My Chemistry

“Th l f“The role of assessments in method development”

- A knowledge based approach -

11th ACD/Labs user meeting 8-9 June 2010

Tom van Wijk MsC

Method development and validation departmentMethod development and validation department

Abbott Healthcare products BV

(formerly Solvay Pharmaceuticals)

Weesp, the Netherlands

Requirements for stability indicating HPLC methods for assay and purity

• Good specificity

y y

• Availability column

• One method for DS and DP

• Robust separation

• MS compatible

• Common technique

• Linear, accurate and precise

• Transferable

• Low maintenance

• Short runtimes

• Early or late phase method

• Simple methods

• Fast development

• other• Simple methods

• Minimize number of methods

• other………………..

2Company Confidential© 2010 Abbott

“ ACD usermeeting”“09-junr-2010

Choices during development

• Quality of the method depends on estimations made during development of the methodduring development of the method.– Selection of components– Choice of separation and detection techniqueChoice of separation and detection technique

• How to make sure to make

the right decisions?– Think in advance– Know your shortcomings– Learn from mistakes

“L ki f th i ht t ?!”



“Looking for the right components?!”

Effective development

HardwareInformationInformation

SoftwareEffective development:

b t bi ti f 3 f t4Company Confidential

© 2010 Abbott

best combination of 3 factors“ ACD usermeeting”“09-junr-2010

Hardware and software

• Suitable instruments and methods for impurity profilingHPLC– HPLC

– (HS)GC– (chiral) CE– Other…

• Hyphenation with MS– Correlation between disciplines, techniques or methods

• Use of ACD/AutoChrom software as a tool to allow automated HPLC method screeningautomated HPLC method screening

• Hardware and software covered, but what about the information?


information?


Strategy

Method development

• Method development strategyMethod development

Additional method

ACD Autochrom d t b

Select screening method

HS GCRP HPLC

databaseColumn and pH screening

Select best

Modifier screeningACD/AutoChrom

developmentSelect best

Temperature and gradient Optimization

Select best

development

OK?Real time Method

evaluation

Select best

Test optimum


Application method for release and IPC

Method Optimization


Strategy

• Critical items:– Can we fully rely on hardware and software?Ca e u y e y o a d a e a d so t a e– Choice of technique(s)– Selection of components “needles in the haystack”– Impurity levels 0.05% relative to nominal level active compound

• Perform screening study prior to method development

“Screening phase”component screening study

ChemicalStress tests

Pharmaceutical stress tests

Impurities of synthesis

ScreeningMethods y

Select/pool Samples

Choice of relevant componentsDatabase


Select/pool Samples


StrategyMethod development

• Traditional approach with typical drawbacks:– Selection of relevant components without prior knowledgeSelection of relevant components without prior knowledge– How to handle unknowns during development– ‘Complete’ coverage of impurities?– no compound knowledge available at the analytical work floor

• Solution: knowledge based approach

“Impurity profiling assessments”Create an overview of potential impurities



“Analytical” questionsInformation

• Reports on the analysis of the substance under study (think also of methods used in Discovery)of methods used in Discovery)

– Salt screening– Solubility (water, pH range, organic solvents)– Stability (water, organic solvents, pH range, heat, light, moisture)– pKa, Log P, Log DR t th l i f d ith i il h i t• Reports on the analysis of compound with similar chemistry, e.g. for common degradation patterns or elution behavior.

• Can anyone predict the degradation patterns (consult in anyCan anyone predict the degradation patterns (consult in any case the project chemist), include interaction with excipients

• Use predictive tools to estimate missing properties (ACD/Labs S S ) f


Log D Sol Suite), e.g. for known and potential impurities“ ACD usermeeting”“09-junr-2010

More information…..?Information

• If there are no reports as questioned above, has unreported work been done and, if yes, who are the persons involved?work been done and, if yes, who are the persons involved? (might yield "unofficial" documentation or unwritten "tacit" knowledge)

• Often is more information available:

– knowledge of people (in the mind)d f diff t– used for different purposes

– not per definition shared– not aware of interest to other peoplep p



More information…..?Information

• The impurity landscape may look different for different disciplines

compound A Compound Z

Change route?disciplines.

yieldKgup-scaling

Particle size

Salt selection

Compound B

Compound X

Change route?

Process development

metabolites

d d ti d timpuritiesformulation

Starting materials

yieldKg

Compound YAdapt method?

degradation products

homogeneity

stabilityChange formulation?

NMRLCMSavailability

Out sourcing

NMRLCMS


gintermediates


Knowledge based approachAssessment

• Share knowledge: perform assessments in multi disciplined teams: O -

– Synthesis– Degradation

MetabolismN

Cl

C

N + OH

O– Metabolism– Excipients

• 30-50 potential C

OH

ON OH

componentsN

COH O

COH

Method development is a multi disciplined Activity,

O


not an analytical task!“ ACD usermeeting”“09-junr-2010

Preparation phaseCase study

• Which components selected to develop a new method? 1. N

COH

OC

N

COH

2

1. Main compound

O

C

O -

N + OH

2.

3.

2. Metabolite

3. Degradation product

4 Impurity 1 OH

CON

4.

4. Impurity 1

5. Impurity 2

6. Impurity 3

COH

OCl

5.

6.

13Company Confidential



• Initial

– Create a list of theoretical components includeCreate a list of theoretical components, include• Origin, Probability• UV chromophore, pKa, mass

– Estimate suitable detection method• Choice of ionization source and mode

– Propose technique(s) of analysisPropose technique(s) of analysis• HPLC / GC• But also, e.g. pH of separation

– Enhances the chance of detecting unknowns– Estimation: 80-90% of components can be predicted –

relevant unknowns need to be screened for!


relevant unknowns need to be screened for!



component Origin : probability Information

Create a list of potential components + add information

component Origin : probability (H/M/L)

Information

No structure synthesis Degradation UV pKa Mw Source/mode

1 H Y 8 400 ESI N OHpos/neg

2 L H Y 8 416 ESI pos/neg

3 L M Y 8 5 388 ESI neg

OC

C

O -

N + OH

O

3 L M Y 8.5 388 ESI neg

4 M Y n.a. 201 APCI pos/neg?

NC

OHO

Cl

5 L L Y 7 196 ESI pos

6 H L N 2 204 ESI neg

N

COH

O


O




Selected technique

Estimated technique per component: before start screening

(H/M/L)

No structure synthesis Degradation GC LC

1 H Y L HN OH1 H Y L H

2 L H Y L HMethod 1H 3 10

OC

C

O -

N + OH

O

3 L M Y L M

4 M Y H L

pH 3 or 10

Method 2

NC

OHO

Cl

5 L L N M H

6 H L Y L LC

OH

N

Polar,No UV


CO

No UVMethod 3?



Compounds selected for HPLC development: before start screening

1.O

N

COH

N OH2.ON

COH

OO -

N + OH

COH

CO3.

Create method based on this selection; e.g. wavelength UV, MS tune Increases the chance of detection unpredicted components!

O5.

Increases the chance of detection unpredicted components!



Preparation phase (overview)

Impurity Profiling Assessment

DegradationGenotox

List of components information Degradation

AssessmentGenotox Assessment



Preparation phase (overview)

• Preparation phase

– Make sure estimated Collect

infoPreparation

phasecomponents can be detected in the screening

TheoreticalAssessment

Screeningphase

– Provisional component selectionphase

– Sample selection– Provisional technique

selectionselection



Screening phase

• Write and approve (combined) component screening & method development protocolmethod development protocol

– Forced degradation ; create the impurity• acid/base/oxidation/reduction/temp/humidity/lightp y g• compatibility samples

– Collect/prepare standards (capacity chemist!)C ll t l hi h ld t i th i it– Collect samples which could contain the impurity

• synthetic impurities• stability samples DS *y p• stability and/or compatibility samples DP *

* if available; aid to estimate relevant degradation products20Company Confidential

© 2010 Abbott

* if available; aid to estimate relevant degradation products“ ACD usermeeting”“09-junr-2010

Screening phase

• Start component screening– Start forced degradation experiments

• Develop screening methods, include– Feasibility technique(s)– Tune instrument, ionization source and mode on components

with different characteristics• Analyze samples (20 - 40)

– Orthogonal systemst h i• techniques

• HPLC - 2 pH’s



Screening phase

• Evaluate results– Team (project / technical)Team (project / technical)– Compare with prediction– Determine relevancy of (newly found) components

S l t t ( d ti t l ti )– Select components (and motivate selection)



Screening phaseCase study


Selected technique

Choice of technique per compound

(H/M/L)

No structure synthesis Degradation GC LC

1 H Y L HN OH1 H Y L H

2 L H Y L H

Method 1pH 3 or 10

OC

C

O -

N + OH

O

3 L L N L M

4 L M Y L H

NC

OHO

N Newly found

5 L L N M H

6 H L Y L L

Method 2

COH Method 3

Cl

found


CO

Method 3


Screening phase

• Update impurity profiling assessment– Select components for method development– Select components for method development– Estimation: 90-95% of components can be screened for –

unknowns still need to be monitored!

• Determine suitable development approach• Strategy

f th d• nr of methods• Time available• Life time of the method• personal favorite



Screening phase (overview)

Screening phase– Relevant component

l ti

Collectinfo

Preparationphase

selection– Technique selection– Select samples

TheoreticalAssessment

Screeningphase

– Pool samples (development on 4-6 samples)

protocol

phaseDevelopment phase

– Make sure selected components can be

ProfilingAssessment

data

Developmentphase

components can be separated and detected

– Comply with validation requirements


q


Development phase

• Need for software

1 Detect relevant peaks in samples e g the known or1. Detect relevant peaks in samples e.g. the known or expected components but also unknown impurities (estimated about 10-20%)

2. Correlate relevant peaks between different chromatographic systems

3. Evaluate optimum system



Experience with ACD/AutoChrom

1. Detect relevant peaks in samples e.g. the known or expected components but also unknown impurities• Use of different MS ionization modes is possible, but:

• Generation of huge datasets (processing time)

• No MS matching between positive and negative mode.

• Would enhance detectability of components (see assessments)

• IntelliXtract is suitable in finding peaks of interest but:• IntelliXtract is suitable in finding peaks of interest, but:

• Huge amount of potential components

• Relevant low level/response components can easily be lost during p p y gprocessing

• Known information cannot be used

W ld i ifi tl i


• Would significantly ease up processing


Development phase Missed components only found in

negative mode

• Detect relevant peaks in samplesSelected relevant

t

negative mode

Components foundUnknown relevant

t i iti d components found components in positive mode




2. Correlate relevant peaks between different chromatographic systems

• UV and MS spectra of selected impurities are known prior development (often at different pH)

component screening study Chemical

Stress testsPharmaceutical

stress testsImpurities of

synthesis

Choice of relevant components

ScreeningMethods

Select/pool Samples

Choice of relevant componentsDatabase

• Create a database to be used in autochrom development

• Match factor enhances confidence in proper correlation of component


p p p


Experience with ACD autochrom

3. Evaluate optimum system• Software can be used as a tool but not as a black box applicationSo t a e ca be used as a too but ot as a b ac bo app cat o

• Optimum found depends on the quality of the processing, e.g. correct evaluation only if all (relevant) peaks are (correctly) assigned

• Include weight factors

e.g. for critical separation of impurities with the main compound.

Q i k l ti• Quick evaluation:

UV Overlay of the different subsamples for visual comparison, including zoom in/out




• Effective method development using a screening approach requires software for effective processing, combining both UV and MS data.

• ACD/AutoChrom offers a high potential, but needs i t t b ffi i timprovement to be more efficient

• Use of prior knowledge

• Use of MS capabilities

• Better visual overviews

• Development of a high quality stability indicating method within 2 weeks is shown to be feasible (excluding assessments and compound screening)


assessments and compound screening)“ ACD usermeeting”“09-junr-2010

Acknowledgements

• Norbert Lammers

• Piet Hoogkamerg

• Harm Niederländer

• Andre van Kempen

• Coen Bijl

• Leon van den Bos




My Chemistry

Documents