Integrated Method Development and

Validation

RACI Conference - Chemical Analyses

Dr. Ludwig Huber

[email protected]
 Today’s Agenda
• Lifecycle management of analytical FDA 2014
procedures: development,
validation and routine use
• Using principles of Quality by
Design to get most robust FDA 2013
methods
• Defining validation parameters,
acceptance criteria and test
procedures NATA 2013
• Templates and examples for
efficient and consistent
documentations

© Copyright Ludwig Huber - LabCompliance Slide 2

FDA Guide – Bioanalytical Method Validation

Major differences to the 2001 Guide

• Section on System Suitability testing
• Inclusion of incurred sample
reanalysis
• Level of details on LBA similar to
chromatographic methods
• Concentrations below the LLOQ should
be reported as zeros
• Sample Analysis Reporting should
include: All accepted and rejected
analytical runs

© Copyright Ludwig Huber - LabCompliance Slide 3

 FDA Guide – Analytical Method Validation
• Components of Quality by Design
(QbD)
– Begin with an initial risk
assessment and follow with
multivariate experiments (design
of experiments).
– Lifetime management
• Requires submission of method
development data
– You should submit development
data within the method validation
section if they support the
validation of the method.

© Copyright Ludwig Huber - LabCompliance Slide 4

 Official Guidelines for Method
Validation
• ICH - Guidance for Industry - Q2 (R1) QbD
Text and Methodology components
Must be followed in US and Europe
• FDA: Analytical Procedures and Methods
Validation for Drugs and Biologics (Draft, Feb 2014)
• FDA - Industry Guidance
Bioanalytical Method Validation (Draft, Nov 2013)
• USP <1225>: Validation of Compendial Methods
• USP <1226>: Verification of Compendial Procedures
• USP <1224>: Transfer of Analytical Procedures
ICH = International Conference for Harmonization
USP = United States Pharmacopeia

© Copyright Ludwig Huber - LabCompliance Slide 5

 Method Validation
• The accuracy, sensitivity, specificity, and
reproducibility of test methods have not been
established and documented (W-187)
• Failure to validate analytical test methods used
for API for potency testing. (W-259)
• For example, your firm failed to validate the xxx
compound to quantify Peak A for potency and
robustness.
• Your firm has been unable to determine why the
chromatographic columns of the same make and
model had variability and could not provide
adequate separation (W-259)
www.fdawarningletter.com
© Copyright Ludwig Huber - LabCompliance Slide
Slide6
6
 Method Validation Parameters for
different Method Tasks (ICH Q2)
Identifi- Impurity Impurity
Analytical Task Assay
cation Quantitative Qualitative
Accuracy No yes No Yes

Precision
Repeatability No Yes No Yes
Intermediate No Yes No Yes
Reproducibility No Yes No Yes

Specificity Yes Yes Yes Yes

Limit of detection No No Yes No

Limit of quantitation No Yes No No

Linearity No Yes No Yes

Range No Yes No Yes
Robustness Expected to be done during Method Development

© Copyright Ludwig Huber - LabCompliance Slide 7

 Parameters and Tests (ICH Q2)
Parameter Tests (examples)

Accuracy Minimum at 3 concentrations, 3 replicates

Precision
Repeatability Minimum of 9 determinations over the specified range
Intermediate Over 3 days, 2 operators, 2 instruments,
Reproducibility Only required if testing is done in different laboratories

Specificity Prove with specific methods: HPLC, DAD, MS, dif. columns

Limit of detection Visual approach, S/N >= 3

Limit of Quantitation S/N >= 10, Standard deviation of response

Linearity Min 5 concentrations: visual, correlation coefficient (r)

Range 80 to 120% of test concentration, from linearity tests

© Copyright Ludwig Huber - LabCompliance Slide 8

 Why Should we Change the Traditional
Way
• Problems in routine use, too many failures
• Developers not end-users
• Low emphasis on method robustness and
ruggedness
• Poor knowledge on critical parameters –
– problems during method transfer
• No or inadequate use of risk assessment
• Invested time not very efficient

© Copyright Ludwig Huber - LabCompliance Slide 9

 Possible Conflict of Interests
• Development chemist
– Shortest time possible
• Routine User / QC Director
– No problem during routine use
– No out-of-specification situations
• Quality Assurance
– Enough documentation for inspections
• Regulatory Affairs
– Enough documentation for registration
• Finance
– Lowest development and validation cost
© Copyright Ludwig Huber - LabCompliance Slide 10
 Objectives of the New Approach
• Efforts for method development and validation
should be value adding: building knowledge
• Method will work consistently within its design
space
– Changing people
– Changing material (e.g., chromatographic
column)
– Environment (transfer)
• Focus on critical parameters using a risk based
approach
Compliance is still important !!!

© Copyright Ludwig Huber - LabCompliance Slide 11

 What we really want
• Design a method and validation procedures
to ensure that the method works for the
intended routine use, independently from
– Where it is being used
– Who is using it
– Specified instrumentation
– Actual method parameters, as long as
they are in the defined operating range

Trouble free operation – transfer –

With no method specific OOS results

© Copyright Ludwig Huber - LabCompliance Slide 12

QbD - Background and regulatory Situation
• Principles widely applied in all industries, particularly in car
industry
• Adopted by FDA in the 21st Century cGMP initiative
Reference: Pharmaceutical Quality for the 21st Century: A Risk-Based
Approach (2003)
• Adopted by ICH in Q8: Product Development, 2005,
updated in Q8 (R2), 2009
• In 2006, Merck & Co.’s Januvia became the first product.
• Starting to be adopted to analytical laboratories, e.g., used
to design robustness into analytical methods with the
Analytical Target Profile (ATP) concept
• 2013: FDA/EMA Q&As on method validation by QbD
• 2014: New FDA method validation guide with QbD components

© Copyright Ludwig Huber - LabCompliance

Slide 13
 QbD in Laboratories: Key Applications

• Development and validation of analytical

methods
– HPLC and others
• Transfer or analytical procedures
• Verification of compendial methods
• Analytical instrument qualification
• Dissolution testing
• Near Infrared Spectroscopy (NIR)
method
• Water analysis

© Copyright Ludwig Huber - LabCompliance Slide 14

 EFPIA Positioning Paper
• Establishment of Analytical Target Profile (method
performance criteria, acceptance criteria)
• ATP defines ‘what’ needs to be measured not ‘how’
• ATP is submitted to regulatory agencies and approved
instead of an analytical procedure
• Any analytical method conforming to the approved ATP
can be used
• Alternative methods, e.g., new technology, can be used
through internal change control procedure
In line with FDA‘s general approach for QbD (no re-approval
required as long as working in the approved design space)
Also in line with the European Variation Guideline and with ICH Q8

Reference: Ermer, European Pharmaceutical Review, Vol 10, Issue 3 (2011)

EFPIA = European Federation of the Pharmaceutical Industries and Association

© Copyright Ludwig Huber - LabCompliance Slide 15

 QbD in Laboratories – Current Situation
and key Applications
Situation
• No formal regulations or guidelines, no FDA pilot project
• QbD can be used for all critical analytical quality
parameters
• Some laboratories are starting to adapt QbD for analytical
method validation
• FDA/EMA address methods in Q&As sessions and guide
• EFPIA Positioning Paper
Key Applications
• Development and validation of analytical methods
• Method transfer, disolution testing
EFPIA = European Federation of the Pharmaceutical Industries and Association
© Copyright Ludwig Huber - LabCompliance Slide 16
 EFPIA Positioning Paper
• Establishment of Analytical Target Profile (method
performance criteria, acceptance criteria)
• ATP defines ‘what’ needs to be measured not ‘how’
• ATP is submitted to regulatory agencies and approved
instead of an analytical procedure
• Any analytical method conforming to the approved ATP
can be used (reality or dream?)
• Alternative methods, e.g., new technology, can be used
through internal change control procedure (???)
In line with FDA‘s general approach for QbD (no re-approval
required as long as working in the approved design space)
Also in line with the European Validation Guideline and with ICH Q8

Reference: Ermer, European Pharmaceutical Review, Vol 10, Issue 3 (2011)

EFPIA = European Federation of the Pharmaceutical Industries and Association

© Copyright Ludwig Huber - LabCompliance Slide 17

 Traditional Development & Validation of
Analytical Methods
• Select preliminary method, scope &
Not regulated

Preparation specifications
• Assure performance of equipment
• Assure that operators are qualifified
• Select and optimize method & parameters
Development • Robustness testing
• Define operational limits and SST
• Preliminary validation experiments

• Document final acceptance criteria

Validation
Regulated

• Document final scope

• Perform validation tests, incl. robustness

• Controlled transfer
Routine Operation • Regular review
• Controlled changes & Revalidation

© Copyright Ludwig Huber - LabCompliance Slide 18

 Quality by Design for Analytical Methods
Method development
Specifications Critical Method
Analytical Target Parameters and
Profile, Quality Critical Attributes, Risk
Target Method Assessment
Profile

Continuous Design Space,

Monitoring and Method
Improvements Operational
QC Tracking Ranges

Method
Control Strategy
Qualification
for CMAs
(ICH Q2)
System Suitability

© Copyright Ludwig Huber - LabCompliance Slide 19

 QbD Terms in Method Development
and Validation
Product Development Method Development Examples for Methods
Method Validation
Target product profile Analytical target profile Accurate quantitation of
(TPP) (ATP) impurities in drugs
Quality target product Quality target method LOQ <0.05%, precision
profile (QTPP) profile (QTMP) and accuracy at LOQ
better than 15%
Critical process Critical method Flow rate, temperature,
parameters (CPP) parameters (CMP) pH of mobile phase
Critical quality attributes Critical Method Resolution, peak tailing
(CQA) attributes (CMA)
Proven acceptable Method operational pH ± 1,
range (PAR) design range (MODR) col temperature ± 2

© Copyright Ludwig Huber - LabCompliance Slide 20

 Define the Analytical Target Profile

• Method operational intent (what the method has to measure)

Inputs from end-user department
– Ease of use, analysis cycle time, acceptable solvents,
analysis cycle time
• Method performance characteristics, e.g., precision,
accuracy, specificity, LOD/LOQ, linearity
• Acceptance criteria for method performance characteristics
• Which instruments will be used, where will the method be
used (specific lab, specific site, global)?
Example (incomplete): Quantitative impurity analysis compound at ≥0.05% with
an accuracy and precision of 15% RSD at the limit of quantitation and 5% at
20x LOQ.

© Copyright Ludwig Huber - LabCompliance Slide 21

 Determine Factors Impacting Critical
Method Attributes (CMA)
• Test Conditions Factor3
– HPLC Mobile phase composition, pH Factor2
– Column Temperature, detector wavelength Factor1
MCA
– Sample extraction time
• Material attributes
Factor1
– Matrix, sample stability, sample solubility, Factor2
column batch Factor3

– Reference standards, quality of reagents Use Fishbone diagrams

• Environmental conditions and risk assessment
– Humidity, room temperature, electromagnetic interference
• Random effects
– Analysts, e.g., skill level, thoroughness
– Timing, e.g., day and night shift
– Instrument, e.g., performance, maintenance
© Copyright Ludwig Huber - LabCompliance Slide 22
 Apply Risk Assessment to Support
Defined Criticality of Method Attributes
• Identify parameters with impact on the method’s performance (Risk
Identification)
– Rely on subject matter experts, Brainstorming meeting
– May also go back to development experiments
• Develop a prioritization matrix (Risk Evaluation)
– Look at factors with highest impact on method performance
– Link at specified instrument
functionality, performance and qualification
– Rank, e.g., in three categories: high (3), medium (2), low (1)
• Determine risk priority numbers for individual parameters
Severity Probability Detectability Risk Number

Factor 1 3 3 2 18
Factor 2 2 1 3 6

© Copyright Ludwig Huber - LabCompliance Slide 23

 Example Risk Prioritization Matrix
Impact of Method Parameters on Performance

Method UV
col. %organic
Parameter flow rate pH Wave-
temp. phase
Method Attributes length
LOQ 1 2 1 3 2
Linearity 2 1 1 3 2
Repeatability 2 1 1 2 3
Accuracy 2 3 1 3 1
Specificity 3 1 1 3 2
Risk Priority
10 8 5 14 10
Number (RPN)

1 = low, 2= medium 3 = high impact RPN ≥ 9 included in DOE study

© Copyright Ludwig Huber - LabCompliance Slide 24

 Determine the Method Operational Design
Range (MODR) through DOE
• From the „Critical Method Attribute“ exercise, select factors that
based on the risk assessment will impact method performance.
• Choose levels of each factor (two or higher)
• Select range over which factors will be varied,
e.g., in two level study there will be a high and low level value
– Requires good knowledge of the method
• Use the multivariate experimental design approach
• Define and perform experiments
• Perform statistical analysis of data
• Interpret the data
• Perform follow-up runs (if necessary)

© Copyright Ludwig Huber - LabCompliance Slide 25

 Design of Experiments (DoE)
• Early DoE strategies began in 1920s
• Part of QbD
• Helps to understand the cause and effect relationship
between input factors and output (e.g., test parameters vs.
method performance)
• Most important to determine a method’s robustness
• Typically implemented through simultaneously changing two
or more parameters, reducing the number of experiments
• Facilitated through availability of software, e.g., Design
Expert (www.statease.com), Minitab (www.minitab.com)

Ludwig Huber
© Copyright Ludwig Huber - LabCompliance Slide 26
 Simple DoE Example for HPLC Method
- Impact on Selectivity -
%
Run # pH Col. Temp Column Flow Rate
Org.Phase
1 -1 -1 +1 +1 +1

2 -1 +1 +1 -1 +1

3 -1 +1 -1 +1 +1

4 +1 -1 +1 -1 -1

5 +1 -1 -1 +1 -1

6 +1 +1 -1 -1 -1

-1 = 40 % ACN 4.0 25 ºC 10 cm 2.0 ml/min

60% ACN
+1 = 6.0 40 ºC 20 cm 2.5 ml/min
40% Water
FDA: Need sufficient statistical power to support analytical “Design Space”
© Copyright Ludwig Huber - LabCompliance Slide 27
 Validate the Method for Intended Use

• Formally validate the method following ICH Q2

• Develop a method qualification plan
• Assure that equipment is formally qualified
(specifically spelled out in the new FDA guide)
• Assure that personnel is formally trained
• Perform qualification experiments, including robustness
testing
• Evaluate data and document results
• Write a validation report

© Copyright Ludwig Huber - LabCompliance Slide 28

 Examples for HPLC Robustness Testing
• Deliberately change critical operational limits and evaluate
impact on performance: precision, accuracy
• Include sample preparation and testing parameters
Sample preparation (accuracy)
• Extraction time (-20% of target)
• Extraction temperature (± 5 ºC)
HPLC
• Col Temperature (± 3 ºC)
• Mobile phase composition (± 2%)
• Buffer concentration (± 2%)
• Flow rate (± 0.3 mL/min)
• Detection wavelength (± 1 nm)
• Column Lot (quality, selectivity) Define acceptable ranges !
Ambient temperature/humidity
Stability of samples, standards
© Copyright Ludwig Huber - LabCompliance Slide 29
 Examples for Acceptance Criteria
Quantitative Impurities in Finished Drugs
Parameter Test

Accuracy 90 – 110%, 80 – 120% at specifications limit

Precision
Repeatability <4 % RSD (up to 15% at LOQ)
Intermediate <5.0 % RSD (higher at LOQ)
Reproducibility < 6% RSD (higher at LOQ)
Peak resolution >1.5 (related substances)
Specificity or >2 (main peak)
Peak purity check with UV DAD or MS
Limit of Detection N/A

Limit of Quantitation 0.05%

Linearity visual inspection of linearity curve, r>0.9900

Range o.k. if accuracy, precision, linearity criteria are met

© Copyright Ludwig Huber - LabCompliance Slide 30

 Example: Report Summary Table
Validation
Measure Acceptance criteria Results
Parameter
Recovery – Conc1 97 – 103 % 99%
Accuracy Recovery – Conc2 97 – 103 % 100%
Recovery – Conc3 97 – 103 % 100%
Method
RSD ≤ 1.5 % 0.4%
Precision
Intermediate
RSD ≤ 2.0 % 0.8%
Precision
Specificity Peak Resolution Factor R R for all peaks >1.5 All peaks >2.0
Correlation Coefficient ≥ 0.9900 0.9900
Linearity
Visual inspection of plot Linear response plot Shows linearity
Correlation Coefficient ≥ 0.9900 0.9900
Range Precision at 3 concentrations ≤ 1.5 % <1%
Recovery at 3 Conc. 97 – 103% 99.6%

Column Temp. ±2 C R for all peaks >1.5 R for all peaks >2.0
Mobile Phase ±2 % R for all peaks >1.5 R for all l peaks >2.0
Robustness
Sample extraction time -20 % Recovery in spec. Recovery in spec
Compound stability 6 days <3% degradation <2% degradation

© Copyright Ludwig Huber - LabCompliance Slide 31

 Assure that the Method Remains in a
State of Control
• Run system suitability tests
– Select critical test parameters based on risk
assessment and design space experiments
• Track quality control sample test results
• Thoroughly look at OOS results, and if method
specific, implement a corrective action plan
• Apply rigorous change control procedures

© Copyright Ludwig Huber - LabCompliance Slide 32

 Change Control
• Follow change control procedure
• Assess the impact of each change and
perform risk assessment
• Take advantage of knowledge gained during
robustness testing
• Evaluate if the method parameter change is
within the defined and tested design space
and boundaries (method operational design
ranges)
• If not may have to revalidate the method

© Copyright Ludwig Huber - LabCompliance Slide 33

 Coninually monitor and improve the
Method
• Actively collect inputs from operators on reliability and performance of
the method
• Evaluate customer complaints
• Conduct regular method review, e.g., yearly
• Track and trend system suitability
• Respond to adverse trends before they become problems
• Continually improvement through
– Problem solving and corrective action
– Preventive action
– Verification of correctve and preventive actions

© Copyright Ludwig Huber - LabCompliance Slide 34

 Benefits of QbD for Laboratories
- Example: Analytical Method Validation -

• Facilitates technology innovation (new technology can be

used without FDA re-approval, as long as the Analytical
Target Profile (ATP) is the same (future thinking)
• Technology changes can be implemented without loss of
time – facilitates continuous improvement
• Less analytical method related Out-of-Specification and
failure investigations
• Lower failure rates for method transfer
• Allowed method changes without revalidation well defined
through design space and robustness testing

Ludwig Huber
© Copyright Ludwig Huber - LabCompliance Slide 35
 Industry Barriers to QbD
• Current guidelines not in line with QbD approaches
• Registration currently not based on method performance but
on method conditions
• Low motivation to change
• Only little experience in the industry
• Requires new tools and skills for analysts
• Implementation Challenges
– Collaboration between functions
– Experience with new concepts
– Workload and resource limitations

Ludwig Huber
© Copyright Ludwig Huber - LabCompliance Slide 36
 FDA-EMA Collaborative Research on
QbD for Analytical Methods: Q&As
Question
• What are the Agencies’ views with respect to the use of
analytical target profile (ATP) for analytical methods?

Answer
• In general, an analytical process profile (ATP) can be
acceptable as a qualifier of the expected method performance
by analogy to the QTPP as defined in ICH Q8 (R2).
• However, the Agencies would not consider analytical
methods that have different principles (e.g.,HPLC to NIR)
equivalent solely on the basis of conformance with the ATP. An
applicant should not switch between methods without
appropriate regulatory submission and approval

© Copyright Ludwig Huber - LabCompliance Slide 37

 FDA-EMA Collaborative Research on
QbD for Analytical Methods: Q&As
Question
• What are the Agencies expectations in regulatory submissions for
Method Operational Design Ranges (MODR)?
Answer
• For example, data to support an MODR could include: (a)
appropriately chosen experimental protocols to support the
proposed operating ranges/ conditions; and (b) demonstration of
statistical confidence throughout the MODR.
• Issues for further reflection include the assessment of validation
requirements as identified in ICH Q2(R1) throughout the MODR
and confirmation of system suitability across all areas of the
MODR

© Copyright Ludwig Huber - LabCompliance Slide 38