quality bydesign (qbd) solutions foranalytical ... bydesign (qbd) solutions foranalytical...

Quality by Design (QbD) Solutions for Analytical Method Development

Andreas Tei

Pharmaceutical Segment Manager

A systematic approach to reducing

variability

2

31

Target System

Emulation by ISET

Robustness study by

QbD software

1290

R&D

1260

QA/QC

Method transfer

Method development

System

Use of 1.8 µ particles

and QbD software

Target Systems in

QA/QC labs

QbD Method Development & Method Transfer WorkflowFrom UHPLC to HPLC in a nutshell

2

1290

R&D

1260

R&D

Emulation

Content

• Introduction

- Traditional approach of method development and transfer

- QbD approach for method development

• Agilent solutions for method development-Agilent method development systems

-Intelligent system emulation technology (ISET)

-Method scouting wizard software

-Third party QbD software

• A practical approach of method development under QbD principles- Screening

- Optimization

- Robustness study, Design of Experiments

- Transfer & Verification

Method Development & Transfer WorkflowsA Classic Approach

4

� Parameters determined by trial-and error.

� Robustness tests performed by the one

factor at the time approach (OFAT)

Develop Validate Transfer

R & DQA / QC

Variability of critical methodattributes (resolution, tailing, etc)

Small changes in pH, temperatureor flow rate shows a large effect

OFAT Approach: Method Transfer is a balancing act

10% of analytical methods are discarded per year to avoid high revalidation costs

after the methdod showed instabilities on the QC system

Development

System

Target QA/QC

Systems

Even buffer solutions from different

operators deliver different results

Different results on different systems

Quality by Design - A systematic approach to obtain a“consistent quality” output

System Variabilities

Fixed process

Variable output

System Variabilities

Flexible

process

Consistent output

Quality by

Design

Understand variability

Continuous improvement

Control Variability

ICH Q8-Q11

6

� http://www.fda.gov/downloads/Drugs/Guidances/ucm073507.pdf

7

More Robustness by Applying QbD Principles

�Risk assessment process

� Define Critical Method Attributes CMAs

(resolution, peak tailing, etc.)

� Define critical method parameters CMPs

(flow, temp) which affects the CMAs

�Perform a Design of Experiment (DoE) study by

modifying multiple CMAs in simultaneously

and create a Design Space

8

� Key is the statistical „Design of Experiment“

(DoE) where multiple CMPs will be varied in

each experiment

� A method is robust, as long all changes of CMAs

are within the determined Design Space

QbD Approach for Method Development

Creating a Design SpaceOptimizing through One Variable at a Time vs. DoE

One Variable at a Time

Variable 1

Variable

2

Impact of variables 1 and 2 (e.g. % organic mobile phase and column temperature)

on Critical Method Attributes (CMAs) e.g. peak resolution or %recovery

9

Injection Volume (µl)0.6 3.0

% O

rgan

ic m

obile

phas

e (m

etha

nol)

20

10

Flow Rate (ml/min)

1.6

0.8

Column Temp. (oC)30 40

10

What Are The Advantages ?

� The relationship between the different critical method

parameters (CMPs) will be described and visualized in surface

or contour plots

� Increase of efficiency by avoiding unnecessary experiments

during the method development process (no trial an error!)

� DoE is delivering robust methods compared to the classic

„one-factor-at-the-time“ (OFAT) robustness testing process

Content

• Introduction

- Traditional approach of method development and transfer

- QbD approach for method development

• Agilent solutions for method development-Agilent method development systems

-Intelligent system emulation technology (ISET)

-Method scouting wizard software

-Third party QbD software

• A practical approach of method development under QbD principles- Screening

- Optimization

- Robustness study, Design of Experiments

- Transfer & Verification

Agilent Solutions for QbD Method Development

ISET

Intelligent System Emulation Technology

RA

Remote Advisor

Harmonized Qualification

ACE

Method Development

System & Method Scouting SW

Rapid Development of Robust New Methods Agilent 1290 Infinity II Series Method Development Solution

13

Agilent 1290 Infinity II Method Development Solution

6100 Series MSD and 1290 ELSD are optional detectors to

ensure the most comprehensive compound detection

1290 Infinity II Series Method Development SolutionSchematics

14

8

x

12 + 1

x

12 + 1

= 169

= 1352

1290 Infinity II

MCT

1290 Infinity II

Flexible Pump

External Solvent

Selection Valves

Up to 12 solvents

per SSV!!!

A2A1 B2B1

SSV2SSV1

� 1352 different combinations of column chemistries and eluents

� A nearly infinite number of separation conditions is created by including

different temperature and flow rates as variable parameters

Agilent ChemStation Method Scouting WizardThere is no easier way to set up even complex screening campaigns

15

• Define project

Choose scouting combinations and

base method.

• Select columns

All installed columns are shown

automatically.

• Select solvents

Pump types and valves are

automatically detected.

• Define gradients

Select between different gradients and

temperatures.

• Review and select screening

methods

Check for incompatible combinations.

• Set up samples

Define injection volumes and number of

repetitions. Easy and fastsetup of sequences !

Agilent ChemStation Method Scouting WizardEase of Use: Screening Report

16

Fast screening of mobile and stationary phases with 1290 Infinity II LC –

Analysis of Degradation Products of Metoprolol

Max. R

ete

nti

on

Tim

e

Injections

Bubble size represents the number of integrated peaks and,

consequently, best mobile and stationary phase

combination.

https://www.agilent.com/cs/library/applications/5991-0989EN.pdf

Intelligent System Emulation Technology (ISET)

- Seamless transfer of methods between LCs, regardless of the brand

PO-Nr. G2197AA

Method Transfer Between Different LC Instruments

min3 4 5 6 7 8 9 10 11

mAU

0

25

50

75

100

125

150

175

1100 Series Binary System

1290 Infinity LC

Method transfer from a UHPLC system with a minimized dwell

volume and optimized mixing behavior to any other LC system is

often challenging and affects rentention time and resolution

Example: Method Transfer from a 1290 UHPLC to a 1100 HPLC system

19

Practical aspects how to measure dwell volumes, transfer & optimize methods

20

How To Measure Dwell Volumes

� Add a UV active tracer to solvent B (often acetone)

� Remove the column, fit a restriction capillary

� Run a step gradient from 10%B to 90%B

� Calculate the dwell volume by the delay between

UV response and the step with respect to the flow rate

21

1290 Infinity LC

min0 1 2 3 4 5 6 7 8 9

mAU

0

100

200

300

400

500

600

1100 Series Quaternary LC

Gradient differences due to different dwell

volumes and different mixing behavior

Method Transfer Between Different LC Instruments

Results from a tracer experiment

min0 1 2 3 4 5 6 7 8 9

mAU

0

100

200

300

400

1290 Infinity LC

1 min isocratic hold


22

Method Transfer Between Different LC InstrumentsApproach # 1: Isocratic holding step to synchronize


Still gradient differences due to

different mixing behavior

1260 Infinity LC

1290 Infinity LC

+ 900 µl hold

23

Approach # 1: Applying Isocratic Holding Steps

Results

• Results shows a low consitency

• Requires manual determination of the dwell volume/ isocratic hold

(in solvent delivery systems equipped with dampeners the dwell volume is pressure

dependent and variable)

• Requires modification of the methods

(should be avoided in validated environment,

but doesn’t require revalidation USP Chapter <621>)

min0 1 2 3 4 5 6 7 8 9

mAU

0

100

200

300

400

1290 Infinity LC

1 mL loop installed


24

Method Transfer Between Different LC InstrumentsApproach # 2: Adding a physical void volume

Almost consistency of both

gradient curves


1290 + 1000 µL dwell vol.

1100/1200 Quat Pump

Approach # 2: Adding a Physical Void Volume

Results

25

• Results shows a good consitency

• Manual determination of dwell volumes required (issues of a variable

dwell volume with systems containing damperners)

• All mechanical changes are laboriously not flexible

Agilent Solution:

Intelligent System Emulation Technology (ISET)

26

Programmed gradient

1290 gradient

1200 gradient

0 0.5 1 1.5 2 2.5 3 3.5

0

50

100

150

200

250

300

350

400

min

mAU

InjectionSoftware controlled compensation

of dwell volume differences and

synchronization of mixing

behaviors

Agilent Solution: Method Transfer by ISETResults: 1260 Infinity Binary LC to 1290 Infinity LC

27

min1 2 3 4 5 6 7 8 9

mAU

0

10

20

30

40

Imp

A

Imp

B

Imp

H Imp

J

Imp

F

Imp

K

Pa

race

tam

ol

1290 Infinity LC with ISET

1260 Infinity Binary LC

• Consistency of results

28

• Fast screening of mobile and stationary phases with the Agilent 1290 Infinity LC

and seamless method transfer to an Agilent 1200 Series LC using ISET

Agilent Application Note 5991-0989EN

• Developing faster methods for generic drugs within USP <621>allowed limits


• Effective use of pharmacopeia guidelines to reduce cost of

chromatographic analysis


• Developing faster methods for generic drugs within EP 2.2.46E allowed limits


Agilent Application Notes

Method Transfer by ISET

3rd Party Add-On QbD Software

AutoChrom

ChromSwordAuto

Turn your LC into an automated

QbD method development system!

Only Agilent‘s Method Development Solution is supported by all

three big manufactures of automated method optimization software!

Fusion QbD

29

Chemstation

4.5

pH

Oven Temp

Initial %

Organic

5.0

15.0

30.0 60.0

30

2.5

3.5

Fusion QbD Software (S-Matrix)

�Statistical approach identifying the best separation conditions

�A graphic plot illustrates the position of the best conditions

Sequence of Experiments 3 Dimensional Results Plot

Creating a Design SpaceSimple DoE Example for HPLC Method

Run # % Org.Phase pH Col. Temp Column Flow Rate

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

Level -1 =40 % ACN

-4.0 25 ºC 10 cm 2.0 ml/min

Level +1 =60% ACN

40% Water6.0 40 ºC 20 cm 2.5 ml/min

FDA: Need sufficient statistical power to support analytical “Design Space”

31

CMPs: Impact on Selectivity, Five Variables Two Levels

List of CMPs and CMAs for robustness

testing

* The coded names are used in robustness model displays

Creating a Design Space (Fusion QbD Software)Robustness Testing & Establishing a Design Space (MODR)

Design Space

32

33

Chromsword Software

• Interactive

• Sample profiling

• Impurities search

• Full factorial

• Statistical DOE

• Interactive

• Find workable method ASAP

• Method scouting

• User defined

• Statistical DOE

ScreeningRapid

optimization

Fine Optimization

Robustness Studies

34

Details of Chromsword Software

• Method development and optimization based on

the solvatic model of RP-LC

• Building retention models after processing results

by applying different variables as listed below

• Organic modifier concentration (C)

• Gradient time (Grad)

• Temperature (T)

• pH

• Multi variables: C+pH; Grad +T; C1 +C2

35

min0.5 1 1.5 2 2.5 3 3.5 4 4.5

Norm.

0

100

200

300

400

500

600

1.970

2.049

2.820

3.070

min1 2 3 4 5 6 7 8 9

Norm.

0

20

40

60

80

100

1.136

1.384

3.232


• Automated resolution optimization of critical peak pairs

36


Design space creation after robustness study is fundamental to the

Quality by Design approach

37

4. Modeling

Temperature

Gradient

ACD/AutoChrom Software

Start

Column, Buffer &Solvent Screening

Select Best

Optimize Gradient & Temperature

End

1. Data Collection

2. Peak Tracking

5. Next

Experiment

Report

3. Interpretation3. Interpretation

Content

•Introduction

– Traditional approach of method development and transfer

– QbD approach for method development

• Agilent solutions for method development–Agilent method development systems

–Intelligent system emulation technology (ISET)

–Method scouting wizard software

–Third party QbD software

• A practical approach of method development under QbDprinciples

– Screening

– Optimization

– Robustness study, Design of Experiments

– Transfer & Verification

39

31

Target System

Emulation by ISET

Robustness study by

QbD software

1290

R&D

1260QA/QC

Method transfer

Method development

System

Use of 1.8 µ particles

and QbD software

Target Systems in

QA/QC labs

QbD Method Development & Method Transfer WorkflowFrom UHPLC to HPLC in a nutshell

2

1290

R&D

1260

R&DEmulation

40

Overall workflow which

consists of four main steps

namely

• Step # 1: Screening

• Step # 2: Optimization

Screening

• Column Chemistry

• pH conditions

• Organic Solvent Selection

• Gradients, temperatures

Agilent Method

Scouting

WizardOptimization

• Optimize Gradient Profiles

• pH conditions

• Flow rates, temperaturesQbD Software

QbD Based Method Development Workflow

min10 10.2 10.4

998 Set by User

| || |' ' ' ' '

+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

min10.5 10.55 10.6 10.65 10.7

998 Set by User

| || |' ' ' ' '

++++++++++++++------+++++++++----------+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

min2.5 5 7.5 10 12.5 15 17.5 20

mAU

0

50

100

150

200

250

300

350

10.543

min2.5 5 7.5 10 12.5 15 17.5 20

mAU

0

50

100

150

200

250

300

350

10.007

OptimizationResults: peak purity and separation after optimization 99.8%

99.8 % purity

Before

Optimization

After

Optimization

41

Purity< 90%

42

Overall workflow which

consists of four main steps

namely

• Step #1: Screening

• Step #2: Optimization

• Step #3: DOE & Emulate

Screening


• pH conditions



Agilent Method

Scouting

WizardOptimization


• pH conditions

• Flow rates, temperaturesQbD Software

QbD Software

Design of Experiments

• Multivariate study

• Robust region

• Design Space


43

Target System

Emulation by ISET

Robustness study by

QbD software

Method transfer

Method Transfer & VerificationFrom UHPLC to HPLC

2

1290

R&D

1260

R&DEmulation

Method Transfer From UHPLC To HPLC

44

Crawford Scientific Calculator

http://www.crawfordscientific.com/

HPLC_Method_Transfer_On-line.htm

http://www.unige.ch/sciences/pharm/fanal/lcap/telechargement.htm

HPLC Calculator University of Geneva

45

HPLC design space with new CMA

Robustness StudyModeling a HPLC design space on an emulated 1260 system

Critical Method

Parameters (CMPs)

Proven

Acceptable

Range (PARs)

Critical Method

Attributes (CMAs)

Column: Agilent ZORBAX

Eclipse Plus C8 4.6X150

mm, 3.5 µm

No. of peaks (>40)

API resolution (>4)

Peak purity (≥ 98%)

Peak tailing (<1.3)

Strong solvent: Methanol

% Strong solvent: 87.5.% ± 1.5%

Aqueous solvent pH: 7.7 ± 0.1

Gradient range: 5% to

87.5%

Oven Temperature: 37°C

Gradient time: 45 min

Flow rate: 1.4 mL/min

Wavelength: 292 nm

HPLC design space parameters

46

Proof Of RobustnessConditions applied from center point and the four corner points of the Design Space

min10 20 30 40 50

mAU

050

150

250

350

min10 20 30 40 50

mAU

050

150

250

350

min10 20 30 40 50

mAU

050

150

250

350

min10 20 30 40 50

mAU

050

150

250

350

min10 20 30 40 50

mAU

050

150

250

350

% B max: 86 %; pH: 7.6

% B max: 89 %; pH: 7.6

% B max: 87.5 %; pH: 7.7

% B max: 89 %; pH: 7.8

% B max: 86 %; pH: 7.8

A

B

D

T

C

API Rs> 4 and API tailing = 1.2 for all runs

Critical Method Attributes remain within the limits

47

Overall workflow which has

four main steps namely

•Step #1: Screening

•Step #2: Optimization

•Step #3: DOE & Emulate

•Step #4: Verify

Screening


• pH conditions



Agilent Method

Scouting

WizardOptimization & DOE


• DOE

• Creating a Design SpaceQbD Software

ISETQbD Software

Emulate Target System

• Transfer gradient profile

• Emulate target system

• DOE

• Creating a Design Space


Target System

Verify on Target System

• Apply Method

• Verify Results

• Validate Results

Verification Of The Final Method With The Target SystemResults: 1260 Infinity data compared to 1290 Infinity data in emulation mode

48

Final gradient

Time %B

0.3 5

45.6 87.5

49.2 87.5

49.5 95

49.8 95

50.1 5

53.1 5

mAU

24.091

API Rs = 4.2

API tailing = 1.2

min10 20 30 40 50

0

100

200

300

400

23.344

1290 emulated as 1260

mAU

100

200

300

400

min10 20 30 40 50

API Rs = 4.1

API tailing = 1.2

0

23.347

24.115

1260 target system

Critical Method Attributes are within the defined limits

Conclusions

49

�Agilent Method Scouting Wizard Software supports

automated method development workflows

�Significant time saving has been achieved by using sub-2-

micron columns for method development workflows

�ISET supports seamless method transfer processes to other

LC-systems

�Additional 3rd Party QbD software is used to create more

reliable and robust methods

Agilent Application Note

50

Title: QbD Based Method Development on an

Agilent 1290 Infinity UHPLC system Combined

with a Seamless Method Transfer to HPLC

Using Intelligent System Emulation Technology

Type: Application Note

Publication Number: 5991-5701EN

Pages: 8

Target segments: Pharmaceutical QA/QC

Language: English

Author: Vinayak A.K.

LitStation Availability: May 5, 2015, CPOD

Agilent Application Notes

QbD based Method Development

• Quality-by-Design Approach to Stability Indicating Method

Development for Linagliptin Drug Product

– Agilent Application Note 5991-3834EN

• Automated QbD Based Method Development and Validation of

Oxidative Degraded Atorvastatin


• Development of an UHPLC Method for Azithromycin Tablets Using

ChromSword Auto Software


• QbD Based Method Development on an Agilent 1290 Infinity UHPLC

system Combined with a Seamless Method Transfer to HPLC Using


- Agilent Application Note 5991-5701EN

51

Agilent Instrument Control Framework (ICF)Seamless Integration Of Innovative Technology

Empower

ICS/OIP Adapter

WATERS

Chromeleon

DDK Adapter

THERMO

Analyst

AB SCIEX I/F

AB SCIEX

LabSolution

Shimadzu I/F

SHIMADZU

• One consistent driver set developed by Agilent

• Developer support provided by Agilent

• Significantly less effort / faster time to market

ICF

Agilent

Instrument

RC.NET

API

Driver

ICF

Agilent

Instrument

RC.NET

API

Driver

ICF

Agilent

Instrument

RC.NET

API

Driver

ICF

Agilent

Instrument

RC.NET

API

Driver

Waters Empower 3: Ready for Agilent 1290 Infinity II Agilent LC and ICF Releases

Agilent LC Hardware

1290 Multisampler

LC Drivers 2.10

Time A

Agilent ICF Software

ICF A.02.03

LC Drivers 2.10

Time A

+ 6 weeks

ICF A.02.03 DU1

LC Drivers 2.11

Time B

+ 6 weeks

Waters Empower Software

Waters ICS 2.1 HF1 71600xxx

ICF A.02.03 DU1

LC Drivers 2.11

Oct 16th 2015

Work in Progress: Thermo Chromeleon Software

Chromeleon 7.2 SRx

ICF A.02.02/3

Agilent LC 2.10/2.11

1290 Infinity II LC

LC Drivers 2.11

Time B

Agilent LC Instrument Control in Empower 3Required software modules to support Agilent 1290 Infinity II

CDS and CDS specific

adapter from Waters(ICS= Instrument Control Software)

Agilent ICF Adapter

Waters Empower 3

Adapter: Waters ICS 2.1 HF1

Agilent ICF A.02.03 DU 1 HF2

Agilent

1290 LC

RC.NET

API

Core

Driver

A.02.11SP1 Agilent RC.Net Drivers

Agilent 1290 Infinity II under Empower 3How does it look like?

LC Status Interface

Method Interface

Agilent’s ICF ConceptPlease download our Technical NotesICF website: www.agilent.com/chem/icf

ICF Release Notes on agilent.com ICF Publications / Technical Notes

• ICF Technical Overview, Pub. No.: 5990-6504EN

• Using the Agilent Instrument Control Framework to control the Agilent 1220 Infinity LC system through

Waters Empower software, Pub. No. 5990-9399EN

• Using the Agilent Instrument Control Framework to control the Agilent 1260 Infinity LC through Waters

Empower software, Pub. No 5990-9092EN

• Using the Agilent Instrument Control Framework to control the Agilent 1290 Infinity LC through Waters

Empower software, Pub. No 5990-9093EN

• Using the Agilent Instrument Control Framework to control the Agilent 1260 Infinity Bio-inert Quaternary LC

through Waters Empower software, Pub. No 5990-9354EN.pdf

• Agilent 1290 Infinity LC with ISET under Waters Empower Control - Emulation of Agilent 1100 Series LC

5991-2275EN

• Operating the Agilent 1260 Infinity LC with Dionex Chromeleon software using Instrument Control

Framework, Instrument control and performance, Pub. No 5990-7232EN

• Operating the Agilent 1290 Infinity LC with Dionex Chromeleon software using Instrument Control

Framework, Instrument set up and performance Pub. No 5990-7215EN

*SPRS: Skills & Resources Planning System

57

THANK YOU

QbD process terminology Analytical QbD

terminology

Examples

Analytical Target Profile

(ATP)

Accurate quantitation of API

without interferences from

degradants

Quality Target Product

Profile (QTPP)

Quality Target Method Profile

(QTMP)

pKa, Log P, Solubility

Critical Process Parameters

(CPP)

Critical Method Parameters

(CMP)

Flow rate,

Temperature, pH

Critical Quality Attributes

(CQA)

Critical Method Attributes

(CMA)

Resolution, Peak

Tailing, Peak Capacity

Control Strategy pH ± 0.1; Wavelength

± 2 nm

QbD terminology in Method DevelopmentAppendix

Analytical QbD Terminology

59

• Fast screening of mobile and stationary phases with the Agilent 1290 Infinity LC

and seamless method transfer to an Agilent 1200 Series LC using ISET


• Developing faster methods for generic drugs within USP <621>allowed limits


• Effective use of pharmacopeia guidelines to reduce cost of

chromatographic analysis


• Developing faster methods for generic drugs within EP 2.2.46E allowed limits


Appendix: Agilent Application Notes

Method Transfer by ISET

Appendix : Agilent Application Notes

QbD based Method Development

• Quality-by-Design Approach to Stability Indicating Method

Development for Linagliptin Drug Product


• Automated QbD Based Method Development and Validation of

Oxidative Degraded Atorvastatin


• Development of an UHPLC Method for Azithromycin Tablets Using

ChromSword Auto Software


• QbD Based Method Development on an Agilent 1290 Infinity UHPLC

system Combined with a Seamless Method Transfer to HPLC Using


- Agilent Application Note 5991-5701EN

60

quality bydesign (qbd) solutions foranalytical ... bydesign (qbd) solutions foranalytical...

Documents