposure
To
xic
ity
DRUG MODEL
CLINICAL TRIAL MODEL
80 140 200
020
40
60
Black Female80 120 160
010
30
Other races Female
Patient Population
Baseline Body Weight TIME
% D
rop
-ou
t
%Adherence%
Pati
en
ts
DiseaseDrugTrial
Models
Core Development Strategy
DesignMoleculeScreening
Patient Population Dose Selection
Approval Criteria
Individualization
Value of Disease
DrugTrial
Models
Core Development Strategy for Testosterone Suppressants
Quantitative
analysis
Reporter
Gene Assay
Preclinical
Clinical Trial
Simulation
Dose
optimization
in cancer
patients
Pivotal trial
|----*2 mo-----|*Actual execution time.- it does account for time spent accumulating resources.
|----*2 mo-----||----*2 mo-----||----*3 mo-----||---------*12 mo--------------|
- Early screening of compounds based on IC50
value.
- High thr’putmethod to filter thousands of compounds
- Based on prior experience, a few potential entities will be selected for the next phase
IC50
PKPD data
- In vitro IC50 as a guide for preclinical dose selection
- Animal modelsto measure all possible biomarkers e.g. GnRH, LH, T and Drug conc.
- Invitro and preclinical data for clinical dose and regimen selection
- Clinical development plan
- Pilot study for dose optimization thr’ innovative trial designs
PKPD data
From Pravin Jadhav, VCU/FDA
Parkinson’s Disease ModelKey Questions
• How do we discern symptomatic vs
disease modifying benefit in clinical trials?
• What is an acceptable primary analysis for
approval?
– Influence of different drug effects, drop-out mechanisms, endpoints and analysis methods
Symptomatic or Protective?
10
12
14
16
18
20
0 6 12
Time, months
To
tal U
PD
RS
Placebo
Drug A
Drug B
Symptomatic or Protective?
10
12
14
16
18
20
22
24
26
28
0 6 12
Time, months
To
tal U
PD
RS
PlaceboSymptomaticProtectiveSymp+Prot
Parkinson’s Disease
Dr. Bhattaram and Siddiqui are the project leads:
FDAStatistics, Clinical, Policy Makers
ExternalStatistician, Disease experts
Parkinson’s Disease
Collect data
Trial#4
Trial#3
Trial#2
Trial#1
Data
9mo+follow-up200NDA
9mo+follow-up900NDA
1yr+follow-up400NIH
1yr+follow-up400NDA
Trial Duration#PatientsSource
Patient Population Model
Baseline UPDRS
Pa
tie
nts
, %
Observed
Simulated
PASS
DevelopDisease
DrugTrial
Models
Disease Progression
NEJM, 351, Vol 24, 2498-2508
Drop-out Model
Weeks
Dro
p o
uts
, % Patients with
faster progression
Patients with no orshallow progression
Change in UPDRS at 26 wks
Pa
tie
nts
, %
Observed
SimulatedPASS
Model Qualification
Discern Symptomatic vs. Protective Effects: Delayed Start Design
0
5
10
15
20
25
30
0 20 40 60
Weeks
UP
DR
S
If drug is protective then patients who received drug longer will have
lower scores compared those who receive drug late.
Placebo
Drug
Drug Protective
False Positive Rate
BOCF: Baseline observation carried forwardLOCF: Last observation carried forward
-MNAR
67.6565.1038.8029.1512.254.70MAR (Lack of benefit+Toxicity)
6.4522.5511.2511.507.554.95MAR (Lack of benefit; unequal drop-outs)
13.2024.7030.7022.6016.355.15MAR (Lack of benefit; equal drop-outs)
4.9518.2011.755.805.005.20MCAR
Baseline CarryForward
Group Max
Group Mean
LOCFAvailable cases
Active PhasePlacebo Phase
Dropout Scenario
Parkinson’s Disease Model
• Collected a large database of clinical trials
• Extracted patient population, placebo/disease progression, drug effect (not shown) and drop-out information.
• Simulations to answer the key questions mentioned earlier are in progress
• These findings will be published soon and technical details will be presented publicly soon.
BiomarkerBiomarker--Survival Relationship is Valuable Survival Relationship is Valuable
for Efficient Drug Developmentfor Efficient Drug Development
-100% -50% 0 50% 100%0.0 0.5 1.0 1.5 2.0
01
23
Re
lati
ve
Ris
k o
f R
en
al F
lare
Estimated RRLL of 95% CLUL of 95% CL
Re
lati
ve
ris
k o
f D
ea
th
Tumor Size “Change”
An inter-disciplinary team at FDA is developing a tumor
size-survival relationship for Non-small lung cancer
• Screening for drugs• Dose selection• Verification of endpoints
• FDA will be in a positionto guide sponsors earlyin drug development
Dartois C, Sung C, Wang Y,
Ramchandani R, Rock E, Booth B, Gobburu J.
Manage Knowledge
Knowledge
Placebo & Disease Models
Information
• Biomarker-Endpoint •Time course• Drop-out• Inclusion/Exclusioncriteria
• Parkinson’s• Obesity, Diabetes• Tumor-Survival
• Rheumatologic condition• HIV• Epilepsy• Pain• Osteoporosis
Increase Review Efficiency
Data submission standards - CDISC
Review Tools
Database
Time has come for
expecting more fromour software
Diversify Expertise
Clinician
PKPDExpertStatistician
Train Scientists
Increase awareness among peers
Train graduates/post-docs
PhRMA should support
•• FDA trains fellowsFDA trains fellows
•• ACCP PM websiteACCP PM website
•• AAPS FellowshipAAPS Fellowship
Integrate pharmacometrics into mainstream drug development
• Message from FDA is loud and clear
– Increased number of NDA/INDs with pharmacometrics analysis
– More ‘customers’ seek consults
• Pharmacometricians need to be part of
drug teams and build strategic
relationships
Increase Sponsor-FDA Interaction
EOP2A meetings provide an excellentopportunity for Sponsor and FDA toexchange science on a less formal basis
We encourage Sponsors to orient thereviewers to the ClinPharm portion ofthe NDA
• FDA Pharmacometrics in collaboration
with Industry is planning on two meetings
early 2007, focusing on:
– Disease models
– Pharmacometrics tools
• FDA Pharmacometrics to set up an
external website – discussions ongoing
Increase Sponsor-FDA Interaction
How Can Industry Help FDA Make Drug
Development Process and Regulatory Review More Successful?
• Collect PK, biomarker and clinical endpoint data over range of doses in late clinical trials
• Apply quantitative methods early and continuously during IND period
• Initiate increased communication with FDA during mid/late stages (e.g., EOP2A)
• Focus on adequate identification of optimal dosing regimens in late clinical studies
Advice on drug development strategy and trial design:
Prerequisites
• Leverage prior knowledge
– Disease models
• Efficient tools
• Build inter-disciplinary expertise
• Integrate pharmacometrics into decision
making
The best way to predict the future
is to create it.
Peter Drucker
