analytical quality by design: what does it look like? › › resource › ... · 3ich q8(r2)...

36
Analytical Quality by Design: What does it look like? Jeremy Springall PhD CASSS DC Discussion Group April 2019

Upload: others

Post on 31-May-2020

7 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

Analytical Quality by Design: What does it look like?

Jeremy Springall PhDCASSS DC Discussion Group April 2019

Page 2: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

Presentation Summary

What is AQbD?Case Study 1: Risk AssessmentsCase Study 2: Design of ExperimentsCase Study 3: Control Charting

Page 3: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

We ensure the following:• The product is of the desired quality to ensure it falls within the

range which has been demonstrated to be safe for patients• The product is consistent with what we have made before• Ensure the product is potent

This is very challenging and requires in depth method and product understanding as well as suitable control to ensure delivery of all the statements above

Ultimately this requires analytical methods that are well understood, robust and under control

Why are Analytical Scientists Important?

3

Page 4: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

AQbD It’s been a long road!

4

2007 2009 2010 2012 2013 2014 2015 2016 2017 2018 2019

USP Performance Based Monographs

EFPIA Pharm Tech AMQBD

Paper

PDA Workshop on implementing AQbD

EMA FDA Pilot Assessment of QbD including

AQbD

PDA Workshop Analytical Science

and AQBD

IQ Survey Pharm Tech

ICH Q12, Q2, Q14 and joint industry

working groups e.g. MHRA/BP

USP Stimuli Articles and workshops

IQ Industry White Paper

& EFPIA paper on Analytical

Procedure Lifecycle Management

Current Status & Opportunities

GSK Pharm Tech Paper

Adapted from P. Borman (GSK) USP AQbD workshop 2018

Page 5: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

What is Quality by Design (QbD)?

5

Quality Target Product Profile3

Critical Quality Attributes

Risk Assessment4

Design Space

Control Strategy5

Continued Process Verification

Analytical Target Profile

Critical Method Attributes

Risk Assessment

Method Operable Design Region (MODR)

Control Strategy

Continued Method Verification

Process Analytical Method1,2

Method Design and Development

Method Performance Qualification / Verification

Continuous Method Verification

1P. Nethercote et al. Pharm. Tech (2010), 34, 2; 52–592USP Stimuli to the Revision Process on Lifecycle Management of Analytical Procedures: Method Development, Procedure Performance Qualification and Procedure Performance Verification (2013) 3ICH Q8(R2) Pharmaceutical Development (2009)4ICH Q9 Quality Risk Management (2005)5ICH Q10 Pharmaceutical Quality System (2008)

Page 6: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

What is Analytical Quality by Design?

6

Analytical Target Profile (ATP)

Establish method operable design region

(MODR)

Perform Method

Development

Continued Monitoring

Verify MODR, validate

NOC

Design of Experiment (DoE)

Verification

Analytical Method

Page 7: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

The combination of all performance criteria required to ensure the measurement of a quality attribute/s is/are fit for its intended purpose and produces data which can be used with the required confidence to support specification pass/fail decisions.

This is analogous to the QTPP described in ICHQ8

Uses of the ATP1. Direct selection of analytical technology2. Support risk assessment and evaluation of variable to develop a full

understanding of how the parameters affect the reportable result3. Defines when method development is complete4. Directs continuous improvement of the method throughout lifecycle

Analytical Target Profile (ATP)

7

2Pharmaceutical Research and Manufacturers of America Analytical Technical Group, European Federation of Pharmaceutical Industries and Associations (EFPIA) Analytical Design Space Topic Team. Implications and opportunities of applying QbD principles to analytical measurements. Pharm Technol. 2010; 34(2):52–59342(5) Stimuli to the Revision Process: Analytical Target Profile: Structure and Application Throughout the Analytical Lifecycle, Barnett. K,L (2016)

Page 8: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

Tools for understanding the root causes of method variability and then driving the assessment of these identified parameters and their impact on the reportable results

Tools can include but not limited to:• Ishikawa Diagrams (Fishbone diagrams)• Failure mode and effect analysis

Univariate and multivariate methodologies can be used to assess the risk of the identified parameters

Risk Assessments

8

442(5) Stimuli to the Revision Process: Analytical Control Strategy, Kovacs. E (2016)5Borman, P., The Application of Quality by Design to Analytical Methods. Pharm Technol. 2007; 31(10):142‐152

Page 9: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

The combination of parameter ranges which have been evaluated to meet the ATP criteria. In other words, the MODR constitutes a region within which changes can be made without impact on the reportable results

Analogous to Design Space in ICHQ8

Once the MODR has been identified it needs to be verified and the normal operating conditions (NOCs) are the conditions and ranges to be stated in your SOP

Method Operable Design Region (MODR)

9

Page 10: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

The aim is to develop a planned set of controls to ensure a true result is being generated by the method. These might include a series of proactive and reactive controls:• Laboratory controls• Procedural controls • System Suitability criteria• Sample replicates

These method controls are to ensure the risk identified by the risk assessments is controlled and the method is performing as per the ATP criteria.

Method Control Strategy

10

Page 11: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

Case Study 1: Risk Assessments

Page 12: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

Risk Assessment workflow

12

Ishikawa Diagram

Failure Mode and Effect Analysis (FMEA)

Method Development

Risk Reduction

Page 13: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

Ishikawa Diagrams

13

• Start from prior knowledge• Use Ishikawa tools to classify the method parameters• It is important to capture all of the sources of possible method variation and the

classification is less important

Page 14: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

Failure mode effect analysis (FMEA) for Risk MitigationA tool for quantifying and prioritizing risk assessments in a process, product or system and then track the actions taken to mitigate the risk

14

Score Severity Score Probability

1 No impact on reportable result or method performance 1 No history of occurrence with this method or similar

methods or published data

3 No impact on reportable result; low impact on method performance 3 No history of occurrence with this method but

concerns from other similar methods or published data

5 Low impact on reportable result and/or medium impact on method performance 5 History of occasional occurrence with this method but

controls are now in place

7 Medium impact on reportable result and/or high impact on method performance 9 History of occurrence with this method and controls

may not be robust

9 High impact on reportable result or method performance or no information

𝑀𝑒𝑡ℎ𝑜𝑑 𝐷𝑒𝑣𝑒𝑙𝑜𝑝𝑚𝑒𝑛𝑡 𝑝𝑟𝑖𝑜𝑟𝑖𝑡𝑖𝑠𝑎𝑡𝑖𝑜𝑛 𝑆𝑒𝑣𝑒𝑟𝑖𝑡𝑦 𝑆 𝑋 𝑃𝑟𝑜𝑏𝑎𝑏𝑖𝑙𝑖𝑡𝑦 𝑃Risk value Table (Severity X Probability)

Effect Value Mitigation

Low X ≤ 12 Optional

Medium 12 < X < 40 Recommended to mitigate if possible

High ≥ 40 Must mitigate

1 3 5 7 91 1 3 5 7 93 3 9 15 21 275 5 15 25 35 459 9 27 45 63 81

Severity

Prob

ablility

Page 15: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

FMEA combined with Ishikawa Diagram

15

Prior observation, knowledge (from this or other projects) or

theoretical concern

Impact on commercial operation

(SEVERITY)

Justification for SEVERITY scoring

Probability of OCCURRENCE

Justification for Probability of OCCURRENCE scoring

Development Prioritization

Number (DPN) = S X O

Column temperature 7deviation from the set temperature

impacts the resolution of main peak to fragment peaks

3Control range is entered into the method to

ensure control from the instrument 21

Autosampler temperature 9if sample is not chilled the product can

under changes prior to injection 3Control range is entered into the method to

ensure control from the instrument 27

Injection volume 3injection volume can be adjusted to fit within the linear range of the method 5

requires analyst to adjust injection volume based on concentration, include SOP requirement for injection volume based on concentration with

example calculations

15

MPA Buffer composition 9

mispreprataion of mobile phase A can result in the incorrect amount of acid

being used and therefore will negatively impact the separation

5Must mitigate and understand the impact

further during method development 45

MPB Buffer composition 9

mispreprataion of mobile phase B can result in the incorrect amount of acid

being used and therefore will negatively impact the separation

5Must mitigate and understand the impact

further during method development 45

UV detection wavelength 3if the incorrect wavelength is used then the sensitivity of the method could be

comprimised3

Controlled by instrument method within a set range 9

We can then use the DPN to determine parameters for method development

Page 16: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

How the FMEA Drives Method Development

16

Prior observation, knowledge (from this or other projects) or

theoretical concern

Impact on commercial operation

(SEVERITY)

Justification for SEVERITY scoring

Probability of OCCURRENCE

Justification for Probability of OCCURRENCE scoring

Development Prioritization

Number (DPN) = S X O

Injection volume 3 injection volume can be adjusted to fit within the linear range of the method 5

requires analyst to adjust injection volume based on concentration, include SOP

requirement for injection volume based on concentration with example calculations

15

MPA Buffer composition 9Mis-preprataion of mobile phase A can result in the incorrect amount of acid

being used and therefore will negatively impact the separation

5 Must mitigate and understand the impact further during method development 45

MPB Buffer composition 9mispreprataion of mobile phase B can result in the incorrect amount of acid

being used and therefore will negatively impact the separation

5 Must mitigate and understand the impact further during method development 45

UV detection wavelength 3if the incorrect wavelength is used then the sensitivity of the method could be

comprimised3 Controlled by instrument method within a set

range 9

Our highest scoring parameters become the dimensions of our DoE for assessment• Lots of parameters use a Plackett Burman screening design• Few identified parameters use a fractional factorial design

Page 17: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

After method development, MODR generation and determination of the NOCs it is important to rescore the risk based on the new data.

We can then determine the types of controls required to ensure a robust and suitable state of the method

Re-scoring Risk

17

Ishikawa Diagram Section

Prior observation, knowledge (from this or other projects) or theoretical

concern

Impact on commercial operation

(SEVERITY)

Justification for SEVERITY scoring Probability of OCCURRENCE

Justification for Probability of OCCURRENCE scoring

Development Prioritization

Number (DPN) = S X O

Mitigation Plan after method developmentImpact on commercial

operation (SEVERITY)

Probability of OCCURRENCE

Development Prioritization Number after mitigation plan

(DPN) = S X O

Method Autosampler temperature 7no temp control could lead to sample degradation over the course of a

sequence5 Some molecules require temperature control

others do not 35 0

Method Capillary conditioning 7If the capillary is not in ideal

conditioning the separation might not occur optimally

5 Historical methods 35 0

Method Capillary balance 9 Sample does not enter detection window 3

Historical MEDI3902 method requires specific balancing of the capillary for

acceptable method performance27 0

Method Sample drawing 5Highly viscous solutions can be slow

to draw and can deliver lower amounts onto the cpaillry

3 Observed for sucrose containing mastermixes and using automated sample preparation 15 0

Method Focussing time 7 inadequate focussing time leads to non-completed separations 3 To be determined through experimentation 21 0

Method Buffer / sample composition 9Mastermix composition is directly

responsible for separation performance

9 To be determined through experimentation 81 0

Method Prepared sample stability 7 If a long sequence is being run the sample could degrade over time 3 To be determined through experimentation 21 0

Method Sequence 2Sufficient buffers are provided to

the system for a sequence, manufacturer demostrated

performance for 90+ injections3 Mitigated by the autosampler temperature 6 0

Original Risk Score

Mitigation Plan

Re-scoring Risk

Page 18: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

Case Study 2: Why use DoE for Method Development?

Page 19: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

General model for DOE

19

Controllable variablesx1 x2 …… xp

z1 z2 …… zp

Uncontrollable variables

Inputs (Factors)

Outputs (Responses)

Use of DOE to:• Evaluate effect of most influential parameters• Identify the interactions between parameters• Optimize the best operating conditions

“Design of experiments (DOE) is a test or series of tests in which purposeful changes are made to the input variables of a process so that we may observe and identify corresponding changes in the output response” from Douglas Montgomery – Introduction to statistical quality control

Page 20: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

Why use a DoE methodology for method development?

Adopting a DoE approach ensures a thorough assessment of risk and enables the generation of an appropriate mitigation strategy

OFAT finds a local optimum; DOE can find a robust optimum

OFAT Design – varying one factor at a time provides estimates of effects at set

conditions and no interaction effects

Factorial Design – provides estimates of effects at differentconditions and therefore enables you to estimate interactions

Fractional Factorial - Optimisation DOE

Full Factorial – Central Composite Facing Design

Page 21: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

Design of experiment can be used to assess the design space and any potential factor interactions on the result in a stepwise approach

The DoE Methodology for Developing Robust Analytical Methods

21

Screening OptimizationRobustness

Page 22: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

Platform analytical methods reduce the uncertainty / risk for the manufacturer• Less method development• Less qualification and / or validation

BUT,• More challenges for new molecular formats e.g. ADCs, bispecifics etc.

What is the benefit of a platform method?

22

Peptide‐protein conjugateAntibody‐drug conjugate (ADC) Peptide Fc‐fusionImmunocytokines

Page 23: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

A method can be thought of as three unit operations. For platform methods the measurement and replicate strategy are fixed.

Therefore, we have to focus on the sample preparation operation

How can we make create platform methods for complex molecules?

23

Image taken from 42(5) Stimuli to the Revision Process: Analytical Control Strategy, Kovacs. E (2016)Image taken from USP Analytical

Method Lifecycle Workshop 2018

Page 24: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

Sample Preparation DoE

24

Analytical Testing

Controllable variablesMethod, Materials, Manpower, Instrument

Uncontrollable variablesEnvironmental Factors, Experimental variability

FactorspHTemperature (°C)Time (mins)

ResponsesMain Peak Purity (%)Fragment (%)Number of Peaks

Page 25: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

Initial screening studies determined that sample buffer pH could cause method induced fragmentation

Therefore, DoE principles were applied for optimizing the sample preparation for this method

Initial Sample Preparation Screening Data

25

Minutes6.25 6.50 6.75 7.00 7.25 7.50 7.75 8.00 8.25 8.50 8.75 9.00 9.25 9.50 9.75 10.00 10.25 10.50 10.75 11.00 11.25 11.50 11.75 12.00 12.25 12.50 12.75 13.00

AU

0.000

0.005

0.010

0.015

0.020

0.025

0.030

0.035

0.040

0.045

0.050

AU

0.000

0.005

0.010

0.015

0.020

0.025

0.030

0.035

0.040

0.045

0.050

PDA - 220nmINT-005 - T - NR

PDA - 220nmINT-005 - P - NR

PDA - 220nmINT-005 - C - NR

Red – pH 9.0, Black – pH 8.0, Blue – pH 7.0

Buffer % Main % Leading Peak % LMWS

pH 7.0 91.1 7.0 2.0

pH 8.0 84.2 10.7 5.1

pH 9.0 73.8 18.9 7.3

% Area

Page 26: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

Fractional Factorial Design • Three factors at two levels• Three center points• 11 injections/run

• Input from subject matter experts is typically used to select factors and ranges

• DoE provides the specific combination of these parameters over the ranges to efficiently test their impact as main effects and interactions

DoE Design

26

Factor Levels ResponsesSample Buffer pH

-1, 0, +1Main Peak Purity (%)

Fragments (%)Number of Peaks, etc.

Incubation Temperature (°C)Incubation Time (mins)

Page 27: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

Using a Least Squares model for effect leveraging the model fit is very good, which gives confidence that there is predictive power for the design space

The model highlighted a number of significant primary and secondary interactions within the design space

DoE Model Fit

27

Page 28: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

Final conditions were selected by applying limits to the result responses, the red square depicts the robust operating space

After selecting the conditions samples were run on orthogonal methods to ensure the data was comparable and the method was then implemented in the QC environment

DoE Final Condition Selection

28

Page 29: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

Case Study 3: Control Charting

Page 30: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

These control charts show the method is in a good state of control and demonstrates ruggedness of the method.

The levels of variation are similar to the variability seen across the MODR

Control Charting for Method Control

30

Page 31: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

Control Charting for Investigational use

31

The cIEF example below shows that there was an observed increase in the % basic species area which corresponded to a decrease in % main species and an increase in the number of peaks detected

Looking back at the DoE data for this method we identified a potential root cause from a urea additive mispreparation.After reviewing the ELN, this was confirmed to be the root cause and the samples were re-tested

Page 32: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

Conclusions

Page 33: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

By adopting an AQbD approach we gain:

• Greater method understanding • Ensure greater method control through identification of potential risk to

the reportable result• Clearly define the method purpose• Capture method changes easily through a change control process• Fewer investigations related to analytical method performance

It is an iterative process and knowledge management is essential for full adoption of this approach

Conclusions

33

Page 34: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

Acknowledgements

Analytical SciencesDouglas JohnsonKristin Schultz-KuszakKristen GonzalezSam ShepherdMichael MartinelliCarrie SowersEric MeinkeMethal AlbarghouthiWei XuXiangyang Wang

34

MS&T AnalyticalDerek Schildt

AZ GTO BiologicsChris Larkin

Quantitative SciencesGuillermo Miro-QuesadaJames Savery

Page 35: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

35

Page 36: Analytical Quality by Design: What does it look like? › › resource › ... · 3ICH Q8(R2) Pharmaceutical Development(2009) 4ICH Q9 Quality Risk Management (2005) 5ICH Q10 Pharmaceutical

Confidentiality Notice This file is private and may contain confidential and proprietary information. If you have received this file in error, please notify us and remove it from your system and note that you must not copy, distribute or take any action in reliance on it. Any unauthorized use or disclosure of the contents of this file is not permitted and may be unlawful. AstraZeneca PLC, 1 Francis Crick Avenue, Cambridge Biomedical Campus, Cambridge, CB2 0AA, UK, T: +44(0)203 749 5000, www.astrazeneca.com

36