Predictive Stability in
Pharmaceutical Development
Helen Williams
Drug Delivery and Formulation Summit, Berlin
14th March 2018
Overview
➢ Predictive stability tools
ASAP
Packaging predictions
➢ Case studies
Phase I API and Tablet ASAP studies
Phase III Tablet ASAP study
NDA packaging predictions
➢ IQ Risk Based Predictive Stability Working Group
➢ Conclusions
2
Predictive Stability Tools
3
Accelerated Stability Assessment Program Studies
Based on the Arrhenius equation modified for solid state degradation
If measure how reaction rate
changes with temperature &
humidity, can determine Ea and
ln (A) and B and via extrapolation
determine the reaction rate at any
given temperature and humidity.
4
D. Genton et al, J. Pharm. Sci. 66, 5 (1977) 676-680.
K. Waterman and S.Colgan, Regulatory Rapporteur, 5 (2008) 9-14.
K.Waterman et al, Pharm. Res., 24 (2007) 780-790.
Packaging Predictions
Depends on permeation of the barrier
Difference in partial pressure of water
between ambient environment and
headspace
Headspace
Equilibrium between headspace and tablets
is defined by the Gravimetric Vapour
Sorption isotherm of the tablets
5
Predictive Stability - Applications
ASAP and packaging predictions can potentially be utilised to reduce stability testing
requirements for regulatory submissions…..
➢ In early clinical trial submissions to support initial shelf life/retest claim
➢ To support API route/process changes during development
➢ To support minor formulation changes
during development
➢ To justify excipient selection
➢ To compare batch to batch stability
characteristics
➢ For specification justification
➢ To support pack changes during
development and post approval
➢ As supporting information in a marketing
submission to then support changes post approval
6
Case Study 1 - Phase I API and Tablet ASAP Studies
Applications of predictive stability tools
1. Drug substance retest period prediction - ASAP study
2. Drug product shelf life prediction – ASAP study
3. Regulatory Applications
7
Case Study 1 - Phase I API and Tablet ASAP Studies
1. Drug substance ASAP study
Temperature (°C)
Humidity (% RH)
Storage Time (weeks)
Initial
Initial
0 (3 repeats), X
50
75
3, 6, S
60
30
3, 6, S
60
75
3 (5 repeats), 6, S
70
11
2, 3, S, X
70
75
2, 3, S, X
80
30
2, 3, S, X
Impurities analysis by LC at all time points
S = spare samples,
X = XRPD sample,
8
No degradation was observed
Case Study 1 - Phase I API and Tablet ASAP Studies
2. Drug product shelf life prediction – ASAP
ASAP studies for 2 formulation strengths, 5 and 50 mg
Protocol for impurities and appearance
Temperature
(°C)
Initial
50
60
60
70
70
80
Humidity
(%RH)
Initial
75
30
75
11
75
30
S = spare sample
9
Storage Time
(weeks)
0
3, 6, S
3, 6, S
3, 6, S
2, 3, 6, S
2, 3, 6, S
2, 3, 6, S
Protocol for dissolution
Temperature
(°C)
Initial
40
40
40
[ ] = optional
Humidity Storage Time
(%RH)
(weeks)
Initial
0
11
[4]
55
[4]
75
4
Case Study 1 - Phase I API and Tablet ASAP Studies
2. Drug product shelf life prediction – ASAP
No change in dissolution profile or
appearance during study
“Good” model
10
Case Study 1 - Phase I API and Tablet ASAP Studies
2. Drug product shelf life prediction – ASAP
Predictions at 25°C/60% RH
support an initial shelf life of
12 months but suggest the
product is likely to achieve 5
years
11
Case Study 1 - Phase I API and Tablet ASAP Studies
3. Regulatory Applications
➢ Drug substance and drug product ASAP data was presented in the
Phase I regulatory submission to support an initial 12 month shelf
life and retest period, in the absence of long term stability data but
with a commitment to set down ICH compliant stability.
➢ Submitted to UK and USA
➢ Accepted with no related questions in USA.
➢ MHRA (UK) requested ICH compliant stability data
➢ Responded with 3 month ICH compliant stability data across drug
substance and product to support 12 month shelf life/retest period.
12
Case Study 1 - Phase I API and Tablet ASAP Studies
3. Regulatory Applications
For drug product we included a comparison of the ASAP predictions with
the ICH compliant stability data in the response.
25°C/60% RH
13
40°C/75% RH
Case Study 2 - Phase III Tablet ASAP Study
➢ 200 mg tablet formulation, phase III
➢ Manufacturing change from wet granulation to roller compaction
➢ Chemically stable product, 2 years long term stability data on wet
granulation product
➢ Reduction in dissolution rate observed when exposed to high
humidity, so packed with desiccant in bottle
Needed to quickly determine stability characteristics of the new roller
compaction formulation …………. ASAP study
14
Case Study 2 - Phase III Tablet ASAP Study
Temperature (°C)
Humidity (% RH)
Storage Time (weeks)
Initial
Initial
0 (3 repeats), X
50
75
4, 8, S
60
11
4, 8, S
60
75
4, 8 (5 repeats), S
70
11
4, 8, S, X
70
75
4, 8, S, X
80
30
4, 8, S, X
Impurities analysis by LC at all time points
S = spare sample, as optional additional time point
X = XRPD sample
15
Case Study 2 - Phase III Tablet ASAP Study
Dissolution Protocol
Temperature (°C)
Humidity (% RH)
Storage Time (weeks)
Initial
Initial
0
40
30
6, S
40
55
6, S
40
75
6, S
24 tablets set down for each time point
(n=6, two dissolution methods pH 1.2 and 4.5, and spares)
16
Case Study 2 - Phase III Tablet ASAP Study
➢ No degradants detected during the 8 week ASAP study
➢ No change in physical form of API detected (XRPD)
➢ No change in dissolution profile of tablets stored at 40°C/30% RH
➢ Evidence of
dissolution rate
slowing in tablets
stored exposed
to higher
humidities
➢ All still pass
specification
(>75% dissolved
at 30 minutes)
17
Case Study 2 - Phase III Tablet ASAP Study
Conclusions
➢ Roller compacted tablets are chemically stable
➢ The tablets should be packed with desiccant to minimise the risk of a
reduction in dissolution rate on storage
➢ Demonstrated equivalence with wet granulation tablet formulation
ASAP data (impurities and dissolution) was included in the Phase III
IND and IMPD submissions, as supporting information.
18
Case Study 2 - Phase III Tablet ASAP Study
Regulatory Submission
➢ ASAP study on 200 mg roller compaction batch
➢ At time of IND submission 1 month ICH compliant stability data on
200 mg roller compaction development batch
➢ At time of IMPD submission 3 months stability data on 200 mg roller
compaction development batch and 1 month on clinical batch
➢ Alongside 2 years stability data on 200 mg wet granulation batch
➢ Claimed a 12 month initial shelf for new 200 mg roller compaction
formulation in the bottle packed with desiccant
➢ Submitted in Canada, US, Ukraine, Russia, Taiwan and Poland
➢ Accepted with no regulatory questions
19
Case Study 2 - Phase III Tablet ASAP Study
Further ASAP study
➢ Six months later, a lower dose tablet was required for a different
clinical trial (100 mg, roller compaction, common granule)
➢ No real time stability data on 100 mg formulation
➢ Performed a second, very similar ASAP study on 100 mg tablets
➢ Again no degradation was observed during the study and a small
reduction in the dissolution rate was observed at 40°C/75% RH.
➢ Demonstrated equivalence between the 100 mg and 200 mg roller
compaction formulations in terms of stability characteristics
20
Case Study 2 - Phase III Tablet ASAP Study
➢ No long term stability data on new 100 mg roller compaction tablets,
ASAP study data only was presented for this formulation
➢ Alongside 2 years stability data on 100 mg wet granulation batch
and 6 months stability data on two 200 mg batches roller compaction
tablets.
➢ Claimed a 18 month initial shelf for new 100 mg roller compaction
formulation in bottle pack with desiccant, based on equivalence with
200 mg roller compaction formulation.
➢ Committed to setting down ICH stability for 100 mg roller compaction
formulation
➢ Submitted to Ukraine, Russia, Canada, US, Taiwan, Poland and
Japan
➢ Accepted with no related regulatory questions
21
Case Study 2 - Phase III Tablet ASAP Study
Packaging Predictions
Predictions were also performed to determine the tablet water content
and humidity inside the pack on storage at 25°C/60% RH for 3 years.
22
Case Study 2 - Phase III Tablet ASAP Study
Packaging Predictions
Predictions suggest that the humidity in the bottle will remain below
30% RH over 3 years stored at 25°C/60% RH.
23
Case Study 3 - NDA Packing Predictions
The two proposed commercial packs are in HDPE bottles:
1. Existing bottle containing 30 capsules, 12 months stability data across 3
batches available
2. New bottle containing 60 counts, 6 months stability data available
Packaging predictions were performed for the new bottle at 25°C/60%RH for 24
months.
Compared the predictions to stability data for the existing bottle configuration.
24
Case Study 3 - NDA Packing Predictions
Both bottle configurations
offer similar moisture
protection.
These predictions were
presented in the NDA and a
24 month shelf life was
granted for both bottles
configurations.
25
IQ Working Group - Risk Based Predictive Stability
➢ Risk Based Predictive Stability (RBPS) IQ Working Group was set up in 2015
➢ A survey of all IQ member companies was conducted in 2016
➢ 19 companies responded of which 16 used predictive stability studies
➢ 10 companies reported using predicted stability data in submissions,
6 in clinical and 5 in marketing submissions with 2 post approval changes
➢ 23 countries were reported to have accepted predicted stability data in clinical
submissions and “worldwide” acceptance was reported in Marketing submissions.
26
H. Williams et al., "Risk-Based Predictive Stability–An Industry Perspective," Pharmaceutical Technology 41 (3) 2017
http://www.pharmtech.com/risk-based-predictive-stability-industry-perspective
IQ Working Group - Risk Based Predictive Stability
➢ Currently working on harmonising how predicted stability data is presented
in clinical submissions, with the aim of publishing both a template and a
worked example during 2018
➢ Collecting case studies of predicted stability data in clinical submissions for
a further publication
➢ Comparing different predictive stability models across the industry
27
Conclusions
➢ Predictive stability tools can be used to support development.
➢ Zeneth and BDE tools can aid understanding of potential degradation
mechanisms across both substance and product.
➢ ASAP and packaging predictions are powerful tools to predict chemical
degradation, water content and pack humidity and can influence pack selection,
pack configuration, storage conditions, shelf life/retest period claims,
specification setting and control strategy.
➢ Pharma companies are working together through IQ Working Group to
harmonise approaches and influence regulatory authorities.
➢ Regulatory acceptance of predicted stability data is growing.
28
Acknowledgements
Confidentiality Notice
Faye Turner
Andrew Brookes
Carolyn Gordon
Jonathan Bright
Emily Roddy
Emily MacDougall
Pam Harrison
Magnus Fransson
Johan Remmelgas
Nadim Akhtar
James Mann
Anna Powell
Keith Parker
John Nightingale
Angela Currie
Darren Gore
Andrew Phillips
Ben McKeever-Abbas
Andrew Poulton
Paul Cronin
Debbie Lane
Edward Griffin
Members IQ Working Group
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
29
Predictive Stability Tools
➢
➢
➢
➢
30
Determine risk of autoxidation
➢ Determine potential degradants
Performed early in development
➢ Performed early in development
Used to compare lead compounds
before forced degradation studies
If autoxidation risk identified,
and method development
followed up with EPR
➢ Used to identify possible excipient
spectroscopy analysis to monitor
incompatibilities to support
for radical formation
formulation development
Predictive Stability Tools
➢ Used throughout development to
predict drug substance or product
stability
31
➢
➢
➢
➢
Used to support pack selection
Determine the need for desiccant
Set water content specifications
Used to support pack changes
during development
➢ Used to support pack changes
post approval
Bond Dissociation Energy Tool
The BDE values can be accurately
calculated for hydrogen atoms in a
molecule using relatively fast DFT
methods1. The risk of autoxidation
can then be estimated.
Initiation
Propagation
Termination
BDE < 87 kcal/mol = risk for
autoxidation
32
1T.
In principle, the chain reaction can be fast
as long as the breaking CH-bond is
weaker than the OH bond (87 kcal/mol)
that is formed in the hydrogen transfer.
Andersson, A. Broo, E. Evertsson, J. Pharm. Sci. 103, (2014), 1949-1955
Bond Dissociation Energy Tool
c
b
b
a
a
3 types of non-equivalent protons
Septet of septet of quartets
(196 lines!)
a
a
b
b
c
a
b
BDE = 96 kcal/mol
c
a
b
BDE==72
72 kcal/mol
BDE
kcal/mol
Electron Paramagnetic Resonance spectroscopy performed
to monitor for free radical formation
33
Simulation based on coupling constants from 2 P. Brovetto, G. Bussetti, Rendiconti del
Seminario della Facoltà di scienze dell'Università di Cagliari, 39 (3-4) p.387
Zeneth
Zeneth is designed to predict potential degradants of a compound from the chemical
structure.
➢ 446 known and validated transformations (version 7.0, 2016 knowledge base)
➢ Data from published sources and data donated by a consortium of member organisations
➢ Drug Substance and Drug Product stability
Benefits of Zeneth3:
➢ Provides degradation information when no experimental data is available
➢ Data generated can be used to support regulatory submissions
➢ Assist in the selection of excipients, highlighting potentially problematic ones
➢ Helps in assignment of structure to forced degradation work
➢ Prevents the overlooking of possible pathways through unbiased application
of transformations
➢ Presence or absence of conditions (light, water, oxygen, radical initiators,
peroxides, metal, temperature, pH)
➢ Relative likelihood of competing reactions
34
3 http://www.lhasalimited.org/products/zeneth.htm
Zeneth
Example query
compound
Zeneth predictions displayed
in a tree layout. Zoom out for
high level view & zoom in for
compound specific detail
Transformation
description and
references to
supporting
literature
Simplified tree
view of results
35
Zeneth also includes a list of common excipients
and is able to predict whether they are likely to
interact with the query compound
Table to show all
degradants generated. Can
be filtered upon molecular
formula or molecular mass
Phase I formulation development - tablet
2. Drug product shelf life prediction – ASAP
➢ When comparing both
strengths of tablets, it can be
seen that the error for the
50mg tablets is much larger
than for the 5mg
➢ This is because the 50mg
tablets did not degrade as
much as the 5mg, therefore
increased extrapolation
required to reach
specification
36
Phase III Tablet ASAP Study Protocol
37
Temperature (°C)
Humidity (% RH)
Initial
Initial
50
75
60
11
60
75
70
11
70
75
80
30
Phase III Tablet – ASAP Study Set down
➢ The tablets were set down in glass jars, with inserts containing the tablets
and a further insert containing the relevant salt solution.
➢ Salt solutions were used to control the humidity - lithium chloride 11% RH,
magnesium chloride 30% RH, sodium chloride 75% RH and Amebis humidity
capsule (U036) for 55% RH.
➢ Coated tablets are usually allowed to
pre-equilibrate with humidity for at least
24 hours before exposure to high
temperatures
➢ The jars were then placed in ovens
➢ 44 tablets for the initial analysis were
placed in the fridge, also stored in the
same inserts.
38