Designing quality in
Colin R Gardner,
Currently: CSO, Transform Pharmaceuticals Inc
Lexington, MA, 02421
Formerly: VP Global Pharmaceutical R&D
Merck & Co Inc.
Acknowledgement for useful discussions:
Dr. Scott Reynolds,
Executive Director, Pharmaceutical Development,
Merck and Co Inc.
Presentation to Manufacturing Subcommittee of the FDA
Advisory Committee for Pharmaceutical Science.
Sept 17 , 2003
www.transformpharma.com
Summary
• Continuum of process development activities from
NCE selection through manufacturing
• Fundamental NCE characterization and process
development leads to meaningful control points
• Success of the scale up exercise is judged by
rational comparison of meaningful process and
product parameters
• Fingerprint parameters are identified to monitor
process robustness and used to flag issues before
control is lost
Issues within the industry
Drug Discovery / Development /
Marketing
Discovery Development Market
Targets
Pre-
Hits Submission& Lifecycle
clinical Phase 1 Phase 2a/b Phase 3
Leads Approval Management
Development
Candidate
2-5 yrs 0.5 - 2 yrs 1 - 2 yrs 1.5 - 3.5 yrs 2.5 - 4 yrs 0.5-2 yrs 10-20 yrs
R&D takes
6.5 - 13.5 years
Up to $800MM
Challenges: - Find safe and effective drugs
- Speed to market Source: PRTM
Drug company products
•The API
•The marketed dosage form(s)
• Approved label claim used to position
product in the market
Drug company products
•The API
•The marketed dosage form(s)
• Approved label claim used to
position product in the market
Intra-company Consequences
• R&D heads focus on potency, selectivity, safety and clinical
response
• do not uniformly recognize the importance of
investment in process chemistry and formulation
development
• Inexperienced clinical staff often set timelines and targets
independent of product development capabilities
• The goals and rewards of Discovery, Development and
Manufacturing staffs are often not aligned
• CEO’s have not regarded manufacturing excellence as a
competitive advantage
Issues created by the
regulatory agencies
• Depth of understanding of process
engineering
•Timeframe to review and understand the
regulatory filing
•Training of compliance inspectors –
especially for PAI’s
PAI examples
Scaling up a suspension formulation
drug excipients drug excipients
Batch size Biobatch Commercial
10 liters batch 100 liters
Mixing 15 mins 45 mins
time
FDA inspector conclusion:
“The processes are different”
Suspension formulation preparation
and filling
Preparation
tank Re-circulating
filling line
Filling Pump
tank
Filling Preservative
points adsorption to tubing
FDA inspector conclusion:
“Any stoppage of the filing process > 15 mins should result
in destruction of the entire batch
What can we do about this situation?
Manufacturing processes start with
the choice of the NCE, its form and
formulation
We must link discovery, early
development, process scale-up and
manufacturing
Industry role
• Develop methodologies to improve:
• Candidate selection
• Form selection and Formulation design
• Process development and optimization
• Process control
• Scale-up and tech transfer
• Process validation
• Process monitoring and continuous improvement
• Demonstrate reduced risk to regulatory agencies
• Obtain regulatory relief
• Demonstrate value to company management
HOW?
1. Picking better development
candidates:
Building in “developability”
Pre-clinical Research &
Early Development Process
Genomics Libraries
Early Discovery
Target HTS Sample collection
Hits to Leads
Discovery Development
Synthetic
Scale up chemistry
100-500mg
Lead optimization
in vitro selectivity
Animal model F& GLP
Probe Tox, F Tox
in vitro metabolism PK, met
Ph I
Animal model 2-4 cmpds
efficacy Process Pharm.
chemistry Sci.
in vitro Tox
Lead optimization
New R&D Challenges
Resource
constraints
Discovery Pharmaceutical Time
revolution constraints
Development
Drug Preclinical Clinical
Discovery development development
Pre-clinical Research &
Early Development Process
Genomics Libraries
Early Discovery
Target HTS Sample collection
Hits to Leads
Discovery Development
Synthetic
Scale up chemistry
100-500mg
Lead optimization
in vitro selectivity
Animal model F& GLP
Probe Tox, F Tox
in vitro metabolism PK, met
Ph I
Animal model 2-4 cmpds
efficacy Process Pharm.
chemistry Sci.
in vitro Tox
Lead optimization
Candidate selection:
Building in “Developability”
Lead
(active molecule)
Potency Physical properties Metabolism
Potency
Selectivity
Metabolism
Selectivity
Best leads
Physical / chemical
LO properties
Biopharmaceutics
(optimized molecule)
2. Form and formulation selection
Product Development Timeline
Develop First Drug Substance Transfer to Validation Launch
Synthetic Supplies Manufacturing Quantities
Route • Develop Process and Scale-up
• Establish Specifications PAI
Non GLP IND/Phase Safety Assessment
Probes I/II Safety
• Extended Safety Studies Carcinogenicity
• Degradate Qualification
Discovery
• Preformulation Studies Product Development Launch
• Biopharm Evaluation Quantities
Launch
• Formulation • Phase I/IIA • Composition • Process Transfer to Validation
Design Formulations & Process Development Manufacturing PAI
• Analytical Defined and Scale Up
Methods • Probe • Biobatch
Stability • Specifications
• MCSS
Clinical Program
• Phase I/IIA • Phase IIB •Phase III
• Wide Dose Dose •Final process
Range Range
• Multiple •>1/10 scale
Formulations
First in Phase Phase File Approval
Man IIB III NDA
WMA
$5-10MM $250-800MM
3-10 years 4-8 years
crg development timeline
Exploration of solid forms
Traditional High throughput
process impurity process impurity
or degradate or degradate
3 1
2
process (t,T) process (t,T)
5 4
solvent solvent
Weakly Crystalline Anhydrous Form
Solubility >100 mg/mL
Crystalline Trihydrate
Solubility ~0.73 mg/mL
Ritonavir: HIV protease inhibitor
H3C CH3
O O
H
H 3C N N
N N N O
H H
CH3 O OH
H3C S S
N
Case history:
1992 ABT-538 discovered
1996 Launch of semi-solid capsule/polymorph I
1998 Polymorph II appears, <50% solubility
Product pulled from the market
1998 - 1999 Massive effort to reformulate the product
1999 Reformulated softgel capsule launched
Summary of Ritonavir Crystal Forms
IV
mp 122 °C mp 125 °C mp 80 °C mp 97 °C mp 116 °C
Launch in 1996
Launch in 1996
Summer of 1998
Summer of 1998 Morissette et al. PNAS 100, (2003).
TransForm 2002 – 6 week effort
2002 5 forms found
TPI 745: New salt form with
improved solubility
Solubility
New TPI Form Has Faster Onset
4
30 mpk P.O.
2 10 TPI-745A
TPI-745B
Parent
Salt form with
“solubility
modifier”
4
1.5 10
4
1 10
Cmax Tmax AUC
5000 TPI-A 23.2±6.2 1.3±1.0* 139±26
TPI-B 19.6±4.6 2.1±1.1 135±24
T-745 21.4±4.0 2.8±1.6 150±43
0
0 2 4 6 8
time, hours
Faster Onset, Increased Bioavailability
and Linear Dose Response
New form & formulation combination significantly improves
dissolution, resulting in better onset and bioavailability
30
TPI-336
Marketed capsule
Neat chemical in capsule
Solution in 2:1 PEG/water
20
10
0
0 5 10
Dose, mg/kg
3. Process development
The current norm The future
raw material properties raw material properties
3 1
2
process process
5 4
conditions conditions
environmental environmental
Pharmaceutical Process Development:
Objectives
• Provide a continuous link from early phase
characterization to final manufacturing process
• Define process based on unit operations
approach
• Provide a road map for tracking success of
scale up activities and technology transfer
• Enable effective process monitoring and
improvements
Pharmaceutical Process Development:
Initial Design
• Identify parts of process which are most
susceptible to failure upon scale-up
• Conceptual “scale down” of the final
manufacturing process into the pilot plant
and the lab
Process Understanding
• Determine fundamental process constraints
– Where appropriate, utilize unit operations
which are most forgiving – lower risk
• Identify underlying principles which control
process
– Avoid “black box” analysis
– Identify appropriate process parameters to
monitor and control - value of PAT
- provides confidence
about process robustness
Pharmaceutical Process Development:
Optimization
• Optimization Studies
– Find regions of process parameters where
performance is most stable
raw material properties raw material properties
– Design process to operate within this region.
3 1
2
process
5 4
process
conditions conditions
environmental environmental
Process optimization
Process most stable
Target values
Region where process is unstable
Pharmaceutical Process Development:
Optimization
• Optimization Studies
– Find regions of process parameters where
performance is most stable
raw material properties raw material properties
– Design process to operate within this region.
3 1
2
•Process Robustness
process
5 4
process
conditions conditions
–Stress ranges of variables
–Include ranges in materials,
environmental environmental
environmental conditions, process
parameters
Process optimization
Region where process Process most stable
is robust Target values
Region where process is unstable
Pharmaceutical Process Development:
Process Control
• Define process through measurable,
quantitative endpoints – PAT?
• Eliminate dependence upon qualitative
endpoints
• Evaluate how process can respond to
variations in process equipment performance
and/or raw materials characteristics
• Provide continuous fingerprint of process
performance – NOT regulatory specifications
Pharmaceutical Process Development:
Continuous Improvement
• “Hooks” for future process improvement.
– Plan into development program collection of
“fingerprint” data for future comparisons
– Design validation protocols to collect similar
“fingerprints”
– Use in manufacturing to continuously
monitor process operation and status
Process optimization
Region where process Process most stable
is robust Target values
Fingerprint region to
monitor process
robustness and
Region where process is unstable prospectively identify
drifts
Summary
• Continuum of process development activities from
NCE selection through manufacturing
• Fundamental NCE characterization and process
development leads to meaningful control points
• Success of the scale up exercise is judged by
rational comparison of meaningful process and
product parameters
• Fingerprint parameters are identified to monitor
process robustness and used to flag issues before
control is lost
