1 WHO/ECTC_DRAFT/21 MAY 2015 2
ENGLISH ONLY 3
4
Guidelines on the stability evaluation of vaccines for 5
use under extended controlled temperature conditions 6
7
8
NOTE: 9
This document has been prepared for the purpose of inviting comments and suggestions on the 10
proposals contained therein, which will then be considered by the Expert Committee on 11
Biological Standardization. Publication of this early draft is to provide information about the 12
proposed WHO Guidelines on the stability evaluation of vaccines for use in a controlled 13
temperature chain to a broad audience and to improve transparency of the consultation process. 14
15
The text in its present form does not necessarily represent an agreed formulation of the 16
Expert Committee. Written comments proposing modifications to this text MUST be 17 received by 21 June 2015 in the Comment Form available separately and should be addressed 18
to the World Health Organization, 1211 Geneva 27, Switzerland, attention: Department of 19
Essential Medicines and Health Products (EMP). Comments may also be submitted electronically 20
to the Responsible Officer: Dr Kai Gao at email: [email protected]. 21
22
The outcome of the deliberations of the Expert Committee will be published in the WHO 23
Technical Report Series. The final agreed formulation of the document will be edited to be in 24
conformity with the "WHO style guide" (WHO/IMD/PUB/04.1). 25
26
© World Health Organization 2015 27 All rights reserved. Publications of the World Health Organization can be obtained from WHO Press, World Health 28 Organization, 20 Avenue Appia, 1211 Geneva 27, Switzerland (tel.: +41 22 791 3264; fax: +41 22 791 4857; e-mail: 29 [email protected]). Requests for permission to reproduce or translate WHO publications – whether for sale or for non-30 commercial distribution – should be addressed to WHO Press, at the above address (fax: +41 22 791 4806; e-mail: 31 [email protected]). 32 The designations employed and the presentation of the material in this publication do not imply the expression of any opinion 33 whatsoever on the part of the World Health Organization concerning the legal status of any country, territory, city or area or of its 34 authorities, or concerning the delimitation of its frontiers or boundaries. Dotted lines on maps represent approximate border lines 35 for which there may not yet be full agreement. 36
37 The mention of specific companies or of certain manufacturers’ products does not imply that they are endorsed or recommended 38 by the World Health Organization in preference to others of a similar nature that are not mentioned. Errors and omissions 39 excepted, the names of proprietary products are distinguished by initial capital letters. 40 41 All reasonable precautions have been taken by the World Health Organization to verify the information contained in this 42 publication. However, the published material is being distributed without warranty of any kind, either expressed or implied. The 43 responsibility for the interpretation and use of the material lies with the reader. In no event shall the World Health Organization 44 be liable for damages arising from its use. 45
46 The named authors [or editors as appropriate] alone are responsible for the views expressed in this publication. 47
Page 2
1
This document provides guidance to National Regulatory Authorities (NRAs) and manufacturers
on scientific and regulatory issues to be considered in evaluating the stability of vaccines for use
under extended controlled temperature conditions (ECTC). It should be read in conjunction with
the existing guidelines on the stability evaluation of vaccines published by the WHO. The
following text is written in the form of WHO Guidelines rather than Recommendations because
vaccines represent a heterogeneous class of agents and the stability testing programme will need
to be adapted to suit the product in question. WHO Guidelines allow greater flexibility than
Recommendations with respect to specific issues related to particular vaccines.
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Page 3
Contents 1
2
1. INTRODUCTION ..................................................................................................................... 4 3
2. SCOPE ........................................................................................................................................ 5 4
3. GLOSSARY ............................................................................................................................... 5 5
4. GENERAL CONSIDERATIONS FOR THE EVALUATION OF VACCINES FOR USE 6
UNDER ECTC ............................................................................................................................... 7 7
5. STABILITY EVALUATION OF VACCINES FOR USE UNDER ECTC ....................... 12 8
6. MONITORING ECTC ........................................................................................................... 17 9
7. SUGGESTED PRODUCT LABELLING INFORMATION FOR USE UNDER ECTC. 18 10
AUTHORS& ACKNOWLEDGMENTS .................................................................................. 19 11
REFERENCES ............................................................................................................................ 23 12
APPENDIX: PRODUCT SPECIFIC ECTC EVALUATION OF A MODEL 13
MONOVALENT POLYSACCHARIDE CONJUGATE VACCINE..................................... 24 14
15
Page 4
1. Introduction 1
2
Vaccines are complex biologics and may undergo degradation during long term storage within a 3
cold chain (e.g. 2-8⁰C) and this is typically enhanced at higher temperatures. Therefore, 4
establishing the stability characteristics of products is a critical element of the overall evaluation 5
by NRAs to ensure that licensed vaccines remain efficacious at the end of their shelf-life when 6
stored under the approved conditions. In response to the stability assessment needs identified by 7
NRAs, WHO developed guidelines on the stability evaluation of vaccines to assist its Member 8
States (1). While the dependence of vaccine quality on cold chain storage is well understood, it is 9
also recognized that immunization programmes in certain regions face substantial challenges in 10
the field with maintaining cold chains, especially during the final stage of distribution in remote 11
areas (2-4). To address these distribution challenges and expand immunization programmes into 12
specific regions, the World Health Organization (WHO) developed a programme referred to as a 13
“Controlled Temperature Chain” (CTC) (4). This programme requires that a vaccine exhibit a 14
stability profile suitable for a single exposure to at least 40°C for a minimum of three days just 15
prior to administration, while remaining compliant with the approved vaccine specifications. 16
Additionally, the programme requires that the CTC provision be included in the licensure by the 17
relevant RNA and the WHO pre-qualification (5). 18
19
During the consultations for this document, the term “Extended Controlled Temperature 20
Conditions” (ECTC) was proposed to distinguish regulatory requirement from programme 21
aspects. This terminology is used throughout this guidance. An ECTC assessment establishes the 22
short term performance of a vaccine at temperatures above those of a typical cold chain and 23
could consider any temperature above the traditinal 2-8°C cold chain that might support vaccine 24
distribution. This is independent of the specific programmatic requirements of the WHO CTC 25
programme. Vaccines licensed for use under ECTC are required to have sufficient information 26
regarding the approved conditions (e.g. maximum temperature and time) on the package insert. 27
28
An example of an approved ECTC product, compliant with the WHO’s CTC programme 29
requirements, is the Meningitis A conjugate vaccine MenAfriVac. The ECTC evaluation and 30
CTC approval for MenAfrVac has made it possible to distribute this vaccine to populations that 31
would otherwise have been difficult to immunize due to limitations in the availability of 32
traditional cold chain (6-7). ECTC labelling allows greater flexibility in vaccination campaigns 33
by reducing health worker burdens and saving refrigeration and generator infrastructure costs, as 34
Page 5
well as addressing the difficulties associated with the distribution of vaccines on wet ice. 1
Additionally, this “on label” NRA approved approach under ECTC avoids “off label” vaccine 2
administration, which is inconsistent with official guidance on best practice (8). 3
4
As mentioned above, this guidance document arises from the WHO immunization programme 5
requirements (2-3) and the resulting discussions held by an international group of vaccine 6
stability experts at WHO-sponsored consultations in Ottawa Canada in December 2012 (9), and 7
in Langen, Germany in June 2013 (10). This ECTC guidance is intended as a supplement to the 8
broader existing WHO “Guidelines on Stability Evaluation of Vaccines” (1), and focuses on 9
ECTC specific issues not covered in the existing guidance with as little overlap as possible. The 10
key elements of this document are the application of the mathematical modeling and statistical 11
concepts in the existing stability guidance (1), as well as in related publications (11-12), to 12
address the unique short term needs required in some cases for vaccine distribution and use (9, 13
10). Early dialogue between manufacturers and regulators, as well as with public health officials 14
in the immunization programmes is recommended, so that those vaccines compatible with ECTC 15
use can be evaluated for licensure by the appropriate NRA. 16
17
2. Scope 18
19
These guidelines describe criteria for the approval of short term temperature conditions above 20
those defined for long term storage of a given vaccine, where the vaccine is exposed to these 21
short term conditions immediately prior to administration. 22
23
This document does not provide guidance on stability evaluation of vaccines inadvertently or 24
repeatedly exposed to temperatures for which they were not licensed. 25
26
3. Glossary 27
28
The definitions given below apply to the terms used in these guidelines. They may have different 29
meanings in other contexts. 30
31
Cold chain: The system used for keeping and distributing vaccines in good condition consists of 32
a series of storage and transport links, all designed to keep vaccines within an acceptable 33
Page 6
predefined temperature range until they are used, typically 2-8⁰C but other approved 1
temperatures can be specified. This is considered a long term storage condition for many 2
vaccines. 3
4
WHO Controlled Temperature Chain (CTC) Programme: A specific innovative approach to 5
vaccine management allowing vaccines to be kept at temperatures outside of the traditional cold 6
chain of 2-8⁰C for a limited period of time under monitored and controlled conditions, as 7
appropriate to the stability of the antigen. Current WHO programme conditions for CTC include 8
a single exposure just prior to administration, tolerating ambient temperatures of at least 40°C for 9
a limited duration of at least three days, with the CTC provision included in the licensure by the 10
relevant RNA and the WHO pre-qualifcation. 11
12
Extended Controlled Temperature Conditions (ECTC): Approved short term temperature 13
conditions above those defined for long term storage, transportation and use for a given product, 14
immediately prior to administration. 15
16
Real-time and real-condition stability studies: Studies on the physical, chemical, biological, 17
biopharmaceutical and microbiological characteristics of a vaccine, during and up to the 18
expected shelf-life and storage periods of samples under expected handling and storage 19
conditions. The results are used to recommend storage conditions and to establish the shelf-life 20
and/or the release specifications. 21
22
Accelerated stability studies: Studies designed to determine the impact over time of the 23
exposure to temperatures higher than those recommended for storage regarding the physical, 24
chemical, biological, biopharmaceutical and microbiological characteristics of a vaccine. When 25
the accelerated temperature conditions are equivalent to or higher than the ECTC condition 26
under evaluation, the accelerated stability data can be considered for support of allowed 27
exposure conditions. 28
29
Shelf-life: The period of time during which a vaccine, when stored under approved conditions, is 30
expected to comply with the specification. The shelf-life is determined by stability studies on a 31
number of batches of the product and is used to establish the expiry date of each batch of a final 32
product. 33
34
Page 7
Product Release Model: A model that describes the relationship between release and expiry 1
specifications to ensure that the product will maintain adequate quality throughout shelf-life. 2
3
Stability indicating parameters: Quality parameters (direct or indirect indicators of vaccine 4
efficacy or safety) that are sensitive to storage conditions. They are used to assess product 5
suitability throughout the shelf-life. Determination of these parameters should result in 6
quantitative values with a detectable rate of change. Qualitative parameters such as sterility 7
could also be considered but cannot be included in the statistical analysis. 8
9
Stability of vaccines: The ability of a vaccine to retain its chemical, physical, microbiological 10
and biological properties within specified limits throughout its shelf-life. 11
12
Quality Attributes: These are chemical, physical, biological and microbiological attributes that 13
can be defined, measured and continually monitored to ensure final product outputs remain 14
within acceptable quality limits. 15
16
4. General considerations for the evaluation of vaccines for use 17
under ECTC 18
Use of vaccines under ECTC requires an appropriate vaccine stability assessment and the 19
consideration of the feasibility of compliance with the approved storage conditions in the field. 20
While the stability evaluation principles described here could potentially be applied to data to 21
support multiple temperature exposures for a vaccine, one needs to consider how such exposures 22
would be tracked for specific vaccine final containers. When contemplating the potential 23
difficulty in tracking multiple exposures and assuring that final containers that have reached the 24
maximum exposure limit are discarded, it was concluded that, at this time, guidance for an 25
ECTC label should be limited to a single planned exposure of specified duration within the 26
labelled expiration date. Once a vaccine has been stored under ECTC, it should not be returned 27
to normal cold chain storage (e.g. 2 - 8°C) in order to prevent inadvertent administration of 28
vaccine that is potentially out of specification. As experience with ECTC stability assessment 29
and programme implementation expands, this could potentially be reconsidered in a future 30
guidance update. 31
32
Page 8
An ECTC application could potentially be approved based solely on product specific 1
stability/quality data when the following conditions are met: 2
- the approved product specifications, supported by quality attributes of the clinical lots, 3
remain unchanged and the vaccine is expected to be compliant with these specifications 4
following the normal storage for the full shelf-life including the ECTC exposure; 5
- that the battery of tests performed to assess vaccine stability, which may include 6
additional characterization assays in specific cases, have the capacity to detect changes in 7
potency and/or immunogenicity that are predictive of vaccine clinical performance. 8
9
Additionally, when a manufacturer has accelerated stability data, that brackets the intended 10
ECTC exposure of 40°C, the potential for interpolation of the data to support a 40°C ECTC label 11
could be considered on a case by case basis. The accelerated stability data must be from lots that 12
represent the current manufacturing process. 13
14
Vaccines used under ECTC should be capable of withstanding the approved planned exposure 15
conditions regardless of the shelf-life remaining before expiry. These evaluations must involve 16
statistical analysis of stability data to determine the rates of decay under both the approved long 17
term storage conditions and those of an ECTC exposure. It is essential that adequate potency be 18
available to compensate for any decay over the full approved shelf-life under approved long term 19
storage conditions, plus under the planned ECTC exposure (e.g. 40⁰C for at least three days) to 20
address worst-case scenarios, where the planned exposure occurs within the shelf-life for a 21
vaccine lot that was filled at or near the minimum release potency (MRP). 22
23
Calculations must be made based on an evaluation of decay rates, MRP and appropriate end-24
expiry Lower Limit (LL) potency determined from the manufacturer’s clinical trial data. This is 25
described in the existing stability guidance and subsequent papers (9-10) and is critical for ECTC 26
applications, for which the same principles apply. The Product Release Model (Figure 1) should 27
be developed based on studies using the manufacturer’s assays, along with the quality data and 28
other essential information. The need to label a vaccine for ECTC use will of course require the 29
support of the manufacturer and the approval of the appropriate regulatory authority(ies). 30
31
Page 9
1 Figure 1: Graphic representation of a “Product Release Model” for an ECTC application. The figure 2 illustrates the relationship between the Minimum Release Potency (MRP) specification (50 EU) and the 3 shelf-life (24 months), given the rate of decay (slope) of the potency over both the long term storage 4 temperature (e.g. 2-8°C) and the maximum ECTC temperature (e.g. 40°C), to ensure that the vaccine is 5 above the approved Lower Limit (LL) for the potency (30 EU, supported by clinical lots) at the end of 6 shelf-life. The potency decay assessment should be based on an appropriate statistical analysis, with a 7 given degree of confidence (e.g. 95%), of multiple lots and should include assay variability. As noted in 8 Section 5, typically log-transformation of potency data permits analysis of stability data by linear 9 regression. 10 11
Potency assays used in the assessment of stability should be fit-for-use and typically are 12
validated as accurate, sensitive, robust and stability-indicating. Given the greater chance of 13
product failure after exposure to the high ECTC temperatures, an assay’s ability to detect 14
quality-related outcomes associated with an ECTC exposure should be re-evaluated, even with 15
existing approved assays. In some cases, supplementary potency assays or key stability 16
indicating tests linked with clinical performance of the vaccine (for example, the test for % free 17
saccharide in glycoconjugate vaccines, if not already performed) should be considered. Other 18
assays, as are typically performed during product stability testing, should also be considered for 19
use in an ECTC context. This may include assessment of quality attributes that may themselves 20
affect stability (e.g. moisture, pH), as well as studies that support container integrity under the 21
ECTC (e.g. sterility, specific container integrity tests). It is not possible to perform decay 22
modeling on products with potency assays that have binary outputs (e.g. pass/fail). In those 23
cases, supplemental potency assays that are capable of showing decay of the product’s active 24
ingredient (or that can provide a worst-case estimate of that decay) may be considered for use in 25
ECTC evaluation, recognizing the need for conservatism in interpretation of the analysis and 26
0
30
40
50
60
70
80
-60 6 12 18 24 30 36
Potency range at release
Potency rangeof clinical lots
Shelf-life (months)
Po
ten
cy
(E
U:E
LIS
A U
nit
)
MRP
Availablepotency
LL
2-8ºC
40ºC
Page 10
results. In the case of absence of adequate stability-indicating assays for a vaccine approval of an 1
ECTC label solely on the basis of a quality data assessment would not be possible. 2
3
While ECTC approval is not a recommendation for shipping or storage and unplanned 4
excursions are not within the scope of this guidance, an approved ECTC label could potentially 5
be considered to guide decisions for use in case of temporary temperature excursions. However, 6
given the finite nature of the available potency over the shelf-life for any given vaccine, potency 7
assigned to deal with unplanned excursions with specific final containers earlier in the shelf-life 8
would not be available to support the use of those same containers of vaccines during a planned 9
ECTC exposure later in the shelf-life. This highlights the importance of maintaining the cold 10
chain, prior to the extreme temperature conditions that the vaccine would be subjected to during 11
a planned ECTC exposure. 12
13
For multivalent vaccines, ECTC evaluation must consider all antigens in the product. If one 14
antigen is known to be less stable than other antigens within a specific vaccine, the assessment 15
should be based on this least stable antigen. Potential interference among vaccine components 16
including adjuvants, stabilisers and preservatives may also need to be considered as applicable. 17
18
The focus here will be on how to evaluate and manage the available potency for a monovalent 19
vaccine to simplify the discussion. At release, products must contain sufficient potency to ensure 20
clinical effectiveness through out its shelf-life and to account for assay variability as well as any 21
product decay. If there is insufficient potency available to permit ECTC use, or if it is desirable 22
to extend the time of the ECTC storage condition, several strategies could be considered to 23
enhance the ECTC potential of a vaccine. The shelf-life under the approved long term storage 24
condition could be reduced, which would enhance the available potency that could be applied to 25
an ECTC exposure. An example of shelf-life reduction to extend the ECTC storage time is 26
shown in the Appendix. If such an approach were used, it would be important to give a unique 27
product name to the ECTC version of the vaccine to avoid confusion in the field. The case study 28
in the Appendix, also illustrates that if a manufacturer were able to implement a lower release 29
specification for free polysaccharide (Free PS) for the model conjugate polysaccharide vaccine, 30
this would create a larger differential between the maximum release limit for Free PS and the 31
end-expiry upper limit for free PS, which would enhance the ECTC potential of the product. 32
33
Additional potency for ECTC applications could also be identified by demonstrating through 34
clinical studies that lower potencies are still effective, thus reducing the amount of potency 35
Page 11
required over the shelf-life to assure clinical effectiveness. Alternately, a manufacturer could 1
choose to fill a vaccine to a higher target release potency in order to assure that the vaccine 2
contains sufficient potency to allow for the full shelf-life including an ECTC exposure, provided 3
there are no safety implications. Enhanced ECTC potential could be achieved by reducing assay 4
variability, thus also reducing the amount of potency required to account for potential errors in 5
initial potency assignment and finally, with a more accurate characterization of vaccine stability, 6
a manufacturer could reduce the amount of potency required to account for errors in the stability 7
estimates. 8
9
From a WHO CTC programme implementation perspective, it is considered that an ECTC 10
exposure potential of 40⁰C or greater for not less than three (3) days (72 hours) is required and 11
that clearly, the longer the exposure provision, the better it would be for programme 12
implementation. 13
14
When a lyophilised vaccine is being considered for ECTC applications, a stability analysis 15
should be performed for the reconstituted product using the same rigorous statistical evaluation 16
principles described in this guidance. In situations where exposure to an ECTC storage condition 17
could result in changes in the visual appearance of the lyophilised product that does not impact 18
its clinical performance, the manufacture should provide the relevant supportive data to the NRA 19
considering the ECTC label change. If approved, a description of potential change in visual 20
appearance should be included in the product leaflet/package insert. For liquid multi-dose vials, 21
additional data to support microbial effectiveness of the preservative under ECTC conditions 22
would be required prior to ECTC use. 23
24
In general, additional clinical assessment for a previously approved product which is being 25
considered in an ECTC application, should only be required where a planned ECTC exposure 26
results in a change of product specifications (e.g. lower end of shelf-life potency or a higher 27
release potency beyond existing clinical experience). Field studies, where the clinical evaluation 28
of a vaccine has been performed on a product that has been exposed to temperatures higher than 29
the approved conditions, but without potency and quality testing using the manufacturer’s 30
assays, are not considered acceptable from a regulatory perspective. Clinical studies that are 31
intended to support ECTC applications should be performed using a vaccine with known (or 32
modeled) potency at the time of the ECTC use, as determined by the manufacturer’s assays. 33
34
Page 12
5. Stability evaluation of vaccines for use under ECTC 1
2
Stability evaluation of vaccines planned for use in ECTC need to obtain sufficient scientifically 3
valid data to support regulatory approval of labelling for such use. This requires assurance that 4
there is sufficient potency available, even with lots near end-expiry, to allow for an exposure 5
under ECTC conditions. The best prediction of actual end-expiry potency of any given lot 6
depends upon a variety of factors, including release potency of the specific lot, accuracy and 7
precision of the potency assay, as well as the results of stability studies. Therefore, statistical 8
evaluation is needed in order to be able to state that, with a given (usually 95%) degree of 9
confidence, the potency after an end-expiry ECTC exposure will still be above the minimum 10
threshold needed for product efficacy. It is only through the use of statistical analysis that it is 11
possible to obtain an indication of the level of confidence in results reported at release or in 12
potencies delivered to the vaccine recipient and thus, statistical analysis is required for assurance 13
of the quality of vaccines intended for delivery in a ECTC context. Because the major ECTC-14
related concern is usually that the temperature exposure will reduce potency to unacceptable 15
levels, this discussion will focus on ensuring that the minimum required potency is maintained, 16
recognizing that similar principles may be used to assure that potentially unsafe degradation 17
products do not exceed a safe level. Because there may also be safety considerations if exposure 18
to ECTC leads to undesired degradation products, the potential for their generation during 19
ECTC should be explicitly evaluated. 20
21
Even for products that are sufficiently stable to tolerate an ECTC exposure, poorly designed 22
studies or inappropriate statistical analyses can reduce the likelihood that an ECTC exposure 23
could be justified. This section describes study design and statistical approaches that will 24
improve the likelihood that ECTC exposure can be justified using sound scientific principles. 25
26
While for most licensed products, additional clinical studies will not be required for an ECTC 27
approval, the establishment of the minimum potency specification for a specific vaccine, through 28
the initial licensing studies, is essential for all stability assessments. Changes in specifications 29
(including lowering of end-expiry potency specifications) may require supporting clinical data. 30
Thus, the data package for ECTC applications should include summaries of the initial clinical 31
studies, including the quality data of the clinical lots, to support the end of shelf-life potency 32
specification. 33
34
Page 13
The package should also include stability studies that formally demonstrate that this minimum 1
potency is exceeded throughout the dating period, including the ECTC exposure. Estimates of 2
the rate (or slope) at which the potency decays (hereafter, referred to as “stability estimates”) at 3
the normal and ECTC temperatures, along with an understanding of potential errors in those 4
estimates, are the most important outcomes of these stability studies. The reliability of these 5
stability estimates, as well as the reliability with which the release potency of any lot can be 6
determined, depends in turn upon the potency assay. 7
8
As mentioned in Section 4, stability studies to support vaccine use in an ECTC context should 9
employ the manufacturer’s potency assay, in order to preserve a connection between the released 10
product proposed for ECTC use and the original clinical material used to support product 11
efficacy. It is likely that key parameters of this assay will already be known from assay 12
validation, including the assay accuracy and precision. More reliable estimates of in-use assay 13
precision may sometimes be obtained by other means, for example, by comparing actual with 14
modelled results in the stability analyses. Other data may also be relevant to the estimation of in-15
use assay precision. Because there are several possible estimates of assay precision that could be 16
used for the purpose of ECTC-related calculations, the choice of this estimate should be 17
scientifically justified and when a clear justification cannot be made, the more conservative 18
(from an ECTC label perspective) estimate should be used. 19
20
Statistical analysis of vaccine stability is normally based on a mathematical model and supported 21
by data, that describes the kinetics of potency changes at different temperatures for different 22
periods of time. Statistical release models must support the conclusion that the mean potency of 23
vials in a given lot will meet specifications through shelf life, with a given level (usually 95%) of 24
confidence, given all stability losses including allowed storage periods outside of long term 25
storage conditions. Typically, the rate of change (generally, loss) of vaccine potency is not a 26
simple linear function of time. Log-transformation of potency data usually leads to a more 27
predictive model that permits analysis of stability data by linear regression. Thus, most cases, 28
potency data should be log-transformed before analysis and when log-transformation is not used, 29
a scientific justification in favor of a more relevant model should be provided. Log-30
transformation usually has greater biological relevance (since for most substances, the rate of 31
decay at any point in time depends on the quantity of substance present at that time) than direct 32
analysis of non-log-transformed potency data. Moreover, empirical observation supports that 33
potency decay of many vaccines follows first-order kinetics, which are linear following 34
logarithmic transformation, although low precision of the potency assay can make it more 35
Page 14
difficult to determine if stability results follow the decay model. Also, potency measurements are 1
often log-normally distributed and when this is the case, log-transformation may be required to 2
satisfy the statistical assumptions of the modeling and log-transformation can further improve 3
the precision of the stability estimate. In all cases, the decay model used should correspond to 4
actual product decay kinetics as observed in stability studies, which may support the use of non-5
log-transformed decay models (including linear models), if these models can also be justified as 6
biologically relevant. It should be noted that log-transformation is not always the best approach 7
for stability indicating assays. For example, when degradation products are measured, kinetics of 8
decay normally depend on the quantity of non-degraded material present in the sample, which 9
also can be modeled (see Appendix for an example). Visual examination of the plot of 10
transformed and untransformed stability data can provide an indication of whether mathematical 11
transformations can linearize the decay curve. Sometimes, no biologically meaningful model can 12
be identified that fits the data, as may occur when there are multiple phases in the decay kinetics. 13
In this case, decay estimates during ECTC exposures can be estimated using only beginning and 14
ending results in the stability testing. Where the most appropriate choice of model is unclear, 15
selection of the most conservative option is appropriate. 16
17
Stability studies should properly evaluate the kinetics of decay, including indicating that decay 18
rates (after transformation) are not higher at the end of the observation period than at the 19
beginning. They should include a sufficient number of time points to determine the adequacy of 20
the decay model, while also providing robust stability estimates. Linearity can theoretically be 21
supported using at least three time points; the starting point, the ending point (corresponding to 22
the desired ECTC exposure time) and ideally the midpoint. However, when assay precision is 23
not high, use of additional time points will likely be needed to increase the assurance that the 24
change in ECTC-related stability parameters truly follows a linear model. Inclusion of additional 25
time points can also improve estimates of the true rate of change in potency. If linearity has 26
already been established, more precise estimates of decay rates under ECTC conditions can be 27
obtained by testing sufficient numbers of samples at the post-ECTC exposure as compared the 28
pre-CTC exposure time points. When decay kinetics are linear, testing at time points beyond the 29
proposed ECTC use can also improve the precision of the stability estimate. It is often assumed 30
that decay rates under ECTC conditions will be similar near the time of release and at expiry, but 31
this assumption should be verified as with the rate of decay over the normal storage period. If the 32
decay rate does vary depending on the time from release, modeling may need to take this into 33
account, with consideration of the potential uncertainties added by any assumptions that are 34
made. Testing larger numbers of independent samples (batches/lots) can further improve this 35
Page 15
precision and potentially increase the likelihood that these studies will support ECTC use, but in 1
all cases, a minimum of three lots should be tested. 2
3
Typical stability evaluations often include an analysis for “poolability”, because the presence of 4
an outlier may indicate unacceptable manufacturing variability or could cause a combined decay 5
slope calculated using a small number of lots to be inaccurate. Previous guidance advocates 6
using the worst-case lot for the decay slope estimate when analysis suggests that data from tested 7
lots may not be poolable. However, using the worst-case lot can inappropriately penalize 8
expected variability (over which the sponsor has no control) and can be a disincentive to more 9
complete testing. It is reasonable to include all tested lots in the stability analysis, as long as 10
these lots are considered representative of the licensed (and “in-control”) manufacturing process 11
and a sufficient number of lots to address random variability is included. Thus, data from 10 or 12
more representative lots can be pooled for stability analysis regardless of poolability criteria. 13
Pooling of data from fewer than ten lots (without meeting poolability criteria) should be 14
statistically justified. 15
16
Stability testing normally provides information on expected rates of decay (for the linear model, 17
the decay slope) and standard error of the decay slope at “n” different temperatures of exposure 18
(modeling storage, shipping, post-reconstitution, etc.), plus under ECTC conditions (for time tCTC 19
with decay slope bCTC). Modeling of stability test results also can provide an estimate of the 20
precision of the potency assay. 21
22
A first approximation of the impact of an ECTC exposure is the 95% confidence bound on the 23
expected loss LCTC due to the ECTC exposure. This can be estimated as: 24
25
LCTC = |tCTCbCTC – z1- (tCTC*s(bCTC))| 26
27
where tCTC is the time at ECTC temperatures, bCTC is the decay slope (a negative number) at 28
ECTC temperatures, z1- α is the one sided z statistic, at the confidence level associated with the 29
desired degree of confidence (usually α=.05, for 95% confidence bounds), and s(bCTC) is the 30
standard error of the decay slope. If this amount of additional decay in potency beyond that 31
considered by the product release model is considered acceptable, the product can be accepted 32
for ECTC use. 33
34
Page 16
A more accurate and less conservative estimate can be obtained by calculating the aggregate 1
error associated with all assumptions in the decay model. The product release model (Figure 1) 2
defines the needed potency of the product at the time of release. From this information, it is 3
possible to calculate the statistical Lower Bound (LB) on the mean potency of a product that is 4
released at the minimum release potency and that is exposed to multiple temperature conditions, 5
as follows1: 6
7
LB1-α= MRP + t1 b1 + t2 b2+…+ tn bn+ tCTC bCTC - U 8
9
where 1- describes the statistical confidence level associated with the lower bound ( is 0.05 10
for 95% lower bounds), MRP is the manufacturer’s minimum allowable release potency (usually 11
log-transformed), ti is time at temperature i, bi is decay slope (a negative number, or zero if 12
positive) at temperature i, and U is the combined uncertainty associated with the independent 13
estimation of the numbers on the right side of the equation, for example: 14
15
U= z1- α x sqrt((sassay)2+( t1 s(b1))
2+( t2 s(b2))
2+…+( tn s(bn))
2+ ( tCTC s(bCTC) )
2) 16
17
Where z1- α is the one sided z statistic, at the confidence level associated with the desired degree 18
of confidence (usually α=.05, for 95% confidence bounds), sassay is the assay precision and s(bi) is 19
the precision (standard error) of the decay slope at temperature i. Thus, the expected end-expiry 20
potency is expected to be (with 95% confidence) as low as LB0.05 which accounts for the 21
manufacturer’s minimum release potency, the estimated potency losses at the various 22
temperatures and the associated uncertainty. 23
24
Without an ECTC exposure and with omission of the ECTC-associated terms, the above 25
equations yield the minimum potency that an already-licensed product is expected to maintain 26
throughout its dating period based on the manufacturer’s release model. Inclusion of the ECTC 27
term allows a reviewer to determine the degree to which potency is affected by the ECTC 28
exposure and can allow a determination of whether or not that is acceptable, based on clinical 29
experience with the vaccine at those levels of potency. 30
31
1 As noted, the equations listed are for the general case that could include modeling storage, shipping, post-
reconstitution, etc.. However, when only considering the normal storage condition and a single ECTC exposure, an
example of a simplified form of the equations (which could be applied to each of the three presented) would be:
LB1-α= MRP + t1 b1 + tCTC bCTC - U
Page 17
When the Lower Limit of potency (LL), defined as the minimum potency below which there is 1
some concern about product efficacy (considering the potency results from the clinical lots) has 2
been defined, it is preferred to rearrange the terms of the above equation to determine the 3
minimum release potency (MRP) required to maintain potency through expiry including the 4
ECTC exposure, in essence calculating the minimum amount of potency that must be added to 5
that minimum potency (LL) in order to assure product quality throughout normal storage and 6
handling including an ECTC exposure: 7
8
MRP = LL - t1 b1 - t2 b2-…- tn bn- tCTC bCTC + U 9
10
Manufacturers may also choose to release a product at higher potencies to manage their own 11
risk. This provides a convenient way to ensure that the release model will support ECTC 12
labelling. If ECTC exposure potential cannot be established, then several options could be 13
considered as outlined in Section 4 (see also the “Product Release Model” definition in Section 3 14
and the related Figure 1, in Section 4). It should be noted that the analytical principles 15
represented in the equations above are the same as those in the existing WHO vaccine stability 16
guidance (1) and have been expanded to include ECTC exposure. Additionally, the equations 17
here are not the only ways to represent these calculations and that other approaches that 18
encompass similar statistical principles could potentially be acceptable where justified. 19
20
6. Monitoring ECTC 21
22
All vaccines should be kept in the recommended long term storage conditions with appropriate 23
oversight (e.g. monitoring and recording) prior to ECTC exposure. Use of vaccines under ECTC 24
requires specific monitoring of temperature exposure and time, as well as formal procedures to 25
ensure that approved maximum temperature and time are not exceeded. Unused vaccines that 26
exceed the maximum approved temperature or time should be disposed of by suitable 27
procedures. ECTC temperature monitoring systems need to be able to distinguish vaccine that is 28
still appropriate for use from a vaccine that has exceeded the limits imposed by the data 29
supporting ECTC use. The monitoring requirements for ECTC differ from those for long term 30
storage and transport. The respective monitoring systems for long term and ECTC storage should 31
both be consistent with product stability characteristics. In order to allow an ECTC exposure, the 32
monitoring system should assure that approved long term storage conditions, especially with 33
respect to temperature, are not exceeded. At the time of vaccine approval for an ECTC, the 34
Page 18
manufacturer, the NRA and the immunization programme should work together to ensure that an 1
appropriate monitoring system is in place. 2
3
7. Suggested product labelling information for use under ECTC 4
5
Information on ECTC should be described in the product leaflet/package insert (PL/PI) to 6
provide information for medical practitioners. The statement on ECTC should be a separate 7
paragraph, in the appropriate section of the label (e.g. Storage and Handling). 8
9
The ECTC information in the product leaflet/package insert should be clear, concise and 10
specific. If the vaccine consists of two or more components (e.g. lyophilised vaccine and 11
diluent), ECTC information should be given for all components of a specific vaccine. 12
13
Information to be included in ECTC statements should consider the following, if applicable: 14
- maximum temperature; 15
- maximum time at a specific temperature; 16
- time after opening (or reconstitution or mixture), if applicable; 17
- advice on unopened vials exposed to ECTC (e.g. disposal); 18
- single ECTC exposure within the shelf-life. 19
20
Model PL/PI text for ECTC use 21
The vaccine [and its diluent/solvent or other component] may be kept for a single period of time 22
of up to [x days or x weeks or x months] at temperatures of up to [xoC] immediately prior to 23
administration. At the end of this period, the vaccine [must be disposed of]. This is not a 24
recommendation for storage but is intended to guide decision making when exposure to higher 25
temperatures is unavoidable. [After opening [or reconstitution or mixture], the vaccine can be 26
kept for [x hours or x days] at temperatures of up to [xoC] at which point it must be disposed of]. 27
28
Page 19
Authors& acknowledgments 1
2
The first draft of these guidelines was prepared by: Dr C Conrad, Paul-Ehrlich-Institut, Langen, 3
Germany; Dr E Griffiths, WHO consultant, Kingston-upon-Thames, Surrey, United Kingdom; 4
Mrs T Jivapaisarnpong, Department of Medical Sciences, Bangkok, Thailand; Dr J Kim, 5
Department of Essential Medicines and Health Products, World Health Organization, Geneva, 6
Switzerland; Dr I Knezevic, Department of Essential Medicines and Health Products, World 7
Health Organization, Geneva, Switzerland; Dr P Krause, Center for Biologics Evaluation and 8
Research, Food and Drug Administration, Bethesda, Maryland, United States of America; Dr J 9
Shin, Expanded Programme on Immunization, Regional Office for the Western Pacific, World 10
Health Organization, Manilla, Philippines; Dr D Smith, Centre for Biologics Evaluation, Health 11
Canada, Ottawa, Ontario, Canada; Dr J Southern, Adviser to Medicines Control Council of 12
South Africa, Cape Town, South Africa; Dr T Wu, Centre for Biologics Evaluation, Health 13
Canada, Ottawa, Ontario, Canada; taking into account comments received from: 14
Dr B D Akanmori, Regional Office for Aftrica, World Health Organization, Brazzaville, Congo; 15
Ms A-L Kahn, Expanded Programme on Immunization, World Health Organization, Geneva, 16
Switzerland; Dr A Meek, Department of Essential Medicines and Health Products, World 17
Health Organization, Geneva, Switzerland; Dr C A Rodriguez Hernadez, Essential Medicines 18
and Health Products, World Health Organization, Geneva, Switzerland; Dr S C Da Silveira, 19
Agencia Nacional da Vigilancia Sanitaria, Ministerio da Saude Esplanada dos Ministerios, 20
Brasilia, Brazil; Ms S Zipursky, Expanded Programme on Immunization, World Health 21
Organization, Geneva, Switzerland. 22
Acknowledgement is given to Mr M Walsh, Centre for Evaluation of Radiopharmaceuticals and 23
Biotherapeutics, Health Canada, Ottawa, Ontario, Canada for providing expertise on the 24
mathematical modelling and statistical approach of the section on product specific CTC 25
evaluation of a model of monovalent polysaccharide conjugate vaccine. 26
27
This first draft was based on the Ottawa and Langen CTC consultation reports prepared by: Dr 28
M Baca-Estrada, Centre for Biologics Evaluation, Health Canada, Ottawa, Ontario, Canada; Dr 29
C Conrad, Paul-Ehrlich-Institut, Langen, Germany; Dr E Griffiths, WHO consultant, Kingston-30
upon-Thames, Surrey, United Kingdom; Dr J Kim, Department of Essential Medicines and 31
Health Products, World Health Organization, Geneva, Switzerland; Dr I Knezevic, Department 32
of Essential Medicines and Health Products, World Health Organization, Geneva, Switzerland; 33
Dr P Krause, Center for Biologics Evaluation and Research, Food and Drug Administration, 34
Bethesda, Maryland, United States of America; Dr M (Ferguson) Lennon, Horning, United 35
Kingdom; Dr H Meyer, Paul-Ehrlich-Institute, Langen,Germany; Dr V Oeppling, Paul-Ehrlich-36
Institute, Langen, Germany; Dr M Pfleiderer, Paul-Ehrlich-Institute, Langen, Germany; Dr J 37
Shin, Department of Essential Medicines and Health Products, World Health Organization, 38
Geneva, Switzerland; Dr D Smith, Centre for Biologics Evaluation, Health Canada, Ottawa, 39
Ontario, Canada; Dr R Wagner, Paul-Ehrlich-Institute, Langen, Germany; Dr T Wu, Centre for 40
Biologics Evaluation, Health Canada, Ottawa, Ontario, Canada; Ms S Zipursky, Expanded 41
Programme on Immunization, World Health Organization, Geneva, Switzerland. 42
43
Acknowledgments are extended to the following participants dedicated to presentation and 44
discussion at Ottawa, Canda 4-6 December 2012 and Langen, Germany 4-6 June 2013: 45
Page 20
Dr M-C Annequin, Agence nationale de sécurité du médicament et des produits de santé, Saint-1
Denis, France; Dr M Baca-Estrada, Centre for Biologics Evaluation, Health Canada, Ottawa, 2
Ontario, Canada; Dr K Brusselmans, Scientific Institute of Public Health, Brussels, Belgium; Dr 3
M Chultem, Biologics and Genetic Therapies Directorate, Health Canada, Ottawa, Canada; Dr C 4
Cichutek, Paul-Ehrlich-Institute, Langen, Germany; Mr W Conklin, Merck, Whitehouse Station, 5
New Jersey, United States of America; Dr C Conrad, Paul-Ehrlich-Institute, Langen, Germany; 6
Ms D Doucet, GlaxoSmithKline Biologicals SA, Bruxelles, Belgium; Dr W Egan, Novartis, 7
Columbia, Maryland, United States of America; Dr L Elmgren, Centre for Biologics Evaluation, 8
Health Canada, Ottawa, Ontario, Canada; Dr D Felnerova, Crucell, Bern, Switzerland; Dr S 9
Gairola, Serum Institute of India Ltd., Pune, India; Dr E Griffiths, WHO consultant, Kingston-10
upon-Thames, Surrey, United Kingdom; Dr D A Hokama, BioManguinhos, Rio de Janeiro, 11
Brazil; Dr W Huang, Xiamen Innovax Biotech Co., Xiamen, People’s Republic of China; Mrs T 12
Jivapaisarnpong, Department of Medical Sciences, Bangkok, Thailand; Dr J Kim, Essential 13
Medicines and Health Products, World Health Organization, Geneva, Switzerland; Dr J 14
Korimbocus, Agence nationale de sécurité du médicament et des produits de santé, Lyon, 15
France; Dr P Krause, Center for Biologics Evaluation and Research, Food and Drug 16
Aministration, Bethesda, Maryland, United States of America; Dr H Langar, Regional Office for 17
the Eastern Mediterranean, World Health Organization, Cairo, Egypt; Dr A Laschi, Sanofi 18
Pasteur, Lyon, France; Dr C Lecomte, GlaxoSmithKline Vaccines, Wavre, Belgium; Dr M 19
(Ferguson) Lennon, Horning, United Kingdom; Prof H Leng, Somerset West, South Africa; Ms 20
A Lopez, Biológicos y Reactivos de México S.A. de C.V., Col Popotla, Mexico; Dr A Luethi, 21
Crucell Switzerland AG, Bern, Switzerland; Dr W Matheis, Paul-Ehrlich-Institute, Langen, 22
Germany; Dr D G Maire, Department of Essential Medicines and Health Products, World 23
Health Organization, Geneva, Switzerland; Dr A Merkle, Paul-Ehrlich-Institute, Langen, 24
Germany; Dr H Meyer, Paul-Ehrlich-Institute, Langen, Germany; Dr B L M Moreira, Agencia 25
Nacional da Vigilancia Sanitaria Secretaria de Vigilancia Sanitaria, Brasilia, Brazil; Dr K-T 26
Nam, National Institute of Food and Drug Safety Evaluation, Ministry of Food and Drug Safety, 27
Cheongju, Republic of Korea; Dr R Nibbeling, Institute for Translational Vaccinology, AL 28
Bilthoven, Netherlands; Dr V Oeppling, Paul-Ehrlich-Institute, Langen, Germany; Dr D Mora 29
Pascual, Centro para el Control 27 Estatal de la Calidad de los Medicamentos, Havana, Cuba; Dr 30
M Pfleiderer, Paul-Ehrlich-Institute, Langen, Germany; Dr M L Pombo, Pan American Health 31
Organization, World Health Organization, Washington DC, United States of America; Dr T 32
Prusik, Temptime Corporation, Morris Plains, New Jersey, United States of America; Dr M 33
Reers, Biological E Ltd, Azamabad, Hyderabad, India; Dr C Rodriguez HerNadez, Department 34
of Essential Medicines and Health Products, World Health Organization, Geneva, Switzerland; 35
Mr T Schofield, Medimmune, Gaithersburg, Maryland, United States of America; Ms F A 36
Setyorini, Biofarma, Bandung, Indonesia; Dr S Shani, Central Drugs Standard Control 37
Organisation, FDA, New Delhi, India; Dr I S Shin, National Institute of Food and Drug Safety 38
Evaluation, Ministry of Food and Drug Safety, Cheongju, Republic of Korea; Dr J Shin, 39
Department of Essential Medicines and Health Products, World Health Organization, Geneva, 40
Switzerland; Dr S C Da Silveira, Agencia Nacional da Vigilancia Sanitaria, Ministerio da Saude 41
Esplanada dos Ministerios, Brasilia, Brazil; Dr D Smith, Centre for Biologics Evaluation, Health 42
Canada, Ottawa, Ontario, Canada; Dr M Vega, Centre for Genetic Engineering and 43
Biotechnology, Havana, Cuba; Dr R Wagner, Paul-Ehrlich-Institute, Langen, Germany; Dr T 44
Wu, Centre for Biologics Evaluation, Health Canada, Ottawa, Ontario, Canada; Ms S Wong, 45
Vaccine and Biological Stability, Merck, Whitehouse Station, New Jersey, United States of 46
Page 21
America; Dr M Zeng, National Institutes for Food and Drug Control, Beijing, People’s Republic 1
of China; Ms S Zipursky, Expanded Programme on Immunization, World Health Organization, 2
Geneva, Switzerland. 3
4
The second draft of these guidelines was prepared by Dr C Conrad, Paul-Ehrlich-Institut, 5
Langen, Germany; Dr I Feavers, National Institute for Biological Standards and Control, Potters 6
Bar, United Kingdom; Dr E Griffiths, WHO consultant, Kingston-upon-Thames, Surrey, United 7
Kingdom; Mrs T Jivapaisarnpong, Department of Medical Sciences, Bangkok, Thailand; Dr J 8
Kim, Department of Essential Medicines and Health Products, World Health Organization, 9
Geneva, Switzerland; Dr I Knezevic, Department of Essential Medicines and Health Products, 10
World Health Organization, Geneva, Switzerland; Dr P Krause, Center for Biologics Evaluation 11
and Research, Food and Drug Administration, Bethesda, Maryland, United States of America; 12
Dr J Shin, Expanded Programme on Immunization, Regional Office for the Western Pacific, 13
World Health Organization, Manilla, Philippines; Dr D Smith, Centre for Biologics Evaluation, 14
Health Canada, Ottawa, Ontario, Canada; Dr J Southern, Adviser to Medicines Control Council 15
of South Africa, Cape Town, South Africa; Mr M Walsh, Centre for Evaluation of 16
Radiopharmaceuticals and Biotherapeutics, Health Canada, Ottawa, Ontario, Canada; Dr T Wu, 17
Centre for Biologics Evaluation, Health Canada, Ottawa, Ontario, Canada; taking into account 18
comments received from: 19
Dr B D Akanmori, Regional Office for Africa, World Health Organization, Brazzaville, Congo; 20
Dr A Alsalhani, Access Campaign/Médecins Sans Frontières, Paris, France; Dr M-C Annequin, 21
Agence nationale de sécurité du médicament et des produits de santé, Saint-Denis, France; Dr B 22
Bolgiano, National Institute for Biological Standards and Control, Potters Bar, United Kingdom; 23
J Bridgewater, Center for Biologics Evaluation and Research, Food and Drug Administration, 24
Bethesda, United States of America; Dr A Chang, M.S.E Bioengineering Innovation and Design 25
Candidate, Johns Hopkins University; Dr A Cheung, Center for Biologics Evaluation and 26
Research, Food and Drug Administration, Bethesda, United States of America; Ms D Doucet, 27
GlaxoSmithKline Biologicals SA, Brussels, Belgium; Dr J Du, National Institutes for Food and 28
Drug Control, Beijing, People’s Republic of China; Dr W Egan, Novartis, Columbia, Maryland, 29
United States of America; Dr S Gagneten, Center for Biologics Evaluation and Research, Food 30
and Drug Administration, Bethesda, United States of America; Dr D Garcia, Agence nationale 31
de sécurité du médicament et des produits de santé, Lyon, France; Dr E Griffiths, WHO 32
consultant, Kingston-upon-Thames, Surrey, United Kingdom; Dr T Guo, National Institutes for 33
Food and Drug Control, Beijing, People’s Republic of China; Mr K Hicks, Sanofi Pasteur, Lyon, 34
France (Consolidation on behalf of IFPMA); Ms A Juan-Giner, Epicentre/Médecins Sans 35
Frontières, Paris, France; Ms A-L Kahn, Expanded Programme on Immunization, World Health 36
Organization, Geneva, Switzerland; Dr U Kartoglu, Department of Essential Medicines and 37
Health Products, World Health Organization, Geneva, Switzerland; Dr B-G Kim, National 38
Institute of Food and Drug Safety Evaluation, Ministry of Food and Drug Safety, Cheongju, 39
Republic of Korea; Dr J Korimbocus, Agence nationale de sécurité du médicament et des 40
produits de santé, Lyon, France; Dr T-L Lin, Center for Biologics Evaluation and Research, 41
Food and Drug Administration, Bethesda, United States of America; Dr D Petit, Expanded 42
Programme on Immunization, World Health Organization, Geneva, Switzerland; Dr M Ramos, 43
Public Health England, London, UK; Mr T Schofield, Medimmune, Gaithersburg, Maryland, 44
United States of America; Dr S C Da Silveira Andreoli, Agencia Nacional da Vigilancia 45
Sanitaria, Ministerio da Saude Esplanada dos Ministerios, Brasilia, Brazil; Dr D Smith, Centre 46
Page 22
for Biologics Evaluation, Health Canada, Ottawa, Ontario, Canada; Dr T Prusik, Temptime 1
Corporation, Morris Plains, New Jersey, United States of America; Dr J Shin, Regional Office 2
for the Western Pacific, World Health Organization, Manila, Philippines; Ms S Zipursky, 3
Expanded Programme on Immunization, World Health Organization, Geneva, Switzerland. 4
5
Acknowledgments are extended to the following participants dedicated to presentation and 6
discussion at Geneva, Switzerland 24 March 2015: 7
Mr F S Adeyemi, Drug Evaluation and Research Directorate, National Agency for Food and 8
Drug Administration and Control, Abuja, Nigeria; Dr M Allin, Worldwide Regulatory Strategy, 9
Pfizer, Brussels, Belgium; Ms Y Bai, Center for Drug Evaluation, China Food and Drug 10
Administration, Beijing, Peoples Republic of China; Dr J Bergers, National Institute for Public 11
Health and the Environment, Center of Biological Medicines and Medical Technology, 12
Bilthoven, Netherlands; Ms P Carneiro, Institute Butantan, Sao Paulo, Brazil; Dr C Conrad, 13
Paul-Ehrlich-Institut, Langen, Germany; Ms D Doucet, GlaxoSmithKline Biologicals SA, 14
Brussels, Belgium; Ms N Dubois, Pfizer, Brussels, Belgium; Dr A Fauconnier, Federal Agency 15
for Medicines and Health Products, Brussels, Belgium; Dr I Feavers, National Institute for 16
Biological Standards and Control, Potters Bar, United Kingdom; Dr S Gairola, Serum Institute of 17
India, Hadapsar, Pune, India; Mr K Gopinathan, Biological E Ltd, Azamabad, Hyderabad, India; 18
Dr E Griffiths, Kingston-upon-Thames, Surrey, United Kingdom; Dr K Hicks, Sanofi Pasteur, 19
Lyon, France; Ms A-L Kahn, Immunization, Vaccines and Biologicals Department, World 20
Health Organization, Geneva, Switzerland; Dr U Kartoglu, Department of Essential Medicines 21
and Health Products, World Health Organization, Geneva, Switzerland; Mr Y Kaushik, Bharat 22
Biotech International, Hyderabad, India; Dr B-G Kim, National Institute of Food and Drug 23
Safety Evaluation, Ministry of Food and Drug Safety, Cheongju, Republic of Korea; Dr J Kim, 24
Department of Essential Medicines and Health Products, World Health Organization, Geneva, 25
Switzerland; Dr I Knezevic, Department of Essential Medicines and Health Products, World 26
Health Organization, Geneva, Switzerland; Dr J Korimbocus, Agence nationale de sécurité du 27
médicament et des produits de santé, Lyon, France; Dr P Krause, Center for Biologics 28
Evaluation and Research, Food and Drug Administration, Bethesda, United States of America; 29
Mr A Kukrety, Food and Drug Administration, New Delhi, India; Dr J S Lloyd, Immunization 30
Supply Chain, Ferney Voltaire, France; Dr B L M Moreira, Agencia Nacional da Vigilancia 31
Sanitaria Secretaria de Vigilancia Sanitaria, Ministerio da Saude Esplanada dos Ministerios, 32
Brasilia-DF, Brazil; Dr A Luethi, Crucell Switzerland AG, Bern, Switzerland; Dr A Meek, 33
Department of Essential Medicines and Health Products Department, World Health 34
Organization, Geneva, Switzerland; Dr D M Pascual, CECMED, Habana, Cuba; Mr C Perrin, 35
Médecins Sans Frontières, Paris, France ; Dr T Prusik, Temptime Corporation, New Jersey, 36
United States of America; Ms C Rodriguez Hernandez, Department of Essential Medicines and 37
Health Products, World Health Organization, Geneva, Switzerland; Mrs J Rogers, Biological 38
Products Unit, Food and Drugs Authority, Ghana; Ms C Schmit, GlaxoSmithKline Biologicals 39
SA, Brussels, Belgium; Ms P Patricia Said, Bio Farma, Bandung, Indonesia; Mr T Schofield, 40
Medimmune, Gaithersburg, United States of America; Dr D Smith, Biologic and Genetic 41
Therapies Directorate, Health Canada, Ottawa, Ontario, Canada; Dr J Southern, Adviser to 42
Medicines Control Council of South Africa, Cape Town, South Africa; Ms S Wong, Merck & 43
Co. New Jersey, United States of America; Dr T Wu, Biologic and Genetic Therapies 44
Directorate, Health Canada, Ottawa, Ontario, Canada. 45
Page 23
References 1
2
1. Guidelines on stability evaluation of vaccines. Annex 3 in: WHO Expert Committee 3
on Biological Standardization. Fifty-seventh report. Geneva, World Health 4
Organization, 2011 (WHO Technical Report Series, No. 962). 5
http://who.int/biologicals/vaccines/stability_of_vaccines_ref_mats/en/index.html 6
2. WHO Strategic Advisory Group of Experts on Immunization (SAGE) April 2012 7
http://www.who.int/immunization/sage/meetings/2012/april/presentations_backgroun8
d_docs/en/index.html 9
3. WHO Immunization Practices Advisory Committee (IPAC) 2012 10
http://www.who.int/immunization/policy/committees/IPAC_2012_April_report.pdf 11
4. McCarney S, Zaffran M. Controlled Temperature Chain: the New Term for Out of the 12
Cold Chain. Ferney: PATH; 2009. www.technet-13
21.org/index.php/resources/documents/doc_download/725-controlled-temperature-14
chain-the-new-term-for-out-of-the-cold-chain 15
5. Controlled Temperature Chain, Vaccine management and logistics – logistics support, 16
http://www.who.int/immunization/programmes_systems/supply_chain/resources/tool17
s/en/index6.html 18
6. D Butler. Vaccines endure African temperatures without damage: Anti-meningitis 19
campaign in Benin delivers more than 150,000 doses with no losses from excess heat. 20
http://www.nature.com/news/vaccines-endure-african-temperatures-without-damage-21
1.14744#/ref-link-1 22
7. S Zipursky, M Djingarey, J-C Lodjo, L Olodo, S Tiendrebeogo, and Oliver 23
Ronveaux. Benefits of using vaccines out of the cold chain: Delivering Meningitis A 24
vaccine in a controlled temperature chain during the mass immunization campaign in 25
Benin 2014, Vaccines; 2014, 32:1431-1435 26
8. Use of MenAfriVac™ (meningitis A vaccine) in a controlled temperature chain 27
(CTC) during campaigns, 28
http://www.who.int/immunization/documents/WHO_IVB_13.04_5_6/en/ 29
9. Meeting report of WHO/Health Canada Drafting Group Meeting on Scientific and 30
Regulatory Considerations on the Stability Evaluation of VAccines under controlled 31
temperature Chain, Ottawa, Canada, 4-6 December 2012 32
http://who.int/biologicals/areas/vaccines/CTC_FINAL_OTTAWA_Web_Meeting_re33
port_25.11.2013.pdf?ua=1 34
10. Meeting report of WHO / Paul-Ehrlich-Institut Informal Consultation on Scientific 35
and Regulatory Considerations on the Stability Evaluation of Vaccines under 36
Controlled Temperature Chain, Paul-Ehrlich-Institut, Langen, Germany, 4-6 June 37
2013 http://who.int/biologicals/vaccines/CTC_Final_Mtg_Report_Langen.pdf 38
11. Philip R. Krause. Goals of stability evaluation throughout the vaccine life cycle. 39
Biologicals; 2009, 37: 369-378. 40
12. Timothy L. Schofield. Vaccine stability study design and analysis to support product 41
licensure. Biologicals; 2009, 37: 387-396. 42
43
Page 24
Appendix: Product Specific ECTC Evaluation of a Model 1
Monovalent Polysaccharide Conjugate Vaccine 2
3
The model vaccine and the stability data presented in this section were developed based on 4
Health Canada’s overall experience with conjugate vaccines and do not represent characteristics 5
or data from any specific product. The analysis presented is also applicable to other stability-6
indicating parameters, such as vaccine potency. The vaccine example under evaluation is a 7
monovalent conjugate vaccine composed of purified capsular polysaccharide (PS) covalently 8
attached to diphtheria toxoid (DT) protein. The final vaccine product is a non-adjuvanted liquid 9
formulation and presented in single-dose vials. The normal storage temperature for this model 10
conjugate vaccine is 2-8ºC, with a dating period of three years and the temperature under 11
consideration for the ECTC application is 40ºC. The quality attributes monitored in routine 12
stability studies intended for licensure included total PS, free PS, molecular size distribution, free 13
protein, pH and sterility. Free PS is considered a key stability indicating attribute for this 14
polysaccharide conjugate vaccine, since in general this parameter is linked to the clinical 15
performance of this type of vaccine. The specification for free PS for this model conjugate 16
vaccine was set as “Not More Than (NMT) 15%” at release and “Not More Than 25%” at the 17
end of the shelf-life. A review of manufacturing data indicated that, at release, 90% of the 18
commercial lots contained less than 10% of free PS and 10% of lots contained free PS in the 19
range of 10-13%. In addition, vaccine lots containing 5-25% free PS were shown to be safe and 20
immunogenic in clinical studies. 21
22
Stability data 23
Real time and real condition stability studies were conducted to establish the shelf-life under 24
normal storage conditions (2-8ºC) and to support the ECTC application. Although a minimum of 25
three lots are required for statistical modelling, analysis of larger data set (more lots) leads to 26
more accurate estimates. In this example, routine stability monitoring tests were performed for 27
four commercial vaccine lots stored at 2-8ºC and for an additional four commercial lots stored at 28
40ºC. In addition, O-acetyl content, NMR spectrum and immunogenicity (rSBA and IgG) in a 29
mouse model were also performed to characterize vaccine lots exposed to 40ºC condition. 30
Analysis of routine monitoring data revealed that total PS, molecular size distribution, free 31
protein and pH were stable for all lots stored under the 2-8ºC and 40ºC conditions. However, an 32
increase of free PS was observed for all lots as summarized in Tables 1 and 2. 33
Table 1: Summary of free PS content at 2-8ºC 34
Page 25
Lot # Free PS (Not More Than 25%)
0 3M 6M 9M 12M 18M 24M 30M 36M
1 7.53 9.58 10.73 11.17 12.54 13.51 16.07 NT 16.05
2 7.01 9.36 10.77 10.32 10.59 11.92 14.60 14.56 15.03
3 2.38 6.01 8.13 7.46 8.94 9.37 10.08 11.09 10.88
4 5.71 6.15 7.85 7.77 9.02 10.87 14.37 12.21 13.77
M: month. NT: not tested. 1
2
Table 2: Summary of free PS content at 40ºC 3
Lot # Free PS (Not More Than 25%)
0 1W 2W 3W 4W 6W 8W 10W 12W
5 2.01 2.38 4.81 6.18 9.39 11.39 13.55 13.67 13.88
6 1.74 5.71 4.64 5.37 8.16 9.08 9.98 11.77 14.37
7 5.43 10.48 10.49 10.59 13.94 15.35 15.66 15.57 16.88
8 5.21 8.05 9.45 9.05 12.71 15.10 15.72 15.73 17.26
NMT: not more than. W: week. 4
5 Statistical analysis 6
In this example, the increase in the stability-indicating free PS is due to the hydrolysis of bound 7
PS. The rate of free PS increase is the same as the rate of the decrease of bound PS. Therefore, 8
the bound PS at each test point can be calculated from the free and total PS based on mass 9
balance. The hydrolysis of bound PS can be analyzed as a first order reaction, at a decay rate that 10
is proportional to the concentration of bound PS. An initial analysis demonstrated that the free 11
PS data did not fit a linear regression well, with or without log transformation (plots not 12
provided). Therefore, the rate of free PS increase was analyzed indirectly through the modelling 13
of bound PS and logarithmic transformation of the bound PS content at different test points 14
yielded data that was more amenable to linear regression analysis. Thus, free polysaccharide data 15
obtained in stability studies was converted to percent bound polysaccharide (Tables 3 and 4), and 16
then subjected to log transformation. A release model was developed to characterize the 17
relationship between bound PS at release and end-expiry, permitting evaluation of potential 18
ECTC use. 19
Table 3: Summary of bound PS content at 2-8ºC 20
Lot # Bound PS (NLT 75%)
0 3M 6M 9M 12M 18M 24M 30M 36M
1 92.47 90.42 89.27 88.83 87.46 86.49 83.93 NT 83.95
2 92.99 90.64 89.23 89.68 89.41 88.08 85.40 85.44 84.97
3 97.62 93.99 91.87 92.54 91.06 90.63 89.92 88.91 89.12
4 94.29 93.85 92.15 92.23 90.98 89.13 85.63 87.79 86.23
M: month. NT: not tested. 21 22
Table 4: Summary of bound PS content at 40ºC 23
Page 26
Lot # Bound PS (Not Less Than 75%)
0 1W 2W 3W 4W 6W 8W 10W 12W
5 97.99 97.62 95.19 93.82 90.61 88.61 86.45 86.33 86.12
6 98.26 94.29 95.36 94.63 91.84 90.92 90.02 88.23 85.63
7 94.57 89.52 89.51 89.41 86.06 84.65 84.34 84.43 83.12
8 94.79 91.95 90.55 90.95 87.29 84.90 84.28 84.27 82.74
W: week. 1
2
Statistical analysis was performed using R, version 3.1.1 (1) to estimate the loss of bound PS 3
under both the 2–8°C and 40°C storage conditions with 95% confidence. The analysis was 4
undertaken in the following steps: 5
1. For each stability lot, the percentage of bound PS at each test point was log transformed 6
and the slope was calculated using a linear regression model. Plots of the linear 7
regression fit for all stability lots are presented in Figure 1. 8
2. Lots 1, 2, 3 and 4, monitored at 2-8°C, were assessed with respect to slope variability, 9
which was considered acceptable to use the linear regression model with a pooled (mean) 10
slope for all four lots. The same analysis was also applied to the data set (lot 5, 6, 7 and 11
8) at 40ºC, which supported the use of a pooled slope. 12
3. The assay variability (Sassay) was calculated as the standard deviation of the residuals 13
using the regression fit for the corresponding data set. Residuals were calculated using 14
the following formula: 15
Residual = observed bound PS – predicted bound PS using the pooled slope. 16
4. The uncertainty was calculated using formula described in Section 5 and two examples 17
are provided below: 18
- The uncertainty (U) at 2-8°C for 36 months = z0.95 ∙ sqrt[(sassay)2+( t2-8 ∙ s(b2-8)]
2 19
= 1.644854 ∙ sqrt[0.012728112 + (36 ∙ 0.0001845235)
2] = 0.02361 20
- The uncertainty (U) at 2-8°C for 36 months followed by 3 days at 40°C 21
= z0.95 ∙ sqrt[(sassay)2+( t2-8 ∙ s(b2-8))
2 + ( tCTC s(bCTC)
2)) 22
= 1.644854 ∙ sqrt [0.012728112 + (36 ∙ 0.0001845235)
2 + (0.1 ∙ 0.007765961)
2 ] 23
= 0.02366 24
5. The decay of bound PS was estimated using the linear regression model and the formula 25
provided in Section 5. An example is provided below to illustrate the calculation of the 26
total decay of bound PS at 2-8°C over 36 month, plus 3 days at 40ºC. This is based on the 27
worst case lots, which contain 85% bound PS at release. 28
- First step: the log transformed total decay of bound PS is: 29
= (t2-8 ∙ b2-8) + (tCTC ∙ bCTC) - U 30
= 36 ∙ (-0.002430492) + 0.1 ∙ (-0.074298917) - 0.02365824 = - 0.118585836 31
Page 27
- Second step: the logtransformed bound PS at end of storage (2-8°C plus ECTC) 1
is: 2
= loge(bound PS at release) - [(t2-8 ∙ b2-8) + (tCTC ∙ bCTC) + U] 3
= loge85 - 0.118585836 = 4.32406542 4
- Third step: bound PS at the end of storage (2-8°C plus ECTC) = e4.32406542
= 5
75.4949 6
- Forth step: the decay of bound PS at 2-8°C over 36 month plus 3 days at 40ºC is: 7
= bound PS at release – bound PS at the end of storage (2-8°C plus ECTC) 8
= 85 - 75.4949=9.5051 9
Figure 1. : Plot of bound polysaccharide data for four conjugate vaccine lots 10
11
Table 5: Summary of statistical analysis of bound PS data at 2-8°C 12
Data set Shelf-life
(month)
Pooled slope
(per month) s(b2-8)* Sassay ** U***
Decay of bound
PS (%)
4 lots (0-24M) 24 -0.003311 0.0002765 0.01146 0.02178 8.1851
4 lots (0-36M) 36 -0.002430 0.0001845 0.01273 0.02361 8.9388
*: standard error of slope. 13 **: assay variability, estimated as standard deviation of residuals. 14 ***: combined uncertainty, calculated using the formula described in Section of 5 of this document. 15
16
Table 6: Summary of statistical analysis of bound PS data at 2-8°C and 40°C 17
Data set Pooled slope
(per month) s(b40)*
Months
at 2-8ºC
Days at
40ºC Sassay** U***
Decay of
bound PS (%)
4 lots (0-12W) -0.04482 Not done Not done Not done Not done Not done Not done
4 lots (0-4W) -0.07430 0.007766 24 7 0.01146 0.02199 9.5207
-0.07430 0.008625 36 3 0.01273 0.02366 9.5051
*: standard error of slope. 18 **: assay variability, estimated as standard deviation of residuals. 19 ***: combined uncertainty, calculated using the formula described in Section of 5 of this document. 20 21
2-8
4.45
4.50
4.55
0 10 20 30Time (Months)
log
e(B
ou
nd
PS
)
lot1234
40
4.40
4.45
4.50
4.55
0 1 2Time (Months)
log
e(B
ound P
S)
lot5678
Page 28
Examination of the plots presented in Figure 1. indicate that, at each of these temperature 1
conditions, the decay of bound PS for each lot appears to be better-modelled by a straight line 2
when compared to the modelling of free PS. This supports the use of a linear regression model of 3
bound PS in this case. However, it is also noted that the estimated slope of bound PS decay is 4
different over different study periods, especially under 40ºC. The decay rate is higher over four 5
weeks (-0.07430/week) when compared to that over 12 weeks (-0.04482 /month) as is shown in 6
Table 6. Differences in rates of change for key quality attributes over a product’s shelf-life are 7
not uncommon, therefore, it is import to highlight the need to characterize trends when modeling 8
the data and the need to estimate the rate of change based on real time data over the full study 9
period. 10
11
Due to limited data points at 40ºC, the rate of bound PS decay used for ECTC application was 12
based on the modelling of -week data set, and a conservative approach was taken to limit the 13
total decay of bound PS to slightly below 10%. The statistical analysis summarized in Table 6 14
revealed an estimated 9.5% loss of bound PS (equal to the increase of free PS) after 36 months 15
storage at 2-8°C followed by three days at 40°C, or 24 months storage at 2-8°C followed by 16
seven days at 40°C. 17
18
Conclusion 19
The application of the “product release model” to the analysis of free PS can be summarized as: 20
Release specification (15%) = End of shelf-life specification (25%) 21
- Estimated combined increase at 2-8ºC and 40ºC 22
(upper bound with 95% confidence level) 23
24
Based on statistical modelling and product-related information, the following can be concluded: 25
- Different specifications for release and end of shelf-life should be established for free PS 26
for this model conjugate vaccine. A specification of “Not More Than 25%” at the end of 27
shelf-life is considered acceptable based on clinical lots shown to be safe and 28
immunogenic in clinical studies. A release specification of “NMT 15%” was considered 29
appropriate based on manufacturing capability, which ensures a high compliance rate for 30
commercial lots at release. 31
- A 36-month shelf-life at 2-8 ºC was determined to be appropriate for this model 32
conjugate vaccine. This conclusion ensures that “worst case lots”, which contain the 33
highest level of free PS permitted by the release specification (NMT 15%), plus the 34
accumulation of free PS during the storage period (8.9388%), comply with the end of the 35
shelf-life specification (NMT 25%). 36
Page 29
- A single storage period of three days at 40ºC, prior to immunization, was considered 1
acceptable, as the free PS of the worst case lots, which contain 15% free PS at release and 2
stored nearly 36 months at 2-8ºC followed by an exposure of three days at 40 ºC, are 3
expected to contain a total of 24.5051% free PS. 4
- If longer than three days at 40ºC is needed, shortening the shelf-life at 2-8 ºC from 36 5
months to 24 months would allow a single storage period of seven days at 40ºC. 6
Alternatively, the release specification of “NMT 15% free PS” might be tightened (e.g. 7
“NMT 13% free PS”), based on additional manufacturing experience, to allow longer 8
than three days of ECTC exposure while maintaining a three year shelf-life at 2-8°C. 9
10
As explained in section 4 of this document, clinical testing of a vaccine stored under ECTC 11
conditions might not be necessary as long as a battery of stability monitoring tests provides 12
sufficient assurance that the critical quality attributes (e.g. potency) of the vaccine meet the 13
specifications supported by the quality of the clinical lots. 14
15
Other than routine stability monitoring assays, the following characterization tests were also 16
performed to assess quality attributes related to vaccine clinical performance after this model 17
conjugate vaccine was stored at 40ºC for 12 weeks: 1) O-acetyl content remained stable; 2) PS 18
structure was confirmed using nuclear magnetic resonance (NMR); 3) Carrier protein integrity 19
was confirmed by the results of an in vivo immunogenicity test. It was noted that the antigen 20
dose used to immunize the mice was within the dose response curve, indicating that the in vivo 21
test was of acceptable sensitivity. In conclusion, the available stability data sets assessing critical 22
quality attributes were considered sufficient to support a single storage period of 3-day ECTC 23
(40ºC) application of this model conjugate vaccine, within the approved 3-year shelf-life at 2-24
8ºC. 25
26
Reference 27
1. R Core Team (2014). R: A language and environment for statisstical computing R Foundation 28
for Statistical Computing, Vienna, Austria. URL http://www.R-project.org/. 29