(very) Pilot Study to Automate Forced Degradation Workflows

David Benstead, AstraZenecaACD/Labs User Group Meeting 27th September 2016

Why Automate Forced Degradation Workflows?

Not a new idea!

Automating some or all of the forced deg workflow has a

large potential payoff

• All development compounds must have a forced deg

• Robust retention of raw data and context is imperative

• Data should be generated as few times as possible to

avoid “surprises”

Forced deg can easily generate 30+ samples per compound

• Significant time cost to manually process them all

2

Stability Understanding

From ICH Q1A:

Forced degradation/stress testing forms part of a wider stability

understanding, and the knowledge is used throughout the product

lifecycle

3

Drug Candidate

•What do we know?

•Structure

•Lab “lore”

Predictions

•Company knowledge

•Literature sources

•In silicoprediction

Reality

•Forced Degexperiments

•Accelerated stability

•Long term stability

Data

•Process

•Interpret

•Store

Knowledge

•Capture

•Immediate use

•Long term use

Automation

Forced degradation studies generate large numbers of samples and

data, which are time consuming to acquire and process.

Can the acquisition and interpretation process be made easier?

4

Drug Candidate

•What do we know?

•Structure

•Lab “lore”

Predictions

•Company knowledge

•Literature sources

•In silicoprediction

Reality

•Forced Degexperiments

•Accelerated stability

•Long term stability

Data

•Process

•Interpret

•Store

Knowledge

•Capture

•Immediate use

•Long term use

Automation

5

Structure In silicoprediction*

Forced Degexperiments

•LC-MS

Data Knowledge

What tools are available, and how can we combine them?

* Zeneth 6.0 (Lhasa Limited) used here. For further info see Parenty et al, Mol. Pharmaceutics, 2013, 10 (8), pp 2962–2974,

doi:10.1021/mp400083h

6

Rule based prediction of degradation products under a variety of

conditions

Relies on existing knowledge base, which can be added to with

custom knowledge

Degradation map can be exported in a variety of formats, including

.sdf

Zeneth Degradation Map

7

Zeneth Degradation Map

8

.sdf

Zeneth Degradation Map

9

.sdf

Q

D1

Spectrus MetID

Spectrus MetID offers a workflow which is very

relevant to forced deg interpretation

Identification of components from a predicted

metabolic transformation map, component

filtering and databasing of transformation,

structural and spectral information.

Could the Zeneth output be used to input

structures and component relationships into the

MetID workflow?10

DegID

DegID is a script to feed the Zeneth .sdf output

into the MetID workflow

11

DegID

12

DegID

13

DegID

14

15

DegID

16

Metasense

Metasense is an evolution of MetID and

incorporates extra features such as time-course

profiling

17

Metasense

18

Multiple chromatograms with

time-course data

Multiple forced

deg conditions

Doesn’t currently

interface with Zeneth

maps – this example

uses MetID map.

Metasense (current output)

19

Degradation

map

Component XIC

TIC

Degradation Profile

Multiple

conditions and

timepoints

Graphical output

for deg map via

Chemsketch

Export as .csv

DegID/”DegSense”

The Good:• Retains reaction information from Zeneth deg map

• Databases reactions, structures, spectra and chromatograms easily with

minimal user data entry.

• Easy visualisation of data.

• Excellent tool for “one place” retention of forced deg data.

The room for improvement: • Use of UV for quantification/retention times

• Isobaric impurities – potentially resolvable with HRMSn data for non-isomers

• Sensible Markush

• Incorporation of non-Zeneth predictions

• Zeneth “intermediates” structures – reactive intermediates contained within

Zeneth .sdf but not part of the degradation map.20

“Degsense” (thoughts on desired output)

21

Degradation

map Component XIC

DAD or flat chrom

UV quantitation at

chosen wavelength

Multiple

conditions and

timepoints

Project name field

Component

UV

spectrum

Is there an opportunity to automatically select optimum wavelength?

Deg map at

different

levels

Degradation ProfileDeg map filter –

predicted,

observed etc.Graphical output

for deg map via

Chemsketch

Export as .csv

Conclusions and End game

ACD workflows can provide an opportunity to automate

aspects of forced deg and degradation knowledge

management

What could an automated forced deg workflow look like?

22

Low pH T8_3_TIC

Retention Time (min)4.54.03.53.02.52.01.51.00.50

-100

-80

-60

-40

-20

0

20

40

60

80

100

Automated

LC-HRMS2

Autochrom10%?

Best API Generic

MethodNo

Yes

“DegSense”Spectrus DB

Acknowledgements

Peter Russel, Albert van Wyk and Richard Lee Advanced Chemistry Development, Inc. (ACD/Labs)

Alex Harding and Lisa BarrLhasa Limited

23

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

24