Lifecycle Approach to Process

Validation
Best Practices and Strategies

July 19th, 2016

Patrick Donohue, Drug Product Development, Janssen R&D

Disclaimer
The contents of this presentation are my personal views
and do not reflect those of Johnson & Johnson or its
affiliates.

2
Do we really know how to
link process capability with
in vivo performance?

Adapted from Ken Hinds, PhD (PDA/FDA JRC, 2014)

3
Stage 1: Focus on Criticality

 Are all CQAs (CPPs/CMAs) created equally?

– 2011 FDA Process Validation Guidance: “With a lifecycle approach

to process validation that employs risk based decision making
throughout that lifecycle, the perception of criticality as a
continuum rather than a binary state is more useful.”

Adapted from Ken Hinds, PhD (PDA/FDA JRC, 2014)

4
Focus on Criticality

 Definition: Criticality is defined as a classification of an

item, (e.g., process, equipment, parameter) that
expresses significance given to the impact of that item,
and should be controlled or monitored to ensure product
quality, safety or efficacy. (PDA TR54)

Criticality = Impact x Uncertainty

 Impact scored via science and prior knowledge

 Uncertainty scored via statistics and process experience
 Criticality as a guidepost for monitoring and control

Adapted from Dr Mike Long, MBB (PDA/FDA JRC, 2014)

5
Focus on Criticality

 Control strategy driven by

relative risk of process
parameters and material
attributes

 Objectives:
– Understand how process
parameter input variation ties
to product output variation
– Understand how starting
material input variation ties to
product output variation
– Design a manufacturing
process that consistently
delivers acceptable product
output variation

Adapted from Ken Hinds, PhD (PDA/FDA JRC, 2014)

6
Focus on Risk

 Do all risks require the same level of mitigation/control?

– PDA TR54: “application of risk management activities relevant to

the type and level of risk inherent in each process will enable
product realization, establish and maintain a state of control, and
facilitate process improvement.”

Adapted from Ken Hinds, PhD (PDA/FDA JRC, 2014)

7
Focus on Risk

 Definition: Risk is defined as the combination of the

probability of occurrence of harm and the severity of that
harm (ICH Q9)

Risk = Severity x Occurrence

 Each attribute identified as critical to quality should be

assessed by severity and occurrence
 Severity scored via science and prior knowledge
 Occurrence scored via statistics and process experience
 Risk as a guidepost for monitoring and control

Adapted from Dr. Mike Long, MBB (personal communication)

8
Focus on Risk

 Control strategy driven by

relative risk of product
attributes

Output (safety/efficacy)
 Objectives:
– Understand how product
attribute variation ties to
clinical performance
– Establish clinically relevant
specifications linking clinical
performance to product
quality Input (CQAs)

Adapted from Ken Hinds, PhD (PDA/FDA JRC, 2014)

9
 Process Validation Sequence

Stage 1 Stage 2 Stage 3

Based on product When is
quality and patient Is the process Are the variables confidence What is looked for
safety known? known? achieved? and for how long?

Process Process Commercial Monitoring &

Process Design
Understanding Qualification Manufacture Improvement

Risk Assessment

10
Process Validation Stage 1

 Process Design:

The commercial manufacturing process is defined during

this stage based on knowledge gained through
development and scale-up activities. (FDA 2011
Validation Guidance)

Stage 1a - Process Development

Stage 1b - Process Characterization (PDA TR60)

11
Process Validation Directives

Manufacturer’s Should:

 Understand the sources of variation

 Detect the presence and degree of variation

 Understand the impact of variation on the process and

ultimately on product attributes

 Control the variation in a manner commensurate with the

risk it represents to the process and product

– FDA 2011 Validation Guidance

12
Sources of Variation

 “Homogeneity within a batch and consistency between

batches are goals of process validation activities.
Validation offers assurance that a process is reasonably
protected against sources of variability that could affect
production output, cause supply problems, and
negatively affect public health.”

– FDA 2011 Validation Guidance

13
Sources of Variation

 Within-batch (Intra-batch) variability

– Can be observed with enhanced/extended sampling

 Between-batch (Inter-batch) variability

– Can be observed with reduced sampling or enhanced sampling

 Analytical method variability

– Can be observed with measurement system analysis techniques,
e.g. Gage R&R

 Unknown variability
– Unavoidable
– e.g. variation in sample handling, shipping, etc.

14
Sampling Strategy

 Standardization of within-batch sampling locations

Example of stratified random sampling:
– Beginning  first 10% of manufactured units
– Middle  middle 20% of manufactured units
– End  last 10% of manufactured units

End Middle Beginning

Samples Samples Samples

Conveyor

15
Stage 1: Process Design Cycle

• Defined from QTPP, CQAs, platform knowledge, etc.

Process Design • Continuously updated during development

Sampling Plan • Estimate Variation

Creation • Determine sample size

Manufacturing & • Explore Design Space

Testing • Challenge Process Parameters

• Identify Sources of Variability

Data Analysis • Quantify impact to CQAs

• Repeat as appropriate during process simulation

Repeat studies, engineering runs, clinical runs, etc.

Proceed to Stage 2

16
Process Validation Stage 2

 Process Qualification

During this stage, the process design is evaluated to

determine if the process is capable of reproducible
commercial manufacturing. (FDA 2011 Validation
Guidance)

Stage 2a – Facility, Equipment, Utility Design & Qualification

Stage 2b – Process Performance Qualification

17
Stage 2: Focus on Risk

 During Stages 1 & 2 the focus is on ‘beta risk’

– “Guilty until proven innocent”, a.k.a. consumer’s risk

– In process validation, beta probability is the chance a batch is

released given that one or more of the manufactured units have
failing attribute levels

– In statistics, beta = probability of committing a Type II Error

Adapted from M. Johnson, ISPE Process Validation Conference, October 2013

18
Stage 2: PPQ Success Criteria

 All results must meet release specifications

Individual Value Plot of CQA 1

B M E B M E
Batch 1 Batch 2 Batch 3 Batch 4 Batch 5
98.0

97.5
CQA 1

97.0

96.5

96.0 Spec

B M E B M E B M E
Sampling Location
Panel variable: Batch

19
Stage 2: PPQ Success Criteria

 Some statistical tools may be unsuitable for the application

of objective pass/fail criteria

 At this stage, limited data from clinical and/or engineering

runs exist  further knowledge is needed about the breadth
of variation from raw material attributes and process
parameters

IF SETTING OBJECTIVE STATISTICAL LIMITS

THEN PROCEED WITH CAUTION!

20
Stage 2: PPQ Data Analysis

 Batch-Specific Tolerance Intervals (TI)

– May be used to directly measure batch homogeneity
– A range of values is estimated for the majority of units produced in a
single batch

 Analysis of Variance (ANOVA)

– May be used to directly measure batch homogeneity and consistency
– May be used to partition variability into ‘Variance Components’
– May be used to estimate differences between sampling locations
(B/M/E) or batches

 Continuous data should NOT be rounded before analysis

– Rounding favored for official documentation (e.g. C of A), but
unfavorable for statistical analysis  may generate statistics that
significantly misrepresent an attribute’s variability

21
Stage 2: CPV Planning

 Within-batch variation and between-batch variation

observed during PPQ can be used to formulate a CPV plan

– “The increased level of scrutiny, testing, and sampling should continue

through the process verification stage as appropriate, to establish
levels and frequency of routine sampling and monitoring for the
particular product and process.”

 2011 FDA Guidance, “Process Validation: General Principles and Practices”

22
Stage 2: CPV Planning

Stage 2B Enhanced Expected CQA Unexpected CQA

Sampling Analysis within- and within- and
between-batch between-batch
variability variability
Yes. CQA meets
next order of Release sampling only Continue enhanced
coverage using TI for Stage 3A sampling for Stage 3A
method
No. CQA does not
meet next order of Continue enhanced Continue enhanced
coverage using TI sampling for Stage 3A sampling for Stage 3A
method

Adapted from Dr. Mike Long, MBB (personal communication)

23
Stage 2: CPV Planning

 Example criteria for batch-specific TIs

Attribute Stage 2b TI Stage 2b TI

Severity Ideal Confidence / Coverage Exceptional Confidence / Coverage

High 95 / 95 95 / 99

Medium 95 / 90 95 / 95

Low 95 / 80 95 / 90

Adapted from Dr. Mike Long, MBB (personal communication)

24
Process Validation Stage 3

 Continued Process Verification

 Ongoing assurance is gained during routine production

that the process remains in a state of control. (FDA 2011
Validation Guidance)

Stage 3a – Short Term CPV

Stage 3b – Long Term CPV

25
Stage 3: Focus on Risk

 During Stage 3 the focus begins shifting to ‘alpha risk’

– “Innocent until proven guilty”, a.k.a. producer’s risk

– In process validation, alpha probability is the chance of rejecting

a batch given that none of the manufactured units have failing
attribute levels

– In statistics, alpha = probability of committing a Type I Error

Adapted from M. Johnson, ISPE Process Validation Conference, October 2013

26
Statistical Process Validation

Given the right investment of time and resources:

 Improved understanding of the manufacturing process

 Improved ability to detect issues and find root causes

 Improved product quality

 Decreased consumer risk

27
Acknowledgements

 Kenneth Hinds
 Mike Long
 Darin Furgeson
 John Motzi

28
Thank you

jdonohue@its.jnj.com
July 19th, 2016
Backup Slides
Sampling and Testing

 Analytical Testing Order

 Can increase one’s ability to detect signals from the

manufacturing process, versus signals from the assay, by
stratifying the sample run sequence

 Natural tendency to segregate samples by sampling

location and test  UNFAVORABLE FOR HIGHLY
VARIABLE ASSAYS
– Sampling location comparisons within individual batches become less
clear as to their cause, from the process or from the assay.

31