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: gaok@who.int. 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 bookorders@who.int). 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 permissions@who.int). 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