Utilizing RM in a Submission for Developing Critical Process Parameters and Critical to Quality Attributes

1
Utilizing RM in a Submission for Developing
Critical Process Parameters and Critical to
Quality Attributes

• Kelly Canter, PhD
• Right the First Time Program Office
• Pfizer Inc., Groton, CT
• FDA/Industry Statistics WorkshopSeptember 2006

2
Outline

• QbD Terminology and Value Proposition
• Risk Assessment Process (Case Study)
• Experiments, PAT and Prioritization
• Creation of Design Space

3
Alignment of ICH Q(8)

• Enhanced knowledge of product performance . .
  .
• Establish range of material attributes,
  processing options process parameters
• Demonstrated product/process understanding
• Results from PAT, DOE, Science of Scaling
• Appropriate application of risk management
  principles
• Establish Design Space
• Flexible regulatory approaches
• Risk based regulatory decisions
• Mfg. process improvements w/in approved design
  space
• Real time quality control Reduce product release
  tests

4
Quality by Design Right First Time
Process Understanding
Continuous Improvement
Process Control
Process Control Strategy
Process Capability Monitoring
Commercializable Manufacturing Process (API or DP)
e.g. Cpk
Change Control Strategy and Implementation
Continuous Improvement (Process Changes)

• Risk Assessment
• Prioritized Experimental Plans
• Prioritized PAT Plans

Regulatory Filing/Approval
Experimentation /Method Dev/Documentation
Design Space Definition
Launch
5
Why Do QbD?(Value Proposition)
Getting at the Right Process Knowledge Value
to Pfizer, FDA and Patients

• Work Impact During Development
• Decrease ICH re-dos
• Decrease Validation re-dos
• Decrease Clinical Batch re-dos
• Transparent assessment of risk
• Prioritization

• Improvements to our Products and Processes
• Decrease Variability
• Assure market supply
• Faster change implementation
• Science support Quality investigtations
• Reduce COG
• Streamline regulatory reviews (SE)
• Framework for decreased regulatory burden
• Standardization

6
Process Understanding
People
Process Parameters
Quality Attributes
Inputs to the process control variability of the
Output
Equipment
I N P U T S (X)
y ƒ(x)
Measurement
y
Process
OUTPUT
Materials
Environment
J. Scott, ASTM, London 2004
7
What is a Quality Attribute?

• Definitions
• Quality Attribute
• A physical, chemical or micorbiological property
  or characteristic of a material.
• Key Quality Attribute (KQA)
• Potential to impact product quality or process
  effectiveness
• Evaluated by an associated analytical method.
• Critical Quality Attribute (CQA)
• impacts the safety or efficacy of a drug products

8
What is a Process Parameter?

• Definitions
• Process Parameters
• Broadly defined as machines, materials, people,
  processes, measurements and environments
• Key Process Parameter (KPP)
• Influences product quality or process
  effectiveness
• Critical Process Parameter (CPP)
• Influences a CQA and that must be controlled
  within predefined limits to ensure the API or
  product meets its pre-defined limits

9
Risk Assessment Work Process
10
Risk Assessment and PrioritizationDecide whats
important to evaluate
Quality Attributes
Process Parameters
Many Ys
Many Xs

• Process
• Consensus decisions
• Use process experience
• Use project process knowledge
• Focus on the Voice of the Customer

• Process
• Cause and Effect Matrix with Effects focused on
  KQAs

Vital Few Ys Key Quality Attributes
Vital Few Xs Key Process Parameters
11
The QbD Work Process at a High Level
12
Risk Assessment Case Study
Dry Granulation Tablet
13
Risk Assessment Objectives

• Gain agreement on process scope
• Decide whats important to evaluate
• Prioritize parameters based on risk
• Gain agreement on high level experimental
  strategy
• Identify and prioritize PAT applications

14
Risk Assessment Work Process
15
Risk Assessment Meeting Participants

• RD
• Co-Facilitator
• API
• Analytical
• Formulation
• Chemical
• DP
• Analytical
• Formulation
• Chemical
• Ext. Subject matter experts
• PAT RD
• Statistician
• Scribe (workbook)
• Line management
• Team Co-Leader

• Pfizer Global Manufacturing
• Co-Facilitator
• API Tech Services
• DP Tech Services
• Manufacturing Supervisor
• QC
• QA
• Team Co-Leader
• Subject matter experts
• PAT PGM
• Line management

16
Risk Assessment Work Flow
Create a Process Map with Focus Areas
Identify all Quality Attributes and Determine How
To Measure
Identify and Prioritize all Process Parameters
(KPPs)
Group KPPs into Experiments
Create PAT Prioritization Matrix
Document
Yellow font Pre-work required.
17
Risk AssessmentStep 1. Create a Process
MapDescribes the composition and boundaries of
each focus area.
Process Step
Commercial Manufacture
Boundaries
Raw Material Dispensing
CP-526, 555-18, Cellulose microcr, PH200,
Calcium Hydrogrenphosphate (amhydrous),
colloidal Silicon dioxide, Croscarmellose Sodium
Raw Material Dispensing
Focus Area 1
Preblending
300 L bin15 minutes
Initial Blend
Initial Blend
Comil0.8 mm sieve
Focus Area 2
Sieving
De-lumped Unlubed Blend
De-lumped Unlubed Blend
300 L bin2 minutes
Focus Area 3
Lube Blend
Lubed Blend
Dry Granulation and Blend
Bepex K 200/50Roll Deep Pocket Screen Size 0.8
mm
Lubed Blend
Focus Area 4
Blending
300 L bin3 minutes
Unlubed Granulation
Unlubed Granulation
300 L bin3 minutes
Focus Area 5
Lube Blend
Final Blend
Final Blend
Focus Area 6
Compression
IMA Comprima 300
Tablet Cores
Tablet Cores
Focus Area 7
Film Coating
Glatt GC 1250
Film Coated Tablets
18
Risk AssessmentStep 2. Identify QAs and How
MeasuredStep 3. Identify and Prioritize
PPsFocus Area 4 - Dry Granulate Blend
Key Attribute Y Y Y Y N Y Y
Rank 7 7 7 7 5 10 10
Process Parameter Sieve Cut Potency Blend Uniformity Particle Size Distribution Mill Choking Surface Area Hardness (Focus Area 6) Content Uniformity (Focus Area 6) Score Exp./Approach
Operator Training Procedures 10 10 10 10 0 10 10 840 FMEA
Roll Force 10 10 10 1 0 10 10 777 DOE
Screen Size 10 10 10 10 0 5 5 632 DOE
Gap Width 10 10 5 5 0 5 5 585 DOE
Material Throughput 10 1 5 10 0 1 1 437 DOE
Roller Compaction Calibration 5 5 5 1 0 5 5 427 FMEA
Sampling Size 10 10 10 1 0 1 5 421 MSA
Roll Speed 5 5 5 10 0 1 1 370 DOE
Equipment Aging 5 1 10 1 0 1 1 286
Transfer Distance into Roller 10 5 1 1 0 1 5 278
19
Risk AssessmentStep 4. Group Key PPs by
ExperimentsFocus Area 4 - Dry Granulate Blend
Raw Materials
Define Process Flowchart
. . . . . . .
Define Focus Areas
. . . . . . .
Identify KQAs and Associated Measurement
. . . . . . .
Identify and Prioritize KPPs
Define Experiments
. . . . . . .
20
Risk AssessmentStep 5. Create PAT
Prioritization Matrix Focus Area 4 - Dry
Granulate and Blend
Focus Area Quality Attributes Metric/Unit Measurement System Probability of Success (H/M/L) Criticality/Benefit (H/M/L) Cost (H/M/L) KeyAttribute (Y/N)
4 Sieve cut potency Intent HPLC M L M Y
4 Flowability L L H Y
4 Blend Uniformity rsd HPLC M M H Y
4 Segregation Index rsd JJ Tester L L H Y
4 Particle Size Distribution Size Sieve Analysis H L H Y
21
Risk AssessmentStep 6. Document the Process
Understanding

• Risk Assessment
• Experimental Strategy
• Protocols
• Primary Data
• Scientific Reports
• Global Document Management System

22
Initial Risk Assessment Complete
23
The Work Process
Risk Assessment
24
Experimental PlanningExample DOEFocus Area 4
- Dry Granulate Blend
Key Attribute Y Y Y Y N Y Y
Rank 7 7 7 7 5 10 10
Parameter Sieve Cut Potency Blend Uniformity Particle Size Distribution Mill choking Surface Area Hardness (Focus Area 6) Content Uniformity (Focus Area 6) Score Exp. Strategy
Operator Training Procedures 10 10 10 10 0 10 10 840 FMEA
Roll Force 10 10 10 1 0 10 10 777 DOE
Screen Size 10 10 10 10 0 5 5 632 DOE
Gap Width 10 10 5 5 0 5 5 585 DOE
Material Throughput 10 1 5 10 0 1 1 437 DOE
Roller Compaction Calibration 5 5 5 1 0 5 5 427 FMEA
Sampling Size 10 10 10 1 0 1 5 421 MSA
Roll Speed 5 5 5 10 0 1 1 370 DOE
Equipment Aging 5 1 10 1 0 1 1 286
Transfer Distance into Roller 10 5 1 1 0 1 5 278
25
Experimental Design for Gerteis Study

• D-Optimal Design
• Process Parameters
• Quality Attributes
• Granulation particle size
• Sieve cut uniformity
• Blend potency uniformity
• Tablet potency uniformity
• Hardness at 7 kP compress. force
• Friability at 7 kP compression force

• Roll force
• Gap width
• Granulating sieve size
• Granulator speed

26
DOE Regression ModelsModel Coefficients (p -
values)
Main Effects Main Effects Main Effects Main Effects Interactions Interactions Quad.
Quality Attributes(Intercept) Roll Force Gap Width Mill Screen Size Mill Speed Roll Force x Gap Width Roll Force x Mill Screen Size Mill Screen Size 2
Gran Particle Size (216) 51 (lt0.0001) --- 68(lt0.0001) --- --- 38(0.0006) ---
Sieve Cut RSD (41.4) -6.4 (lt0.0001) --- 1.2(0.2650) --- --- --- -17.9(lt0.0001)
Log (Gran RSD) (-0.07) 0.10 (0.0758) --- 0.17(0.0051) --- --- --- ---
Log (Tablet Potency RSD) (-0.15) -0.08 (0.0025) -0.06 (0.0308) 0.06(0.0180) --- --- --- ---
CF _at_ Tablet Hard. 7 kP (6.8) 2.0 (lt0.0001) -0.6(lt0.0001) --- --- -0.5(0.0002) --- ---
FRI _at_ Tablet Hard. 7 kP (0.06) --- --- 0.02(0.0320) --- --- ---
27
Requirements to Map Design Space
Boundary Conditions Boundary Conditions
Process Parameters Process Parameters
Gap Width 1.7 3.5 mm
Mill Screen 0.8 1.5 mm
Quality Attributes Quality Attributes
Sieve Cut Variability ( RSD) lt35
Bypass lt15
Compression Force at 7 kP Hardness lt8.5 kN
Tablet Uniformity lt1.0
28
Rationale for Process Ranges within Design Space
(0.8 mm Mill Screen Size and 50 rpm Granulator
Speed)
Yellow Region Acceptable combinations of
process parameters.
Unacceptable space
29
Rationale for Process Ranges within Design Space
Contour Map Bypass Weight

• Bypass weight loss is highest in upper left
  quadrant of Roll Force vs Gap Width

3.8
Statistics and Model
3.2
Response(intercept) RF Coefficient(p-value) GW Coefficient(p-value)  RFGW Coefficient(p-value)
Ln Bypass Wt (0.70) -0.71 (0.0045) 0.37 (0.0479) -0.81 (0.0046)
2.6
Gap Width (mm)
2.0
1.4
4
6
8
10
12
Roll Force
Unacceptable space
30
Conclusions from DOE (D-Optimal)

• Increasing roll force improved (lowered RSD)
  granulation and tablet uniformity.
• Increasing roll force also reduced bypass
• However, increasing roll force increased the
  tablet compressional force required (Safety
  Margin 8.5 kN)
• Acceptable process range for roll force is 5-9 kN
  (see Design Space)

31
The Work Process
Risk Assessment
Experimental Planning
32
Experimental Strategy Prioritization Example
Fractional Factorial (Focus Areas12)
Central Composite Focus Areas 12)
1
Full Factorial w/center
Add axial points to Full Factorial
3
2
Gage RR (Focus Area 3)
FMEA (Focus Areas 23)
4
Etc
33
The Work Process
Risk Assessment
Prioritization
Experimental Planning
34
Building Models KQA f (KPP1, KPP2,
KPPi)Conclusions

• Operating target and ranges were identified for
  each of the following key parameters, key
  attributes
• Roll force (KPP1)
• Impacts particle size, blend uniformity, tablet
  uniformity (KQA1, KQA2, KQA3)
• Gap width (KPP2)
• Impacts tablet uniformity (KQA3)
• Screen size (KPP3)
• Impacts sieve cut uniformity (KQA4)
• Granulator speed (KPP4)
• Not significant for KQAs investigated

35
Control-, Design- and Knowledge space
Knowledge Space
Knowledge Space
Design Space

Control Space
Proven Acceptable Range
Normal Operating Range
36
Design Space
37
Acknowledgements

• Chris Sinko
• Roger Nosal
• Jim Spavins
• Vince McCurdy
• Tom Garcia
• Christina Grillo
• Mary Am Ende
• Dan OConnell