Transcript Pharmacometrics and Biostatistics Interactions at the FDA

PHARMACOMETRICS AND BIOSTATISTICS
INTERACTIONS AT THE FDA
Jeffry Florian, Ph.D.
Division of Pharmacometrics
OCP/OTS/CDER
U. S. Food and Drug Administration
Presented at ASA 2016 Biopharmaceutical Section Regulatory-Industry Statistics Workshop in
Washington, D.C. on September 29th, 2016
www.fda.gov
1
Disclaimer
• The opinions expressed in this presentation are the presenter’s
and do not necessarily reflect the official views of the United
States Food and Drug Administration (FDA).
www.fda.gov
2
Sometimes it may feel like….
Pharmacometrics
www.fda.gov
Statistics
3
A slightly different analogy?
• Similar interests and tools
– Understanding uncertainty and promoting public health
– Statistics, drug development, clinical trial design
4
Organizational Location
Office of Translational
Sciences
Vacant
www.fda.gov
5
Geographical Location
www.fda.gov
6
Divisions Within Each Office
Office of Clinical Pharmacology
Office of Biostatistics
• Immediate Office
• Immediate Office
• Div. of Clinical Pharmacology
I, II, III, IV, and V
• Division of Biometrics I, II,
• Division of Pharmacometrics
III, IV, and V
• Genomics and Targeted
Therapy Group
• Division of Biometrics VI
• Division of Applied
and VIII
Regulatory Sciences
http://www.fda.gov/AboutFDA/CentersOffices/OfficeofMedical
ProductsandTobacco/CDER/ucm106189.htm
www.fda.gov
http://www.fda.gov/AboutFDA/CentersOffices/OfficeofMedical
ProductsandTobacco/CDER/ucm166250.htm#Role
7
Offices Interact in All Activities
Pre-IND
Protocol
Reviews
- Dose finding
- Registration
NDA/BLA
Reviews
Regulatory
Research
QT reviews
Pediatrics
EOP1/2/2A
Pre-NDA/BLA
meetings
meetings
Efficacy
Supplements
Guidance
and
Policy
Labeling
PostMarketing
Biosimilars
8
Various Internal/External
Collaborations
• OCP-OB Collaborative Working Group (2012)
– Good practices for early and timely interactions
– Team building activities between Offices
– Periodic Office meetings for scientific exchange
(Multi-Disciplinary OB-OCP Scientific Exchange)
• External groups facilitating interactions
9
Example 1: Sofosbuvir (SOF, GS-7977)
• Prodrug of a nucleotide analog inhibitor of the hepatitis C virus NS5B
RNA-dependent RNA polymerase
• First-in-class submission (breakthrough designation)
– Broad genotypic activity
• Proposed indication: in combination with other agents for treatment
of chronic hepatitis C (CHC) in adults
• Sofosbuvir was studied in combination with RBV for GT 2 and 3, and
in combination with PEG/RBV for GT 1, 4, 5 and 6.
Example contributed by Karen Qi, Jeff Florian, Wen Zeng, and Dionne Price
10
10
Phase 3 Trial Design: GT 1, 4, 5, 6
Trial Name
Population
Regimen* and Duration
GS-US-334-0110
(NEUTRINO)
Treatment-Naïve
SOF+PEG/RBV 12 Weeks
0
GS-US-334-0110
NEUTRINO
GT1/4/5/6
Treatment-naïve
12
SOF + PEG/RBV
N=327
Weeks
24
36
SVR12
N=Number of subjects; PEG=Pegylated Interferon; RBV=Ribavirin
*SOF (400 mg/day) + PEG (180 g/week) + RBV (1000 or 1200 mg/day)
11
11
Evidence to Support Effectiveness of
SOF+PEG/RBV in Genotype 1 PEG/RBV
Treatment-experienced Subjects
Source
Evidence to support effectiveness of
SOF in PEG/RBV TE subjects
-
Treatment-Naïve, Phase • PEG/RBV TE subjects are represented
III trial (GS-US-334-0110)
within the treatment-naïve population
-
Indirect evidence from
previous PEG/RBV trials
• Confirmation of predictive baseline
factors and response rate in population
-
Indirect evidence from
previous DAA+PEG/RBV
trials
• Comparison with other approved
products suggest a higher response
rate
• Various bridging analyses conducted based on baseline factors,
relative risk, and odds ratio analyses
12
Outcome
• Analyses were presented and discussed at the
sofosbuvir advisory committee meeting
• Majority of the committee agreed with the review
team’s conclusion that this regimen would be no worse
than available treatment options for GT 1 PEG/RBV TE
subjects
• Sofosbuvir labeling reflects that SOF+PEG/RBV can
be used in GT 1 PEG/RBV TE subjects
• Multiple factor analysis is described in the label and
HCV treatment guidelines
• Joint publication of FDA’s analyses and rationale
13
Example 2: MCP-Mod
• “Of the unsuccessful first-time applications, 24 (15.9%) included
uncertainties related to dose selection”
• “Failure to determine the most appropriate dose for clinical use was a
major reason for nonapproval.”
Sacks et al, JAMA (2014)
14
Example contributed by Lei Nie, Dionne Price, Mohamed Alosh, Dinko Rekić, and Yaning Wang
MCP-Mod
• Analysis of dose-response studies has been divided into two
primary strategies
– multiple comparison procedures (MCP)
– model-based approaches (Mod)
• When applied separately, each strategy has shortcomings that may
impact the decision-making process.
• Submission by Janssen Pharmaceuticals/Novartis Pharmaceuticals
– intended to support the use of MCP-Mod as an efficient statistical
methodology (combines both strategies)
• Design stage
– plausible candidate models are selected
• Analysis stage
– assess the dose-response signal using MCP and select ‘best’ model
– Fit the selected model(s) to the data and estimate the target dose
15
MCP-Mod (cont.)
• Office of Biostatistics and Office of Clinical
Pharmacology, Division of Pharmacometrics jointly
reviewed submission
–
–
–
–
Assess materials and simulations provided by the sponsor
Additional sensitivity analyses identified by the reviewers
Identify advantages/disadvantages of MCP-Mod
Worked in close collaboration throughout review
• Review team: We conclude that MCP-Mod either has
similar power to or outperforms alternative
approaches and is fit-for-purpose under the defined
context of use
16
For the
Determination
Letter and
Discipline
Reviews
http://www.fda.gov/drugs/developmentapprovalprocess/ucm505485.htm
17
Example 3: Qualification of Total Kidney Volume (TKV) in
Autosomal Dominant Polycystic Kidney Disease (ADPKD)
Joint FDA-EMA submission from Polycystic Kidney Disease Outcomes Consortium (PKDOC)
PKDOC
approach
Example contributed by John Lawrence, DJ Marathe, James Hung, Sue-Jane Wang
18
TKV Biomarker Qualification submission
• Objective: Clinical trial enrichment in Autosomal Dominant Polycystic
Kidney Disease (ADPKD)
• Stage of Drug Development for Use: All clinical stages of ADPKD drug
development, including proof of concept, dose-ranging, and
confirmatory clinical trials.
• Proposed Context of Use: Baseline TKV can be applied as a prognostic
biomarker that, in combination with patient age, can be used to help
identify those ADPKD patients who are at the greatest risk of advancing
in the course of their disease to a point where there is substantial
decline in renal function as measured by clinically meaningful outcomes
(30% worsening of eGFR , 57% worsening of eGFR (equivalent to
doubling of serum creatinine), and ESRD).
eGFR: Estimated Glomerular Filtration rate; ESRD: End Stage Renal Disease
19
Qualification of TKV in ADPKD
• Biomarker Qualification Review Team (BQRT) conducted
additional analyses and performed model development
and cross validation
o Analyses limited to patients with an eGFR ≥25 and at least 12
years of age, which represent the population likely to be enrolled in
clinical trials (925 subjects with 300 events).
o Some subjects had imaging performed with more than one
modality. FDA reviewers selected MRI data as the first preference,
CT data as the second preference and ultrasound data as the last
preference
o BQRT also carried out an external/independent validation using a
separate internal dataset
20
20
Qualification of TKV in ADPKD
Conclusions:
• Substantial improvement on predictive performance of event risk (based on a
concordance measure for time-to-event data) of a fitted survival model
including log (TKV) as compared to not including log (TKV) based on the Cstatistics using either
– submitter’s registry data in model development (with cross validation)
– in a clinical trial data that is available internally as independent validation.
• Too few ESRD and 57% decline in eGFR events over the time frame of
a feasible clinical trial to perform meaningful analyses
Use Statement:
• TKV, measured at baseline, is qualified as a prognostic enrichment biomarker
– select patients with ADPKD at high risk for a progressive decline in renal function (defined as a
confirmed 30% decline in the patient’s eGFR) for inclusion in interventional clinical trials.
• May be used in combination with age and baseline eGFR for enrichment
21
21
For the
Executive
Summary and
Reviews
http://www.fda.gov/Drugs/DevelopmentApprovalProcess/DrugDevelopmentToolsQualificatio
nProgram/ucm458492.htm
22
Example 4: QT-IRT
• US FDA established the Interdisciplinary Review
Team for QT studies (June 2006)
- Provide reliable, consistent advice to FDA drug review
divisions
- Promote advances in design of QT studies
• Members include medical officer, clinical
pharmacologist, statistician, pharmacologist,
project manager, and data manager
• Responsible for reviewing protocols and study
reports related to QT assessment
• ~45 reports per year, ~70 protocols per year
Example contributed by numerous individuals throughout the years
23
QT-IRT Responsibilities
• Reviews protocols and study reports related to QT
assessment
• Ensures that sponsors and review divisions
consistently receive the best available advice on
these studies
• Participates in internal and sponsor meetings, as
needed
• Establishes and maintains an administrative tracking
system for QT studies
24 24
Extensive Involvement
from QT-IRT
• Policy development
– Q&A for ICH E14, 2008: ECG measurement method, assay sensitivity,
baseline definition
– Q&A Revision 1, 2012: heart rate correction, late stage monitoring, sex
differences
– Q&A Revision 2, 2014: concentration-QT, large proteins, combination
products, special cases
– Q&A Revision 3, 2015: concentration-QT
• Numerous external presentations and workshops
– Drug Information Association, Cardiovascular Safety Research
Consortium
• Advancing science of cardiac safety
– IQ-CSRC study for concentration QT1
– Comprehensive in vitro Proarrhythmia (CiPA) initiative
Darpo et al. 2015
25
Summary
• Our groups interact at the FDA during all stages
of drug development
• Overlapping but complementary skillsets
– Greater understanding when working together
– Various pitfalls, but room for growth
Statistics
Pharmacometrics
26
Acknowledgements
• Sofosbuvir
– Karen Qi, Wen Zeng, Dionne Price, Division of Antiviral Products
• MCP-Mod
– Lei Nie, Dionne Price, Mohamed Alosh, Dinko Rekić, and Yaning Wang
• Total Kidney Volume Biomarker Qualification
– John Lawrence, DJ Marathe, James Hung, Martina Sahre, Sue-Jane Wang,
Hobart Rogers
• IRT-QT Team
– Devi Kozeli, Norman Stockbridge, Christine Garnett, Jiang Liu, Kevin
Krudys, Nitin Mehrotra, Joanne Zhang, Qianyu Dang, John Koerner (and
numerous other individuals throughout the years)
• Yaning Wang
• Office of Clinical Pharmacology at FDA
• Office of Biostatistics at FDA
27
Questions
28