DEBATING THE OPERATING CURVES FOR DDU TESTS ON MARKETED INHALERS

1
DEBATING THE OPERATING CURVESFOR DDU TESTS ON
MARKETED INHALERS
Industrys problem with the current FDA DDU
requirements

• Presented by
• Bo Olsson, AstraZeneca RD Lundon behalf of
  IPAC-RS DDU Working Group
• 27 April 2004

2
Brief History
IPAC-RS developed and submitted PTI test proposal
(Nov. 2001)
FDA asked for clarifications, especially for
situations with off-target batch means (Nov 2002)
Industry comments indicate DDU specifications too
tight
IPAC-RS submitted modified procedure (March 2003)
Form Working Group under ACPS to resolve
remaining issues and finalize test (Jan 2004)
FDA Draft CMC Guidances (1998-1999)
FDA asked for theoretical results on other
quality standards (April 2003)
FDA proposed standards equal to the guidance test
as resolution? (Oct 03)
IPAC-RS submitted requested information and
generalized code (June 2003)
FDA statisticians discussed additional details
(Aug 2003)
3
Joint FDA IPAC-RS Working Group

• Established in 2004 under FDA Advisory Committee
  for Pharmaceutical Science (ACPS)
• Populated by senior representatives from FDA and
  technical experts from OINDP industry
• Robert ONeill (Director, Office of
  Biostatistics, CDER), Chair
• Moheb Nasr (Director, New Drug Chemistry, CDER)
• Badrul Chowdhury (Director, Pulmonary Division,
  CDER)
• Lawrence Yu (Director, Office of Generic Drugs,
  CDER)
• Overall objective to develop a mutually
  acceptable, standard DDU specification (test and
  acceptance criteria) for OINDP, and make a formal
  recommendation to the ACPS in 2004

4
Consensus Reached to Date

• Parametric tolerance interval (PTI) approach is
  an improvement on current DDU test
• Concept requires refinement and further
  development to address regulatory requirements
• It is time to move forward and come to closure
• Working Group is formed to devote necessary
  resources and time to resolve issues (through
  data review and analysis, and development of
  appropriate statistical procedures)

5
Issues where Consensus is Needed

• Evaluation of simulated performance and actual
  performance
• Zero tolerance criterion
• Statistical details of the test design
• Applicability to non-normal distributions
• Better understanding the difference between
  operating characteristic (OC) curves of draft
  Guidances and IPAC-RS proposal

6
Focus of Todays Presentation

• FDA is currently not convinced that industry has
  a problem
• Show data to demonstrate that Draft Guidance
• does not reflect range of OINDP capabilities
• is overly restrictive compared to non-US public
  standards
• does not reflect range of US regulatory practice
  1990-2003

7
Data

• IPAC-RS DDU database (2001)
• Non-FDA public standards (2004)
• FDA approved specifications (1990-2003)

8
IPAC-RS DDU DATABASE (2001)
80 products, 46 816 observations release and
real-time stability data
Multiple strengths counted as different products
9
Product Status
US commercial and development products are
typically also available on or developed for
non-US markets
10
Product Types
11
46 816 observations
CFC susp
HFA susp.
Pre-met. DPI
Device met. DPI
HFA soln.
Nasal
200
Product mean
175
150
125
100
Delivered Dose, LC
75
50
25
0
Products/batches/tests
12
Comparison of Variability grouped by Product Type
13
Comparison of Variability grouped by Product
Status
HFA solution
14
What the Range of Variability Means Practically

• Majority of products (mostly HFA suspension MDIs
  and device metered DPIs) would have high failure
  rate if tested against requirements in Draft
  Guidance
• High failure rates are unacceptable from the
  business perspective
• Substantially increases cost
• Stability failures and recalls
• Process is deemed out of control by the
  compliance division

15
Key Message 1

• The DDU requirements in the FDA Guidances are
  overly restrictive and are not reflective of the
  range of technological capability of OINDP on the
  market and in development

16
NON-FDA PUBLIC STANDARDS (2004)
17
DDU multidose DPI/pMDI Standards in US, Canada,
Europe, Australia
OC curves assuming normal distributed data with
batch mean at 100 of label claim and batch
standard deviation (?) made up from equal parts
of between- and within-container variation.
18
Key Message 2

• Public standards for DDU quality other than the
  FDA Draft Guidances
• are less stringent,
• more accurately reflect the range of OINDP
  capabilities, and
• represent a better balance between process
  capability and assuring public safety
• There appears to be room for alignment of the FDA
  Draft Guidance standard to other public standards
  without compromising public safety

19
FDA-APPROVED SPECIFICATIONS (1990-2003) COMPARED
TO FDA DRAFT GUIDANCE
20
IPAC-RS 2003 Survey

• Requested OC curves for DDU OINDP specifications
  approved by FDA 1990-2003
• 24 approved specifications received
• Frequent exceptions from the draft guidance
  specification
• 16 of 24 are to the right of FDAs curve at 90
  acceptance and batch mean on target
• 8 of 24 are to the left (CFC MDIs and nasal
  suspensions)
• No chronological pattern observed
• e.g., more recently approved products are not
  necessarily all to the right or to the left

21
Batch Mean at 100 Label Claim
22
Batch Mean 5 off Target
23
Batch Mean 10 off Target
24
Batch Mean 15 off Target
25
What Exceptions?

• Inner/outer limits
• Limits on mean
• Through-container-life testing
• Life-stage means
• Sample size
• Number of actuations in a sample
• Outlier retesting
• Other

26
Key Message 3

• The FDA Draft Guidance requirements do not
  represent the actual regulatory DDU
  specifications for many OINDP approved by the FDA
  since 1990
• many OC curves differ substantially from that of
  the Draft Guidance specification
• many OINDP will have difficulty meeting the
  expectations set forth in the Draft Guidance
• the existence of the range of approved
  specifications does not negatively impact public
  health, rather it allows the availability of a
  range of products and product types to meet the
  medical need

27
EPILOGUE
28
Root of the Problem

• We speculate that Draft Guidance specification is
  based on historic FDA experience of
• CFC suspension pMDIs
• Pre-metered DPI
• Nasal pump sprays
• Old analytical methodology and sampling plans
  (excessive priming, beginning only, multiple
  shots)
• Now
• HFA pMDIs, multi-dose DPIs and other technologies
• More discriminating methodology and sampling
  plans
• but same limits

29

• An in vitro performance standard for DDU is
  needed such that it
• does not reject good batches
• provides a reasonable target for developers
• appropriately reflects contemporary formulation
  and device designs and capabilities of available
  manufacturing processes
• avoids frequent exceptions from the public
  Guidance, coupled with uncertain delays and
  negotiations during the review and approval
  process
• avoids unjustified perception that process is out
  of compliance
• Current FDA Draft Guidances specifications do
  not fulfil these needs

30

• Establishing more realistic delivered dose
  uniformity standards will allow guidance to be
  both more historically based and forward looking
• will not change quality of approved products
  (actual specs approved by FDA)
• will be more consistent with public standards
  outside the US
• will stimulate the development and filing of new
  and generic products in the US
• will facilitate and speed up the approval process
  in the US

31
IPAC-RS PTI Test

• Parametric test
• simultaneously controls mean and standard
  deviation
• no zero tolerance
• Suitable for all OINDP product types, multidose
  and single dose
• Flexible
• Same consumer protection as FDA draft Guidance
• Lower producer risk, similar to non-US public
  standards

32
Comparison of OC curvesPublic standards and
IPAC-RS PTI tests
33
Conclusion

• The industry has a problem
• The 2001 IPAC-RS proposal addresses industrys
  concerns
• Joint FDA IPAC-RS Working Group will work to
  resolve issues

34
Acknowledgments

• Dennis Sandell
• Orin Tempkin
• Henrik Thoning
• Keith Truman
• Ed Warner
• Steven White
• Paul Wright
• Bruce Wyka
• Jerry Zhang

• Jeff Blumenstein
• Mark Broughton
• Mary Devlin Capizzi
• David Christopher
• Claire DAbreu-Hayling
• Jennifer Gauvin
• Michael Golden
• Kristi Griffiths
• Niels Hartvig

• Jim Jamieson
• Paul Kovach
• Douglas Lee
• Stefan Leiner
• Svetlana Lyapustina
• Bo Olsson
• David Radspinner
• Lene Garde Rasmussen
• Darlene Rosario

IPAC-RS Companies

• Aradigm
• AstraZeneca
• Aventis
• Boehringer Ingelheim
• Eli Lilly

• Novo Nordisk
• Novartis
• Pfizer
• Schering-Plough

• GlaxoSmithKline
• IVAX
• Kos
• Nektar