WORLD HEALTH ORGANIZATION TLAC Members present:

DEPT: IMMUNIZATIONS, VACCINES Dr B.G. Weniger (chair)

AND BIOLOGICALS Prof J. Colton
Mr R. Davis
Dr S. Deeks
TECHNOLOGIES & LOGISTICS
Prof G. Hammond
ADVISORY COMMITTEE (TLAC) Mr M. Munck
11 - 12 March 2009 Ms D. Phillips
Dr Piyanit T.
Mr S. Spanner
Mr R. Steinglass
Meeting report and recommendations Dr O. Benes

1. Opening introduction ................................................................................ 2

2. Member introduction and information ......................................................... 2
3. Use of Vaccine Vial Monitors (VVMs) in new policies ...................................... 3
3.1 Reviews of VVM development and information .............................................3
3.2 Preliminary session on VVM background and data ........................................ 3
3.3 Invited presentations on VVMs and related issues ........................................ 4
3.3.1 Vaccine behaviour at different temperatures and the use, methods, and
limitations of potency assays ..................................................................... 4
3.3.2 VVM science and technology ...................................................................... 7
3.3.3 History of VVM development, quality control, and application; Policy
considerations for VVMs and experience in the EPI ..................................... 11
3.3.4 VVM validation and lot acceptance ............................................................ 15
3.3.5 Alternative Time-Temperature Indicator (TTI) technologies: mechanisms,
advantages, and disadvantages ............................................................... 16
3.3.6 Summary of TLAC discussion of the use of VVMs in the MDVP revision and
the OCC strategy, including needs for further clarification or research ........... 19
4. Use of vaccines out of the cold chain (OCC) ................................................ 19
4.1 The collaborative methodology used for OCC studies .................................. 19
4.2 Vaccines Out of the Cold Chain: An update on the laboratory studies ............ 20
4.3 Rethinking the cold chain........................................................................ 22
4.3.1 Moving vaccines from cold chain to “cool room” ......................................... 22
4.3.2 Increasing cold-chain flexibility at health centre level ................................. 22
4.4 Summary of OCC session and TLAC discussion and recommendations for
OCC pending working group deliberations ................................................. 23
5. Revision of the multi-dose vial policy (MDVP).............................................. 23
5.1
Analysis of issues and revision of the MDVP and proposed way forward ......... 23
Current challenges and interim solutions for the MDVP................................ 24
5.2
Review of activities for delivery of reduced doses of vaccines and related
5.3
programmatic issues .............................................................................. 26
5.4 Summary of MDVP session and TLAC discussion and preliminary
recommendations on MDVP pending working group deliberations ................. 27
6. Injection Safety ...................................................................................... 27
6.1
Safe Injection Global Network (SIGN): current and future activities of
importance to TLAC ............................................................................... 27
6.2 Preliminary TLAC recommendations on injection safety ............................... 29
7. General comment .................................................................................... 29
8. Administration and interactions ................................................................. 29
9. Adjournment .......................................................................................... 30
10. Disclaimer .............................................................................................. 30

Technologies & Logistics Advisory Committee - Meeting Report

Geneva, 11-12 March 2009 Page 1 of 30
 1. Opening introduction
The second meeting of the Technologies and Logistics Advisory Committee (TLAC) a was held
at World Health Organization (WHO) Headquarters in Geneva on 11-12 March 2009. Eleven
TLAC members and about 29 observers and presenters from WHO Headquarters and WHO
Regional Offices, UNICEF Headquarters and UNICEF Supply Division, PATH, and industry
(GlaxoSmithKline, Merck, Sanofi Pasteur, Crucell, and Temptime) were present.
Dr Jean-Marie Okwo-Bele, the Director: Immunizations, Vaccines and Biologicals (IVB),
opened the formal meeting by welcoming and thanking the TLAC committee members for their
commitment to serve on this advisory committee.

2. Member introduction and information

A TLAC member who had been unable to attend the inaugural meeting in September 2008, Dr
Oleg Benes of Moldova, was introduced to his fellow Committee members and TLAC observers.
Summaries of other meetings at which the activities of TLAC were presented since TLAC’s last
meeting in September 2008 were presented by the TLAC representatives who attended them.
Dr Bruce Weniger summarized the meeting in Geneva in November 2008 of the Strategic
Advisory Group of Experts (SAGE), for which a published report is available online. b Among
various topics considered by SAGE were two relevant to TLAC’s domain. One was the
affordability of injectable, inactivated polio vaccine (IPV) once live, oral polio vaccine (OPV) is
withdrawn for epidemiologic reasons when polio transmission appears interrupted. As IPV is
about ten times more expensive per dose than OPV, one means to mitigate the additional cost
might be dose-sparing by intradermal vaccination using simpler and safer delivery technology
than the traditional Mantoux method. SAGE also discussed and indicated its strong support for
hepatitis B control programs, which might benefit if hepatitis B virus (HBV) vaccines were
shown safe and effective for use outside the cold chain (“OCC”).
Mr Robert Steinglass represented TLAC at the meeting in Tunis in December 2008 c of
the TechNet 21 (Technical Network for Strengthening Immunization Services) Consultation, at
which a variety of topics relevant to TLAC were covered, including new vaccines and their
introduction, vaccine management, and anticipated policy changes.
Mr Steinglass informed TechNet of the current WHO initiatives regarding OCC and a revision of
the Multi-dose Vial Policy (“MDVP”) which guides health workers in how long opened vaccine
vials may safely be used. TechNet participants commended WHO for creating TLAC, and
welcomed its attention to these two initiatives. They also suggested attention to additional
issues on vaccine management and other policies on which WHO could be advised.
The TechNet group expressed hope that new and easy-to-follow policies on OCC and MDVP
would be available soon. Indeed, there was some impatience as such new policies were
needed “yesterday”, and frustration because staff are currently trained in the current MDVP
and a new one will require retraining. The group emphasized the need in developing new
policies to keep the messages simple, to consult widely with stakeholders, to do field testing,
and to measure impact during and after implementation.
It was pointed out that interim solutions would likely be needed while the evidence for such
initiatives and revisions as OCC and MDVP were researched, developed, and field tested to
ensure necessary comprehension by health workers. A novel term noted by Mr Steinglass at
the meeting was ad-hoc-ism, used disparagingly to describe one-off approaches that
accompanied past initiatives, such as the introduction of new vaccines. The implication was
that more consistent and systematic policies should be applied in the future.
Presentations at the TechNet 21 Consultation relevant to TLAC’s current foci of interest
included a review of Vietnam’s experience with plasma-derived HBV vaccine used OCC from

a. http://www.who.int/immunization_delivery/systems_policy/tlac/en/.
b. WHO. Meeting of the immunization Strategic Advisory Group of Experts, November 2008 – conclusions and
recommendations. Weekly Epidemiol Rec. 9 January 2009;84:1-16
(http://www.who.int/entity/wer/2009/wer8401_02.pdf).
c. TechNet 21. TECHNET 2008 Consultation, 2– 4 December 2008, Tunis, Tunisia
(http://www.who.int/entity/immunization_delivery/systems_policy/Technet_report_2008.pdf and
http://technet21.org/fichiers_Powerpoint/TECHNET08_Report_vf-6.pdf).

Technologies & Logistics Advisory Committee - Meeting Report

Geneva, 11-12 March 2009 Page 2 of 30
 1998 to 2000, and then OCC use of recombinant HBV vaccine antigen from 2004 to 2005. Mr
Steinglass reported that the review reported no increase in adverse events, an increase in
coverage for the birth dose, and satisfactory immune responses similar to vaccine kept in the
cold chain. Another presentation on the adoption of MDVP in Turkey (compared to prior
national discard rules), reported coverage increased while wastage went down (the country
also switched from 20-dose to 10-dose vials).

3. Use of Vaccine Vial Monitors (VVMs) in new policies

Vaccine Vial Monitors (VVMs) play a key role in immunization program management, which
would be even more critical upon an eventual adoption of a policy to use specific vaccines OCC
at elevated temperatures, as well as in implementation of a revised MDVP. Following its
September 2008 meeting, TLAC engaged in three major activities to educate itself on the
development, science, technology, and policy regarding VVMs: 1. a review of the scientific
documentation and consultation with experts in the field, conducted in the interval between
TLAC meetings; 2. a preliminary session providing background and data on VVMs held on 10
March 2009 at WHO HQ with key WHO and outside personnel; and 3. a full-day review on
VVMs and related subjects featuring invited experts on day 1 of the formal TLAC meeting of
11-12 March.

3.1 Reviews of VVM development and information

Following TLAC’s September 2008 meeting, d members Prof Jonathan Colton and Dr Bruce
Weniger undertook thorough reviews of existing scientific data and other information relating
to the efficacy, validity, and regulatory status of VVM technology and its underlying
fundamental research methods and results. Sources included WHO, the VVM manufacturer,
vaccine manufacturers, National Regulatory Authorities (NRAs), other experts in the subject in
industry and academia, and any published or available literature.
Prof Colton’s report, e disseminated to TLAC members and WHO in mid-February 2009,
summarized the documents and presentations by WHO and its consultants at the September
TLAC meeting. It provided an overview of the history of the development of VVMs, citing
publications by PATH (Program for Appropriate Technology in Health), WHO, and experts in
time-temperature integrators and kinetics of biological degradation. It described the WHO
aims for the OCC studies to be performed with hepatitis B vaccines and VVM30 labels at the
National Institute of Biological Standards and Control (NIBSC) in the United Kingdom.
Dr Weniger’s report f was completed later and circulated to TLAC members and WHO on 9
March 2009. This document described the remarkable perseverance in the development of this
technological innovation over the two decades from its envisioning by PATH and partners to its
adoption by WHO and introduction into the UNICEF market. The report, however, posed
unanswered questions and identified gaps in evidence, including (1) the increasing
incomprehensibility of evolving performance specifications for VVMs; (2) the absence of
published or furnished confidential data from pivotal studies to determine the mean time until
VVM conversion (plus measures of error such as standard deviation/confidence limits) at the
standardized specification temperatures (+5°C, +25°C, +37°C); (3) the apparent lack of
rigorous cost-benefit analyses and any “post-marketing surveillance” of actual VVM conversion
frequency in the field since introduction of this innovation costing an estimated US$13 million
to US$26 million annually; and (4) the possibility of mismatch between the kinetics of the VVM
endpoint conversion and the associated vaccine’s stability at different temperatures that could
lead to wastage, particularly for relatively heat-stable vaccines such as hepatitis B fitted with
the VVM30, currently the most-stable version of this time-temperature integrator (TTI).

3.2 Preliminary session on VVM background and data

Eight of the 11 TLAC members were able to participate in the late-afternoon preliminary
session regarding VVMs at WHO Headquarters on 10 March. The main presentations at this

d. http://www.who.int/entity/immunization_delivery/TLAC-report_2008-09.pdf.
e. Vial Vaccine Monitors and Out of the Cold Chain Use - Discussion and Recommendations - Version 7 – February 12,
2009
f. Preliminary Review of Vaccine Vial Monitors (VVMs) – Version 10 – 9 March 2009
(http://siamlotus.com/Pubs/TLAC_VVM_Review_V10-2009Mar9.pdf).

Technologies & Logistics Advisory Committee - Meeting Report

Geneva, 11-12 March 2009 Page 3 of 30
 session were by Dr Umit Kartoğlu of WHO, who is the point person on VVMs at the agency,
and by Dr Thadeus (Ted) Prusik of Temptime Corporation, manufacturer of the HEATmarker®
brand of VVMs used on WHO-prequalified vaccines.
Dr Prusik’s presentation was similar to one he would present the following day. He reviewed
for VVMs the chemistry of their color change (i.e., darkening, or more technically: decreasing
reflectance), the Arrhenius equation that predicts such darkening rate by temperature, the
four types of VVMs, their 2002 and 2006 specifications, some data from the Strassburger and
Siegel study of VVM2s in 1992, and descriptions and summary data of testing from their
versions VVM7, VVM14, and VVM30. He also presented background on quality system
regulation, design controls, verification, and validation, and lot release assays. He referenced
general regulatory clearance by the U.S. Food and Drug Administration (FDA) of the
underlying time-temperature indicator technology [as substantially equivalent to predicate
technology marketed prior to 28 May 1976], g with which Temptime has developed 18 versions
that span a range of 3 months at +5°C to 3 years at +22.5°C (including four which
correspond with the VVM2, VVM7, VVM14, and VVM30 applied to non-FDA-cleared vaccines
sold outside the USA).
For the U.S. military, other HEATmarkers are being used to indicate the cumulative
temperature exposure limit for military food rations (3 years at +27°C) and for another
undisclosed military application (6 years at +27°C). In response to questions about the testing
methods used to develop and assess VVMs, such as sample sizes and observation frequencies,
he pointed out that these parameters were not determined by his company; it simply follows
the specifications and guidelines set by WHO. He presented results from a recent study by a
major pharmaceutical company which he said confirmed HEATmarker performance under
isothermal, cyclic storage conditions and uncontrolled shipping studies, which also found
correlation of the color response to protein degradation.
Dr Kartoğlu’s presentation also covered aspects he would present the following day to the full
TLAC, except it also revealed a few examples of actual shelf-life plots (y axis = log scale of
days to loss of potency or to VVM conversion; x axis = temperature) of unidentified WHO-
prequalified HBV vaccines superimposed upon the plots of their corresponding VVM endpoint
specifications, to illustrate the degree of correspondence between the two lines. The
presentation also addressed other issues raised in Dr Weniger’s review.
Also illustrated was the superimposed shelf-life plots of an unidentified brand of an
unidentified type of vaccine and the most comparable available VVM, suggesting a significant
mismatch between stability of the vaccine and of its VVM at +37°C, but not at the
conventional temperature range (~+5°C) intended for its storage. Prequalification of this
vaccine was on hold at the time because of this mismatch at higher ambient temperature. [It
was later approved without the use of a VVM.]

3.3 Invited presentations on VVMs and related issues

The objectives of the all-day session on 11 March were to inform TLAC on the science,
technology, and development of VVMs and similar technologies for developing-country
immunization programmes, and to educate TLAC on the temperature stability of vaccines and
its measurement by potency assays. Finally, the session was to identify gaps in scientific
evidence in relation to VVMs which require clarification or further research.

3.3.1 Vaccine behaviour at different temperatures and the use, methods,

and limitations of potency assays
Dr Roland Dobbelaer, retired expert from the Scientific Institute of Public Health of Belgium,
was invited to review vaccine thermostability and the tools used for its assessment. He began
by pointing out that vaccines are comprised of a wide range of biological/medicinal products,
which thus result in marked differences in their (1) required conditions for storage and
shipping, (2) behaviour at different temperatures, (3) parameters to indicate stability, and (4)
nature and limitations of their measures of potency. He reviewed aspects of composition,

g. Food and Drug Administration. 510K Summary for Public Disclosure: HEATmarker® Time Temperature Indicator.
No. K063759, 15 August 2007 (http://www.accessdata.fda.gov/cdrh_docs/pdf6/K063759.pdf).

Technologies & Logistics Advisory Committee - Meeting Report

Geneva, 11-12 March 2009 Page 4 of 30
 manufacturing, formulation, and delivery route for the major categories of vaccine construct –
whole bacteria, bacterial (subunit) antigens, whole viruses, and viral (subunit) antigens.
Dr Dobbelaer described various production stages and storage times in manufacturing
vaccines, and for the most common vaccines listed their “stability-indicator” parameters that
apply at various points in the process (e.g., in vivo and in vitro potency, virus titre, specific
toxicity, molecular size distribution, pH, etc.). In the case of hepatitis B vaccine, indicators of
stability include in vivo and in vitro potency and the degree of adsorption to adjuvant.
The presentation illustrated results of acceptable potencies measured after short-time
excursions to room temperatures above the recommended storage temperatures (usually
+2°C to +8°C) for vaccines against diphtheria-tetanus-[whole-cell]pertussis (measured in
IU/s.h.d. h ), hepatitis B (measured in RP i ), measles, mumps, and rubella (measured in
CCID50/s.h.d. j), and Haemophilus influenzae type b (measured in percent free saccharide or
in molecular size distribution). Dr Dobbelaer pointed out that the in vivo mouse
immunogenicity assay which produced the results shown for hepatitis B has now generally
been replaced by an in vitro antigen-content potency assay.
He said the data suggested reasonable stability for several vaccines for short-term usage at
the standard +24°C “room temperature” of the Pharmacopeia, but that this is often greatly
exceeded in the tropics. Another caveat was the wide range of currently-approved potency-
indicating tests, which vary in inherent precision and associated capacity to detect possible
degradation when the vaccine experiences deviations from its usual storage conditions.
Next, different types of potency assays were described, with the point emphasized that these
generally intend to measure a biological activity rather than a simple chemical measurement
of concentration of some ingredient. Depending on the vaccine, such assays can be in vivo or
in vitro, can be quantitative or qualitative, can be immuno-biological, immuno-chemical, or
physico-chemical, and may or may not represent the actual functional mechanism by which
the vaccine induces protection in the intended recipient. The next series of slides illustrated
many examples of these, as well as factors that can influence the results of in vivo assays,
such as species or strain of animal used, housing and feeding conditions, injection technique,
and calculation method. Similarly, for both in vivo and in vitro assays, the results can be
influenced by the specific reagents used, incubation times and temperatures, antibody titration
method, and calculation methods.
Dr Dobbelaer summarized by pointing out (1) the wide range of vaccine constructs, (2) that
their stability must be measured with different techniques specific for them at multiple points
in their production, (3) that several common vaccines (DTPw, HBV, HIB, and MMR) have
been found to retain their potency after moderate deviations from their approved storage
temperatures, and (4) that potency-indicating tests vary in their precision and capacity to
detect possible decreases in potency [after such deviations].
The discussion of Dr Dobbelaer’s presentation touched on regulatory thinking regarding
storage temperatures and label changes; the challenging design, validity, and statistical issues
surrounding potency assays and their interpretation; the effectiveness of preservatives under
non-standard conditions; the effect of alternating temperature swings on the degradation
kinetics of complex proteins; the data required for WHO prequalification; and dual
temperature-range labelling.
Discussion of 3.3.1: Vaccine behavior and potency assays
The first question to Dr Dobbelaer was from Prof Colton on how the Dr Eggers pointed out that some vaccines are to be discarded after
speaker had defined “moderate” in terms of permissible short only 6 hours because of reported instability and asked about the data
deviations above recommended storage temperatures of +2 to +8°C. for that. The answer was that most European regulators are actually
The answer was no more than +37°. However, +45°C might be presented with the stability data at 24 hours after reconstitution, at
considered moderate for such vaccines as DTP or HBV, but definitely which time vaccine stability is still technically maintained. But it is these
not for more heat-labile products such as whole cell pertussis and for same regulators who insist on the 6-hour rule because they are
the classical live antigens in MEA, MUM, and RUB vaccines, and uncomfortable with the [small margin of safety of] results at 24 hours.
definitely not for OPV. The same questioner asked about the quite-wide confidence limits

h. IU/s.h.d. = International Units per standard human dose.

i. Relative Potency (R.P.) = the Effective Dosage 50% (ED50, the dose in µg/mL required to seroconvert 50% of the
mice to seropositive for antibodies to HBsAg) divided by the ED50 of hepatitis B standard vaccine.
j. CCID50/s.h.d. = Cell Culture Infective Dose 50% per standard human dose.

Technologies & Logistics Advisory Committee - Meeting Report

Geneva, 11-12 March 2009 Page 5 of 30
 around many of the potency results illustrated in the presentation, WHO consultant Peter Evans pointed out that lab-based potency
sometimes 40 percent, sometimes twice the reference, and how this assays would involve heat exposures under sterile conditions before
might affect planned WHO studies of certain vaccines and their any vials were opened. But in the field, the elevated-temperature
associated VVMs to be taken out of the cold chain (OCC). Dr exposure would likely occur after multi-dose vials were opened and
Dobbelaer responded that this is a problem not only for the proposed after contaminating organisms thus introduced might be growing in
studies of VVMs and their vaccines to be taken OCC but also for competition with the antimicrobial preservatives and in competition with
classical stability studies. One solution, in theory, is to increase the the protein of the vaccine. He asked if such concerns were relevant in
precision of the results by repetition – greatly increasing replicates of terms of the OCC studies that would need to be done. Dr Dobbelaer
the assays on the vaccine. For in vivo studies, this means doing agreed that this raised two key questions: (1) Would the antimicrobial
multiple assays on an increased number of animals, an extremely agent remain effective [at elevated temperatures] to keep the vaccine
costly strategy. safe? (2) What would be the effect on antigen potency by such
Dr Weniger noted the importance of understanding the concept of microbial contamination, for example from damage due to enzymatic
potency and its assays, because of their relevance to issues in the degradation? Dr Dobbelaer concluded that potency could be affected
TLAC portfolio. He asked for confirmation or correction of his own by such contamination.
definition of “potency”, which is “the ability of a vaccine to achieve its Dr Thomas Cherian offered a caveat to the earlier definition of
intended effect”, that is, “its ability to protect the patient from the potency and its relationship to protection, pointing out that vaccines of
disease”. He explained that the impracticality of frequently repeating similar potency might produce different levels of protection in different
field efficacy trials of a vaccine to demonstrate such protection requires populations, providing the prime example of OPV which is sometimes
“bridging” studies to be done, identifying “surrogate markers” and only 60% efficacious in developing countries. So he suggested to
“correlates of protection”, in order that much more convenient assays clarify the defined relationship between potency and protection, one
can be used to predict protection. But ultimately, when we say a could better say that a vaccine of known potency would induce a
vaccine has not lost its potency, it means we have the chain of “similar level of protection” in the various populations, but not imply that
evidence going back to the original human protection trials, so that we it always will protect.
know the vaccine we have just assayed in the lab will still protect the Mr Serge Ganivet asked about Dr Dobbelaer’s statement that the
patient from the disease. As an example, he cited Haemophilus reconstituted vaccines usually had acceptable stability at 24 hours, and
influenzae type B disease, for which human efficacy trials of its HIB wanted to know whether the diluents for these vaccines should
vaccine established that vaccine-induced antibody above a certain continue to be cooled [before reconstitution, to reduce heat stress to
level correlated with those children who were protected by the vaccine, the vaccine], given the increasing demands on space in the cold chain.
while below that level disease was more likely to occur. So one can Dr Dobbelaer’s response was that if the diluent is typical sterile saline
use such a surrogate indicator to be confident the vaccine will protect. or water for injection, or other inorganic solvent, he sees no reason to
Similarly, there may be physico-chemical, microbiological, or other cool the diluent [before reconstitution].
measurements at the lab bench, or in a lab animal, that have been Dr Weniger returned to the issue of clearly defining potency by
“bridged” to correlate with that human antibody level, and that are thus referring to Dr Cherian’s comment that “when one says a vaccine is
validated indirectly all the way back to the definitive human protection ‘potent’ does not necessarily mean that it induces protection. It may not
trial. These may thus become practical and convenient assays for in some cases.” Dr Dobbelaer suggested that this apparent conceptual
“potency” of vaccine lots. paradox boils down to the quality of the correlate of protection being
Dr. Dobbelaer’s response was that this was a fair understanding of assumed and its applicability to the various less-responsive
the meaning of potency, and that, in fact, the specifications for many of populations to which Dr Cherian referred. He mentioned as examples
these potency assays are derived from such clinical data. For example, the new rotavirus and human papilloma virus (HPV) vaccines, for
measures such as virus titers, molecular weight distributions, physico- which the correlates of protection are not fixed for the moment, and
chemical assays, are indeed linkable to some form of clinical efficacy. may differ among different populations. This is why multiple studies of
But of course, the assays are often of surrogates of surrogates of them are being done.
surrogates, but the link can eventually be made all the way back to the Mr Michel Zaffran commented that Dr Dobbelaer’s presentation
definitive human trial. reaffirmed an observation that his group at WHO has been trying to
Dr Piyanit referred specifically about studying HBV vaccine to ask convey: Vaccines are actually more stable than generally believed, and
how many of the three potency assays listed on Dr Dobbelaer’s slide if taken out of the usual +2°C to +8°C temperatures, they do not
10 (the in vivo, the in vitro, and the degree of adsorption to adjuvant) immediately lose their potency. He asked Dr Dobbelaer to speculate,
would be advisable to demonstrate stability after exposure to elevated as a former regulator himself, if an NRA were presented with potency
temperature. In answering, Dr Dobbelaer cautioned that stability is a data on a vaccine kept at a specific higher temperature for a specific
broader concept than just potency. For some vaccines one must look period time, that an NRA would be willing to permit a label change
at more than just one of the in vitro surrogate assays. There is the permitting such a specific time-temperature exposure. Dr Dobbelaer
stability of the vaccine’s toxicity to consider. Toxicity testing may be responded that in Europe it would be difficult, despite actual prior
required to ensure toxicity has not been exacerbated by higher attempts for something similar, for an NRA to give what might be
temperatures. For example, the aluminium salt adjuvant used in considered a carte blanche for such indication because of the need to
hepatitis B vaccine could dissociate from the antigen, changing its define “room temperature”. Another obstacle, Dr Dobbelaer continued,
reactogenicity [adverse reactions induced]. This may not be detected would be that the label change would involve a major change in
by the in vitro assay, and thus might require an assay for degree of [vaccine handling] procedures and training.
adsorption to adjuvant. Dr Dobbelaer then asked the WHO Prequalification unit staff present
In followup, she asked how many samples at a given temperature- if their reviews required additional stability data for such [extreme
time exposure would be needed to demonstrate stability. Dr temperature] in-use conditions that are not usually required by or for
Dobbelaer’s answer was that it would be difficult to a necessary NRAs in developed countries [with temperate climates]. [The response
number of samples, which would depend on the inherent precision of was inaudible.] Dr Dobbelaer then offered a “nutshell” answer to Mr
the assay to be done. One could assume homogeneity between Zaffran’s question. He said that the regulators would fear – which might
different vials from the same lot, but the imprecision of most assays is not entirely be justified – giving such carte blanche because they would
too high to be able to detect differences between vials. So, in theory, not be sure how these [OCC] rules would be interpreted.
perhaps one vial is enough from any lot. While OCC studies might Dr Hammond referred to slide 8 and its table summarizing stability of
propose assays on 10 samples for each temperature-time exposure, if different vaccines at different temperatures, which for HBV vaccine at
one assumes homogeneity of the lot, then in theory even one vial +37°C is reported “for weeks” and at +45°C “for days”. He asked Dr
would be enough. Dobbelaer if this was correct. Dr Dobbelaer said he had no reason to

Technologies & Logistics Advisory Committee - Meeting Report

Geneva, 11-12 March 2009 Page 6 of 30
 doubt the data behind the table, which came from a WHO publication real-life conditions and [vaccine-handling] rules of developing countries,
and were presumably based on experimental studies. So he assumed such as getting stability data at the recommended 6 hour discard time
they are correct. Dr Hammond then asked about slide 15, showing for reconstituted vaccines? Her answer was that prior to
potency assay results after a room temperature excursion. and asked prequalification they request stability data to meet the specifications of
about the before-and-after data points. Dr Dobbelaer replied that there the target countries, including the 6-hour rule, and including particular
was no statistically-significant difference between the heat-exposed WHO specifications such as for MMR vaccine. She said that everything
and non-exposed vaccine, but that is not the same thing as saying that is specific to developing countries would be part of the [UNICEF]
there was no difference. Another questioner asked about the apparent tender and would be checked.
higher point estimate of relative potency after the heat treatment, for Dr Diana Felnerova of Crucell asked about regulatory trends for
which Dr Dobbelaer speculated that, if true, it was possible that the preservatives in vaccines. Dr Dobbelaer observed that the trend has
heat may somehow have improved the effect of the adjuvant. been oscillating – just as for temperatures discussed earlier. He said
Dr Benes asked about the alternating temperatures – low, high, low, the prior European Pharmacopoeia requirement for preservatives in all
high – often experienced by vaccines in the field, and whether it would multi-dose packagings was softened somewhat when the issue of
affect stability compared to vaccine kept at a consistent temperature. thiomersal [known as thimerosal in some countries] mercury-based
The simple answer was “yes”, that oscillating temperatures are a more preservative arose in the USA, the EU, and at WHO. Temporary
severe test. A followup question asked about the stability of entered recommendations in some quarters to try to avoid thiomersal led to
vials compared to un-entered ones. The answer was – setting aside many discussions. The latest consensus is that there was just no good
reconstituted products without preservative and the microbiological alternative to this preservative if one wanted to meet the [multi-dose]
issues they raise – that their inherent stability would be essentially the specifications of the Pharmacopoeia.
same. Dr Ted Prusik of Temptime Corporation responded to an earlier
Another questioner asked about combination vaccines like comment by Mr Zaffran about possibly labelling for “dual temperatures”.
pentavalent DTPw-HBV-HIB and their effect on stability of the He cited the example of insulin, which is labelled for storage at +2°C to
individual components. Dr Dobbelaer’s answered that such +8°C. But once a patient begins to use the insulin, the instructions
combination would not affect the inherent stability of the different state clearly it can be stored at room temperature for 28-30 days.
antigens. He explained that the expiration date and stability is Another example is the GSK NeisVac-C® [meningococcal C conjugate]
determined by the least-stable component, which in a DTPw vaccine vaccine which is stored at +2°C to +8°C, but whose label indicates it
would be the Pw antigen[s]. However, he continued, it is possible for “may be stored at room temperature (up to +25°C) for a single period
one component of a combination vaccine to influence the potency of not exceeding 9 months.”
other components, both in terms of the potency assay and in the Prof Labuza of the University of Minnesota commented on the
immune response measured in human recipients. But this does not fluctuating temperatures question. Theoretically, there would be no
affect stability issues. effect if one were dealing with a pure substance with but a single mode
Mr Steinglass asked about some newer vaccines indicating use of degradation; it would be the same as if the temperature was held
immediately or within 30 minutes or 3 hours after reconstitution and steadily at the mean kinetic temperature of the fluctuations. But with
whether their manufacturers are being overcautious for the developing complex substances at higher temperatures there can be reactions
world market. Dr Dobbelaer responded that it may not be the between small molecular weight carbohydrates and proteins which
manufacturers who are being cautious, as they often provide stability at denature the latter, as well as the formation of disulfide bonds
even longer periods than on the label -- it may be the regulators who producing insolubility. This was found in bovine somatotropin, for
are shortening the labelled time for use. example. Lipids raised intermittently to higher temperatures could
Mr Steinglass directed a question to Dr Nora Dellepiane of the WHO result in a “history effect” with degradation faster than the mean kinetic
prequalification team about whether they require data relevant to the temperature would predict.

3.3.2 VVM science and technology

Dr Ted Prusik, representing the Temptime Corporation, which manufactures all vaccine labels
with VVMs used on WHO-pre-qualified vaccines sold under UNICEF contract, was welcomed as
an expert on VVMs to review their history, science, and technology.
He began with the history of VVMs, which were first attempted by PATH for use on measles
vaccine with a 1970s technology licensed from Allied Corporation, where he previously worked.
Lifelines Technology, Inc. -- the Temptime predecessor company which was started in 1987 --
began working in 1991 on a VVM for oral polio vaccine (OPV), a very heat-unstable product,
using a small grant from USAID to PATH. By 1996, this VVM (now know as the VVM2 version)
became a requirement for use on OPV vaccines in the Expanded Programme on Immunization
(EPI). Six other companies were also working on VVMs at that time, including 3M and CCL
Label. In 1999, the joint UNICEF-WHO policy statement recommended the VVM requirement
for other vaccines, leading to development of the VVM7, VVM14, and VVM30.
Dr Prusik explained that the colour change of a VVM results from a clear diacetylene monomer
slowly and irreversibly darkening as it polymerizes as a result of exposure to heat and time.
The central indicator of a VVM starts out colourless, and as time and heat are integrated, it
darkens to approach and eventually match the colour (actually reflectance) of the fixed-colour
outer indicator ring, which represents its “end-point”, at which time the vaccine should not be
used. He pointed out that the rate of the chemical reaction of the polymerization follows the
Arrhenius equation that predicts shelf life in units of time (e.g., days) across various
temperatures. The “energy of activation” factor in the equation depends of the specific
chemistry of the VVM, and determines the slope of the standard plot of shelf-life (time on
logarithmic y axis) versus temperature (x axis, often converted to °C from the equation’s

Technologies & Logistics Advisory Committee - Meeting Report

Geneva, 11-12 March 2009 Page 7 of 30
 reciprocal of absolute °Kelvin). He showed the superimposed shelf-life plots of the four VVMs
in current use -- VVM2, VVM7, VVM14, and VVM30 – whose numbers represent the targeted
number of days to reach endpoint when held at +37°C.
He next described and compared the 2002 and 2006 specifications for VVMs promulgated by
WHO, through the previous PIS (Product Information Sheets) and current PQS (Performance,
Quality, Safety) prequalification regime. He explained the tolerances of the 2002
specification, k permitting post-mature conversion of up to 5% of VVMs after the upper-limit
[target] time interval at +37°C (e.g., after 30 days for VVM30) and at +25°C (e.g., after 193
days for VVM30), and -- by extrapolation from these higher temperatures -- at +5°C (e.g., >4
years for VVM30). The 2002 specification also permitted premature conversion of up to 5% of
VVMs before 0.75 of the upper-limit time at +37°C (e.g., before 22.5 days for VVM30) and
before 0.6 of the upper-limit time at +25°C (e.g., before 116 days for VVM30). By
extrapolation from results at these higher temperatures, at +5°C up to 5% of VVMs could
prematurely convert (e.g., before 877 days for VVM30).
Dr Prusik explained that the specifications were revised in 2006 l to provide an equivalent
“temperature zone” instead of a “time zone” to measure conformance, which he indicated was
more consistent with how pharmaceutical companies state shelf life, e.g., “2 years at +2°C to
+8°C”. He admitted that Figure 3 in this specification was complicated and somewhat hard to
understand, and that it took him a half hour to do so when it was initially proposed. Although
the figure is challenging, he considered the associated wording clear.
Dr Prusik reiterated from the prior day’s session that the Temptime HEATmarker® technology
had a 510(k) clearance from the U.S. FDA as a generic, stand-alone time-temperature
integrating device that “… changes color in a predictable manner …” and “… has been shown to
react according to the Arrhenius relationship with regard to time and temperature …”, g but
that it had not been reviewed or approved by FDA in connection with use for any specific
biological product such as vaccines. He clarified that regulatory review and approval for a
HEATmarker® TTI to be used with a specific biological product would be the responsibility of
the manufacturer of the biological to initiate.
He then described and showed graphs from an early PATH-funded 1992 study of the VVM2 by
Strassburger and Siegel, in which they stored 12 VVMs each at +8°C, +20°C, +30°C, and
+37°C and observed them for endpoint conversion at 20 equally-spaced timepoints within and
somewhat beyond the expected conversion time. He showed a graph plotting an unspecified
measure of central tendency of the time until endpoint conversion of the 12 VVMs observed at
each temperature on a plot of shelf life versus temperature, and compared these four points to
a straight line drawn between two points of the VVM2 specification: 2 days@+37°C and 225
days@+5°C.
The next series of graphs shown illustrated the results of testing at Temptime of their other
products, the VVM7, VVM14, and VVM30. He said 20 samples were used for these tests at
each temperature, +8°C and +37°C, because that was the number of samples designated in
the written protocol written [by others] for Temptime to follow. “Could one have done more
[samples]?,” he asked rhetorically, and answered himself, “One could have done more.
Whatever was in the protocol was what we followed.” He pointed out the plots for all
conformed to their specifications [see additional discussion in section 3.3.5 below regarding
determination of sample sizes and numbers of different temperatures to study for VVM
testing].
Dr Prusik anticipated the question of “how high could the VVM go?” in terms of temperatures
above +37°C, and showed a shelf-life plot of VVM30 lot RG169/1 which was also studied at
+80°C, showing good alignment of the measured data points to the straight line of Arrhenius
kinetics. Dr Prusik indicated that during their manufacturing process VVM30s are also tested at
destructive extremes, such as +96°C, and that results at this temperature, as well as at
+45°C, at +37°C, and extrapolated to +5°C, were all close to the line of an Arrhenius curve.
He then explained a retrospective validation from 6 random lots of 20 VVM30 samples done at

k. WHO. Specifications for vaccine vial monitors (VVM), E6/IN5. 25 March 2002. (https://apps.who.int/vaccines-
access/vacman/vvm/05.VVM%20Specs%2025%20March%202002.doc)
l. WHO. PQS performance specification: Vaccine vial monitor (30 November 2006). Document WHO/PQS/E06/IN05.1,
2006 (http://www.who.int/entity/immunization_standards/vaccine_quality/who_pqs_e06_in05_1.pdf).

Technologies & Logistics Advisory Committee - Meeting Report

Geneva, 11-12 March 2009 Page 8 of 30
 the time of the 2002 revision of the specifications that were released by the “one-timepoint
method” m.
His slide illustrated that at 22.5 days none of the 20 VVMs had reached endpoint, and that by
30 days all 20 had converted, as they would remain at 37 days, as well. Similar findings were
shown for samples at +25°C. He then showed graph illustrating the linear change in optical
density (OD) over time for VVM30s at +37°C and +25°C, with the former reaching endpoint at
around 27 days, and the latter at 170 days, respectively. Dr Prusik also showed graphs of 6
random lots of VVMs illustrating endpoint conversions for each lot ranging from a mean of
25.8 days to 27.3 days at +37°C and the six means ranging from 165 days to 175 days at
+25°C.
He said Temptime’s principles for design validation follow standard design controls consistent
with the FDA quality system requirement (21 CFR 820). He explained §820.30(f) on “Design
Verification”, which requires a Design History File to document procedures for verifying that
design output satisfies design input. The design input for its VVM technology was simply the
WHO VVM specifications, which are verified by a corresponding protocol, he said. Another
section, §820.30(g) on “Design Validation”, consists of procedures to ensure that devices
conform to defined user needs and intended uses, including testing of production units under
actual or simulated use conditions. This meant, he said, “doing what the customer wants it to
do”, adding that PATH and WHO have done many studies showing how VVMs work, what it
does in the field, and concluded that “it does what they want it to do.” Dr Prusik then
described Temptime’s various quality control measures, including Failure Modules and Effects
Criticality Analysis (FMECA), manufacture according to current Good Manufacturing Practices
(cGMP) for medical devices, and registration of the company under ISO 9001:2000
[subsequently updated to 2008] and ISO 13485:2003. He reported that Temptime is routinely
audited by the vaccine manufacturers to which it sells VVMs and is subject to audit by FDA.
Dr Prusik then described changes in 2004 in the company’s one-timepoint method for VVM lot
release assays on the basis of consultation by an expert in this field. He said this method was
improved because it includes a secondary set of OD limits that are also applied to reduce the
possibility of accepting VVMs that are well beyond the specification limits. These secondary
limits set the acceptable quality level (AQL) at both ends of the specification to 0.1%. In brief,
the method involves taking 6 VVMs from the beginning of each roll and 6 from its end
[~10,000 VVMs per roll], storing them at the test temperature of +37°C for 26.25 days,
measuring their optical densities, and performing statistical calculations to predict with 95%
confidence that no more than 5% would have converted either prematurely or tardily, and
then applying the secondary limits as well. “Outlier” rolls detected by this method are
infrequent, but occasionally happen and are rejected. He said that such one-timepoint testing
of VVM30s is sometimes also done at +45°C for 8.6 days for comparison with the results at
+37°C.
Dr Prusik clarified that VVMs have four years of shelf life from their day of production [when
kept deep frozen], which is dictated by the adhesiveness of the label and not by the durability
of the special ink. He showed examples of other HEATmarker® products used for food storage,
each with different energies of activation than for VVMs, and thus differing slopes of their
“Arrhenius” curves [plots of endpoint means by logarithmic time versus reciprocal of absolute
temperature].
He also indicated that Temptime could readily produce VVMs with other temperature-indicating
characteristics, such as one of longer duration than the VVM30. But it would of necessity still

m. The “one-timepoint method” is an unpublished, unposted protocol for lot release developed by WHO at the
request of vaccine manufacturers to streamline testing of the VVMs they receive for their vaccine vials (personal
communication, Ted Prusik). The assay relies on optical densities (actually reflectance) measured at but a single point
in time after incubation. Based on assumptions of the Arrhenius equation and the activation energy of the VVM
chemical reaction, the method extrapolates from those OD measurements backwards to estimate the day when
already-converted VVM samples had done so and forwards to predict when unconverted ones would eventually do so.
The single timepoint selected for this assay in the 2002 protocol was the midpoint between the upper limit (e.g., 30
days at +37°C for VVM30) and lower limit (e.g., 22.5 days, which is the specified 75% of the upper limit, at +37°C for
VVM30) of the corresponding 2002 specifications. k Thus, at +37°C for a VVM30 the midpoint would be 26.25 days,
when roughly half of VVMs would be expected to have converted -- in other words, the median day of endpoint
conversion. After VVM specifications were again revised in 2006, l the lot release protocol was changed to assess the
VVMs at their upper limit (e.g., 30 days +37°C for VVM30).

Technologies & Logistics Advisory Committee - Meeting Report

Geneva, 11-12 March 2009 Page 9 of 30
 have the same activation energy of existing VVMs, and would thus follow a slope on a semi-
logarithmic shelf-life plot that would be parallel to the Arrhenius curves of existing VVMs,
although crossing the time axis at a different point. In fact, he said, Temptime makes a
HEATmarker® TTI for military rations with a designed endpoint of 3 years at +27°C, which was
even extended to 6 years for another purpose.
The questions and answers for Dr Prusik’s presentation involved interpretation and
understanding of the WHO specifications for VVM performance; how well plots of vaccine
stability correspond with that of the VVMs applied to them; whether VVMs could detect brief
but damagingly-high temperature spikes; explanation of planned OCC laboratory studies by
WHO; comparative function of different VVMs; the predictive value of VVM conversion; how
VVMs affect vaccine management and wastage; the impact of VVMs on very expensive new
vaccines; and human factors affecting the reading of VVMs.
Discussion of 1.1.1: VVM science and technology
At the beginning of the discussion, Dr Weniger thanked Dr Prusik for Dr Chris Nelson of Merck asked about the focus on VVM30 and lack
his participation in the meeting, and asked a first question requesting of focus on VVM2. Can they both be assumed to function similarly? Dr
clarification of the difference between the one-point-in-time lot release Prusik replied that that both VVMs would function similarly [except for
assay he just described and the incomprehensible 2006 VVM their differing time and temperature integrations] and that the focus on
specifications of WHO reproduced in slide no. 6 at the beginning of his the VVM30 was because of WHO’s request to TLAC regarding
presentation. Dr Prusik answered that 99% of others who look at the hepatitis B vaccine in the OCC initiative. Dr Nelson also asked about
specifications have some difficulty understanding them as well. He Slide 6 [showing the main figures of the WHO 2002 and 2006 VVM
clarified that Temptime’s lot release assay is performed at only one performance specifications], and whether traditional epidemiologic
temperature, +37°C, and at only one timepoint, e.g., 30 days for measures such as sensitivity, specificity, and predictive values could
VVM30s [from 2002 to 2006 it was 26.25 days for VVM30s, under a apply here. Dr Prusik was not sure he could answer the question, but
previous protocol n]. It assumes [from the range of optical densities Dr Kartoğlu interjected that such concepts would be covered in the
among the samples] that the same lot would also pass if tested at a afternoon session on VVMs.
+25°C temperature, based on the activation energy of the product Dr Nelson asked for confirmation of his interpretation [of the 2006
which is [known to be] stable. But they do not test at +25°C, which specifications m] that up to 5% of the time a VVM could indicate to
would require waiting 193 days (about 6 months). He pointed out that discard the vaccine when it was still potent, causing wastage (and up
the specifications were not developed by Temptime, but by WHO. to 5% of the time it could still indicate OK usage after it had lost
Temptime simply performs the assays they are asked to do. potency). Dr Prusik’s response was that the VVM served as a
Dr Deeks asked about slide 40, showing side-by-side Certificates of reassurance for use of the vaccine. The vaccine might very well still be
Analysis of lot assays of a VVM30 at +37°C and 30 days, and at potent after the VVM converts because of the safety margin built into
+45°C at 8.6 days. She asked where the 8.6 days came from. Dr its design. The VVM is a safety tool, he added.
Prusik replied that this was the “target time” [of the 2002 version of the Dr Nelson still indicated continuing inability to understand what the
lot release assay], midway between the lower and upper limits of the specifications meant, using an example of 1000 VVMs, for which up to
specification, indicating when half of the VVM30s at that temperature 50 might convert prematurely and result in wastage of good vaccine
would be expected to have reached endpoint. This is the location of the and up to 50 might convert late and result in impotent vaccine being
“diamond” [in Figure 3, at 12.5% below the upper time limit of the WHO used. Prof Colton answered by pointing out that there is intended to be
2002 specifications l]. The +45°C upper limit is 9.9 days. At that a safety margin, in which theoretically the VVM [endpoint] plotline is
temperature, he said, by 7 days one could begin to see some of the always above the [potency] data plot, so that when the VVM expires
VVM30s expire. one is always going to be throwing out potent vaccine, one would
Mr Evans asked two questions: 1. Having been presented with never thus use impotent vaccine. Colton continued that the “wastage”
vaccine stability data [in the prior session] which was graphed quite issue is a misnomer, one is always going to be throwing out good
differently from how graphs of VVM endpoint conversions by time and vaccine. Dr Nelson responded “Can we afford to do that when we have
temperature were illustrated in the presentation under discussion, if vaccines that are relatively more expensive?”
one plots the vaccine stabilities on the same graph with the endpoints Dr Weniger thanked Dr Nelson as a fellow epidemiologist for raising
of their VVMs, do they follow the same “Arrhenius” lines that you are the subject of false positives and false negatives and predictive values.
presenting? 2. According to your VVM curves, it’s quite clear that there He clarified two distinct kinds of “gold standards” that can be used in
may be a [short] period of time, say an hour at +57°C, for which the such assessments. One is how well a VVM may convert to endpoint at
VVM would still indicate the vaccine was OK [not converted to endpoint] the time indicated by its “blue line” [of Dr Prusik’s plot], given that some
but that the vaccine would not be OK [because of heat damage]. Is that VVMs may convert a little early and some a little late relative to the
correct? theoretical time when it should according to its specifications. He
Dr Prusik responded first to the second question by saying that such explained that the other “gold standard” is the one implied in Dr
issues were posed in 1994, 1995, and 1996, and the answer then was Nelson’s comments, of the actual stability or potency of the vaccine on
that there exist [other] technologies to indicate whether such brief which the VVM is applied. That is, there could be a substantial delay
temperature spikes had occurred, and one could apply them [but they between when a VVM may convert [even if right on time according to
had not been]. In paraphrasing the first question as “whether VVMs its specification] and when the vaccine actually reaches a point of
match all vaccines?”, his answer was “absolutely not”, but that it does unacceptable potency, during which interval a still-good vaccine could
not have to match the vaccine exactly. The goal is to have the VVM have been used, but is discarded because of the appearance of its
expire before the vaccine expires. VVM.
Dr Kartoğlu of WHO then clarified some confusion as to the nature of Dr Weniger added that that is the tradeoff he is struggling with, when
the planned OCC studies of vaccines and VVMs to be performed at a very expensive vaccine may be discarded because its VVM converts
NIBSC. He said it was not “7 days at +45°C” but that it “should go until far too early for the vaccine’s actual stability. He said it was fair to ask
all VVMs reach their endpoint”, which should be around 10 days, so Dr Prusik only about his “blue line” [of VVM conversion], and that it was
“we” should stop using this “7 days at +45°C” phrase. more appropriate to direct questions about premature VVM conversion

Technologies & Logistics Advisory Committee - Meeting Report

Geneva, 11-12 March 2009 Page 10 of 30
 well before a vaccine’s loss of potency to the WHO personnel who people perform. He admitted himself being somewhat “color ignorant”,
prequalify the vaccines and approve which VVM is used. perhaps similar to the surprising numbers of people [in the study of
Mr Zaffran responded to Dr Nelson’s comments about discarding perception of HEATmarker® TTIs applied to the Avocet® product in a
potent vaccine by pointing out it was a theoretical issue, as the goal is format similar to VVMs] cited in Dr Weniger’s VVM review, who were
to avoid VVM color change by better managing the vaccine supply unable to correctly interpret the VVM as it approaches its endpoint. He
chain. The VVM thus serves as a tool to facilitate such better practice asked, rhetorically, how does one teach people to better perceive
to ensure the vaccine is used in its best possible condition, as well as VVMs as they get nearer to their endpoint? Is this not an issue of some
providing a safety margin for any particular vial. When such endpoint concern?, Mr Steinglass added.
conversion and discard does take place, it stimulates a management Dr Prusik said that Temptime obtained ophthalmologic consultation
review to avoid its repetition. on the issue of VVMs and color-challenged people. Using various
Dr Dobbelaer suggested another factor in determining VMM filters that illustrate how color-challenged persons perceive, they found
“wastage”, as mentioned above, which would be to consider the there was absolutely no problem with the VVM shades, which do not
situation if there were no VVMs, and all product which happened to be change in color. He indicated that people seem to have different
found to have gone outside the specified +2°C to +8°C conditions of sensitivity to contrast. He said there is a big pharmaceutical company
storage would absolutely have to be discarded. The VVM thus avoids now doing a study of TTIs, and many such companies have “human
such losses for minor and inconsequential deviations from its factors” groups to address such issues. In one such study, with little
recommended storage conditions. training other than what was written on the package, he reported that
Dr Weniger asked Dr Prusik about the more intuitively- 96.1% of the subjects read the TTI correctly. This study only used
understandable 2002 specifications for VVMs, which permitted up to 5 seven different levels of VVM darkening, but Temptime planned to
percent of VVMs to convert prematurely and 5% tardily, but none of the repeat the study with much finer gradations. He said that the human
limited data he had seen indicated there was anywhere near such eye is a good comparator. He said his eyes were pretty good at
frequencies of early or late conversion. In the data he had been shown, reading the contrasts in VVMs, perhaps as good as a densitometer, but,
he could see no examples of conversions occurring outside the central of course, for those who may have poorer vision, “all bets are off”, as
90 percent band, and asked how often this happened. Dr Prusik replied they would be for other visual tasks.
that he did not have such frequency data with him, but he could find Dr Kartoğlu commented on both the “human factors” issue just
out how often VVMs are found to convert outside of the 5% and 95% discussed and the earlier comments on potential discard of still-potent
cutoffs, but that it did indeed happen. but expensive vaccine. He pointed out that a health worker is not
Prof Labuza, referring to earlier statements about lower rates of expected ever to see a VVM that has reached endpoint, because [the
vaccine protection in some populations, implied that the VVM was a darkening indicator] will pinpoint the problem, warn health workers, and
kind of guarantee of goodness of the product, and suggested the need thus lead to action to fix the problem in time. He implied that VVM-
for a “consequence analysis” that might mitigate the risk of non- based vaccine management will not allow VVMs to reach their discard
protection from the vaccine in certain situations and populations. Drs point in a properly managed cold chain. He acknowledged that
Eggers and Weniger responded that this was not a VVM-related issue, misplaced vaccine, say way at the back of a cold room or hidden
which when applied to a vaccine does not intend to guarantee that the behind other product and forgotten for years, will of course have its
patient will be protected if the VVM has not converted. No vaccines are VVM converted when finally found. He said studies of how well health
100% effective and there are many other factors involved in workers can read VVMs when they are 99.9% towards their endpoints
determining whether actual protection will be conferred. is too theoretical a matter. In his field experience, they throw out the
vial when it is anywhere near that close.
Mr Davis commented on the very-expensive vaccines, such as
rotavirus vaccine already introduced into South Africa, followed soon On wastage, he pointed out this may be a necessary and desirable
by pneumococcal vaccine soon to be introduced under bilateral gifting. outcome in such cases as when a child outside of the targeted age for
He point out the importance of getting better data than currently vaccination appears and a dose is given. This dose is considered
available on the proportion of vaccine vials that get discarded because wasted because of the way numerators and denominators are defined
their VVMs reach [endpoint] phases 3 and 4. He said, many people say in calculating wastage. So such terms are dependent on such
we err on the side of caution” [in disposing of VVM-converted vaccine], definitions and perspectives.
which is of course true, but “are we costing ourselves an arm and a leg Dr Benes wondered whether the testing model used for VVMs was
[in doing so]?” He added, we can only find that out in getting the kind of consistent with the actual practices in taking vaccines out of the cold
data he not yet seen. chain for reaching patients with them. The question remained for
Mr Steinglass changed the subject from how VVMs perform to how consideration after the lunch break.

3.3.3 History of VVM development, quality control, and application; Policy

considerations for VVMs and experience in the EPI
Dr Umit Kartoğlu of WHO reported that the VVM concept was conceived by WHO in 1979
following successes with cold chain monitors at higher levels of the cold chain, prompting the
interest in a vial level indicator to extend monitoring to the periphery. During its 19 years of
development (followed by 11 years of implementation), many donors and countries were
involved in supporting the studies, which took place in many countries around the world. In
1996, VVMs were implemented on OPV vaccine. Following a TechNet Copenhagen consultation
in 1998, based on documentation from field studies in Zimbabwe, Turkey, Nepal and Bhutan,
and in consultation with industry, it was decided by WHO and UNICEF to expand VVMs to all
other vaccines. In 1999, WHO and UNICEF issued a policy statement advocating VVMs for all
vaccines. In 2002, a new set of specifications for them were issued by WHO to cover the
additional 3 VVM types, based on the stability information. These specifications and a
verification protocol was revised in 2006 with the introduction of the Performance, Quality and
Safety project at WHO. In order to mark 10 years of successful implementation of VVMs and
to acknowledge efforts put into this tool by individuals, organizations, institutions and
manufacturers, a celebration event took place in Geneva on 3 May 2007.

Technologies & Logistics Advisory Committee - Meeting Report

Geneva, 11-12 March 2009 Page 11 of 30
 Dr Kartoğlu said he would avoid details about VVM quality control and lot assays that the prior
and subsequent speakers would be covering. But in brief, VVMs are manufactured under high-
quality control systems, starting with the ink base preparations, printing, storage, release,
shipment, acceptance and application by their manufacturer and vaccine manufacturers. VVMs
are shipped on dry ice to vaccine manufacturers with an accompanying lot release certificate.
Vaccine manufacturers initially validate different VVM categories against the verification
protocol with end-point optical density (OD) conversion measurements at 30 days at +37°C,
for example for the VVM30. Once validated, some vaccine manufacturers continue using the
same end-point OD conversion measurements as acceptance tests on all lots, while some only
do such readings of ODs on some lots randomly during the year. But in all lots, they measure
the initial reflectance OD of the central endpoint indicator as well as of the peripheral indicator
ring, and compare them to color calibration cards.
Dr Kartoğlu then showed filling line equipment that applies VVMs to either to the vial label or
to the cap of the vial (or neck of the ampoule) depending on the presentation.
He then discussed the two WHO publications on temperature stability of vaccine from 1998
and 2005. He pointed out that WHO never published any document indicating which vaccines
are to be associated with which VVM types. But it is known that all OPVs get VVM2, and HBV
vaccines are expected to get VVM30. However, for other vaccines the appropriate VVM
depends on the vaccine. For example the Japan BCG has longer stability than the Statens
Serum Institute BCG, and thus have different VVM versions VVM7 and VVM14 (which are
indicated around the periphery of each VVM). So WHO looks through the vaccine stability data
to find the best match to the vaccine.
Dr Kartoğlu explained that the original 1997 specification came out only for VVM2 for OPV, and
the later 2002 specifications were developed when the three additional types -- VVM7, VVM14,
and VVM30 -- were implemented. He pointed out that in 2006 the specifications were changed
again mainly because of the shift from WHO’s PIS system that simply listed available items to
the new PQS regime in which WHO prequalifies them. He said the 2002 specifications dropped
a light sensitivity test on VVMs that were in the original specifications because the vaccines
themselves are supposed to be kept in the dark anyway. Between 1998 and 2002, several
meetings took place and studies were done, and TechNet recommended expanding VVMs to all
[EPI] vaccines, and proposals were sent to vaccine manufacturers for feedback, and the final
updated specifications were issued on 11 March 2002. At that time, numbers were added to
the VVM names to avoid confusion of the different versions. Another revision was to move to
+5°C as one of the standard specifications temperatures -- as the midpoint between +2°C and
+8°C -- rather than use the fixed +8°C temperature of the original specifications.
He reiterated Dr Prusik’s earlier mention of the HEATmarker’s 510(k) clearance from the FDA
as a generic TTI which was shown to react according to the Arrhenius relationship with regard
to time and temperature [but not for use with any specific vaccine]. He quoted the FDA as
stating that the TTI “changes color in a predictable manner when exposed to a given
temperature for a specified period of time”. [For the preparation of this report, after the
meeting Dr Prusik added for clarification: Studies of the polymerization of the substituted
diacetylene monomers support this relationship, as do more recent studies involving inks
made from these monomers. He continued: “As with the predicate device, it allows the TTI to
demonstrate that the medical device (vaccines in our case) to which it is affixed has been
exposed to a quantity of heat characterized by a specified time-temperature profile.”]
Dr Kartoğlu discussed at some length the concepts introduced [in Dr Weniger’s review of VVMs]
about "wastage error” and “impotency error" based on the specificity, sensitivity and positive
predictive value of VVMs and the appropriate “gold standards” for their determination. He said
such concepts were not valid simply because VVMs are intended to convert prior to the loss of
potency of the vaccine in order to have a safety margin. The actual stability of the vaccine
should always be better than suggested by the degree of darkening of the VVM indicator. Dr
Kartoğlu believed that the concept of “wastage error” was too theoretical and not pragmatic,
and did not apply to VVMs because they were not invented to be an indicator of whether the
vaccine is potent or not. He pointed out that vaccine potency does not drop precipitously the
day after its expiration date, but that expiry dates on vaccines are based on a decision of how
much risk to accept in setting a cutoff for what is actually a gradual waning of potency. Usually,

Technologies & Logistics Advisory Committee - Meeting Report

Geneva, 11-12 March 2009 Page 12 of 30
 he pointed out, potency would remain acceptable well beyond the expiration date, which is set
where it is to build in a margin of safety.
He then discussed the limitations in temperature monitoring of vaccine refrigerators in the
field, which is mainly done with a thermometer being read and noted twice a day during
working days only. But this does not constitute sufficient monitoring, since there is no record
for the period of time between such spot checks when the temperature may have changed to
quite unacceptable levels. He stressed that VVMs remain the only tool among all time-
temperature indicators that are available at all times in the distribution process through to its
administration to the patient, indicating whether the vaccine has been exposed to a
combination of excessive temperature over time and thus likely or not to have been damaged
by heat. It clearly indicates to health workers whether a vaccine can be used.
Dr Kartoğlu provided the example of the Jogjakarta earthquake in Indonesia in 2006, during
which more than 50 health centres were demolished, and their vaccine stores were exposed to
more than a week of high temperatures. But with VVMs, it was possible for health staff to
rescue from the rubble and return to refrigeration vaccines whose VVMs still indicated "OK"
and discard only those with either converted VVMs or no VVMs at all. In another incident in
India in 2008, one state received VVMs with darkening (but still usable) indicators, prompting
authorities to investigate the cold chain to discover problems with one particular store,
resulting in corrective measures.
Dr Kartoğlu explained how the VVM types are assigned to particular vaccines through the WHO
prequalification process. WHO analyzes and plots the stability data provided for the vaccine
alongside the VVM stability curves to find the best fit. In cases with insufficient data, WHO
requests additional accelerated studies to help decide the best VVM version to use. For
example, submitted data for one vaccine indicated stability through 4 years at +5°C and until
180 days at +25°C (both deserving a VVM30). But at +37°C, the stability was studied only
through 1 week, which was insufficient to determine if a VVM30 was appropriate. So the
manufacturer was asked to extend testing at that temperature for a longer period.
To date, there has been only one case in which the stability profile of a new vaccine did not fit
any of the existing VVM categories. Its data indicated pretty good stability at +5°C, but only
64 days at +15°C, only 2 days at +25°C, and just hours at +37°C. In this case, WHO is
working closely with its manufacturer to identify another solution at the vial level that might
be able to warn health workers of this instability at higher temperatures.
Dr Kartoğlu addressed the paucity since VVM implementation of hard data on the actual
frequencies or proportions of vaccines being discarded due to their VVMs having reached
endpoint. He said it was very difficult to get such data through routine reporting. Routine
monitoring does not usually provide the reasons when vaccine is reported wasted. To get such
data, one must commission a study or create sentinel sites as described in the WHO document
entitled Monitoring Vaccine Wastage at Country Level. n In practice, he said such conversion
should not happen since VVMs warn health workers that there is a problem. It is expected that
responsible staff fix the problem to prevent further degradation.
If implemented successfully, VVM-based vaccine management will not allow VVMs reaching
their discard point in a properly managed cold chain, he said. However, there are cases where
vaccines with VVMs beyond their discard point were discovered in storage facilities during
inventories -- mainly forgotten in a corner behind other things.
Regarding freezing, he indicated that a VVM to indicate such damage was feasible; however,
for now such indicators would have to be applied on the outer cardboard box as opposed to
the vial itself. Vaccines freeze at different sub-zero temperatures, and the resulting damage to
the vaccines is variable. He pointed out that the shake test would remain a useful tool for
alum-containing vaccines, as it visually reveals the actual damage to the vaccine, not some
proxy of the damage.
Dr Kartoğlu concluded that VVMs have become an invaluable tool to increase coverage
through increased access in hard-to-reach communities and in areas with very weak and no
cold chain infrastructure. He ended his presentation by showing a movie about vaccine vial

n. WHO. Monitoring vaccine wastage at country level: Guidelines for programme managers. 2003. Document
WHO/V&B/03.18 (http://whqlibdoc.who.int/hq/2003/WHO_V&B_03.18.pdf).

Technologies & Logistics Advisory Committee - Meeting Report

Geneva, 11-12 March 2009 Page 13 of 30
 monitors, “Five Senses: I trust VVM”, produced for the 10-year anniversary in 2007 of the
implementation of the VVM requirement for EPI vaccines.
The discussion of Dr Kartoğlu’s presentation covered “wastage” and “impotency” concepts;
margins of safety built into VVM conversion time; manufacturer liability concerns; the effect of
OCC on low-potency batches; specifying clear limits for OCC; the WHO prequalification process
and monitoring; the possibility of new VVMs more stable than the VVM30; and vaccine potency
variation thoughout its shelf life.
Discussion of 3.3.5: History of VVM development
Dr Kartoğlu’s question and answer period first covered the concepts that temperature, he worried there might be “issues” for some vaccines.
of “wastage” and “impotency”, when VVM endpoint may not match The vaccine potency might fall below its clinically-validated level
exactly when its vaccine potency drops below an acceptable level. In because it no longer has the normal decrease curve or equivalence
response to an [inaudible] question by Dr Dobbelaer, Dr Kartoğlu curve that it does at +2°C to +8°C. In such cases, one has not taken
pointed out that the levels of acceptable potency for a vaccine are into account the lower release potency of that particular batch because
established by the manufacturers themselves. He said WHO uses that the vaccine validation was generic. He was concerned that low potency
point when assessing the temperature-stability data of vaccines batches may not offer the same guarantee even when subject to VVM-
submitted for licensure and the suitability of its proposed VVM. compliant conditions.
Dr Weniger described the necessary and somewhat arbitrary Dr Cherian commented on the liability issue by asking how one could
dichotomous choice when vaccine developers must set the minimal fix liability when a vaccine did not protect any one individual because
acceptable level of potency for which there is usually a steady no vaccine does so for all its recipients. It is only on a large-scale
degradation according to its time-temperature exposure. In the population basis that liability might ensue, where one gets a lower-
hypothetical case of a live virus vaccine that may start out with one than-expected overall immunity rate, say 70% instead of the usual 90%.
million viable units per given volume, that cutoff might be a decrease of Dr Benes wondered about the need for establishing clear limits of
one log to 100,000 units. He said there would likely be patients who temperature, humidity, sun exposure, etc., when taking VVMs outside
receive a bit more than 100,000 units and fail to respond, while there the normal storage temperature. Mr Zaffran commented that the
will be others who get less than 100,000 units -- a technically impotent discussion was more appropriate for the next day’s focus on the OCC,
dose -- and do become immunized. Regardless of this imprecision, he which is only a “future, possible exploration”. Tomorrow’s topic should
indicated that to assess the theoretical concepts of VVM wastage and not “pollute” today’s discussion of how VVMs work. Another point was
impotency, one would need to use this minimal level of potency as the made regarding low-potency vaccines still within specifications, in that
gold standard of what the VVM conversion ideally should indicate. the prequalification process looks at that range of potencies, and that
Dr Kartoğlu agreed with the need for such a gold standard in that prequalification is not based solely on high-potency batches.
respect. But he thinks that it is not appropriate to apply such a wastage Prof Colton suggested that perhaps the “wastage” term or concept
determination to VVMs because they are intended to convert prior to might be replaced by “efficiency”, determined, for example, by
the loss of potency. They are not an [automotive-style oil ] dipstick, he comparing the number of vials delivered to a country and the number
added. Using the example of the proposed OCC studies at +45°C, he of their doses given.
said the VVM is intended to convert before the vaccine’s potency Dr Dellepiane said that the prequalification process includes looking
reaches unacceptable levels. When it does convert, the vaccine is still at potency data submitted for several lots, not only the best lots
OK and thus not wasted. submitted. WHO also does continual monitoring of the product
Mr Evans responded to a rhetorical question from Dr Kartoğlu about because each manufacturer must conduct stability testing on one batch
the potency or efficacy of a vaccine one day after its formal expiration per year. She said the manufacturers are asked to look at several
date. Mr Evans described the hypothetical case of a vaccine reaching temperatures to ensure the assigned VVM category relies on such data.
its expiration on 31 October, but being administered on 1 November. Dr Weniger recalled Dr Prusik’s landscape-changing revelation
Although the potency might change very little in one day’s time, what earlier in the day that Temptime could produce a VVM45 or a VVM60.
does change significantly is that the manufacturer no longer has This opens up new possibilities [to minimize the mismatch and
liability for that vaccine. Anyone suggesting to use it after its expiration wastage issues], he said. He asked Dr Prusik if Temptime could also
date would assume liability for same. A similar principle would apply, change the activation energy of their VVMs, so as to better match the
he continued, for the storage and use of a vaccine outside of its activation energy of some vaccines [which may follow a different slope
manufacturer’s suggested storage range of +2°C to +8°C on the basis than the VVMs]. The answer was that they have been unable to
of what the VVM was indicating. In such case, the liability “would be change the slope of the VVMs, although they keep trying to do so, but
ours, and we better make sure that we have the science to back up he does not have much hope for success. However, he said, they are
what we are doing”. “I think we have, but that’s what it takes to change”, able to provide a means to indicate an upper-temperature threshold at
he added. a point where a vaccine might experience catastrophic denaturation
Prof Labuza commented on the liability issue, which is why chemical from even a short-duration heat exposure.
tags are not being used much on foods as safety tags. People might Ms Phillips asked about the potencies of vaccine at the beginnings
see the tag had not converted, even if its expiration date was passed, and ends of their dated shelf lives. The answer was that potencies are
and consume it anyway. The manufacturer could not prove when the probably much higher on day one after release than on the last day
food was eaten and thus still face liability. before they expire months or years later. But that the date assigned to
Dr Dobbelaer asked about the differences between batches of its expiration is based an a cautious calculation and estimation that it
vaccine, some of which may pass but be on the low end of potency, will still have more than the cutoff of minimal potency considered
while other batches would be at the high end. Regardless, normally necessary by its manufacturer and NRA to immunize patients. Dr
stability data would guarantee that the vaccine would still be good at Milstien clarified that the vaccine must still meet its release
the end of its expiry date if stored at the recommended cold specifications throughout its shelf life, but it will still be losing potency
temperature. But if one were to put a VVM on a vaccine of low but during that period.
passable release potency, and used it as the criteria of usability above

Technologies & Logistics Advisory Committee - Meeting Report

Geneva, 11-12 March 2009 Page 14 of 30
 3.3.4 VVM validation and lot acceptance
Dr Diana Felnerova of Crucell presented an overview of how a vaccine manufacturer validates
the VVMs that arrive on their labels, which are preprinted by and received from Temptime.
She indicated that the testing is based on validation and lot acceptance tests described in the
WHO document PQS/E06/IN05 containing the 2006 VVM specifications. The main required
equipment includes incubating water baths, alum pouch bags, an X-Rite 404 colour reflectance
densitometer, a calibration card, a heat sealer, and freezer.
The work involves a shipping condition test, a start point determination, determination of the
VVM reaction rate at +37°C, and then a reversion test. Many of these tests involve calculating
differences in the reflectance OD of the central indicator field of the VVM and of the outer
reference ring.
The shipping test verifies if changes in the VVMs occurred after the labels left the VVM
manufacturer’s plant by comparing their reflectance upon receipt with that reported by the
manufacturer on its accompanying Certificate of Analysis. The start point determination
[before the labels are exposed to temperatures above freezing upon the day of filling and
labelling the vaccines] involves measuring 50 samples from each of 3 label lots by measuring
the reflectance of the central indicator and of the outer reference ring. One can also check
VVM quality by analyzing for homogeneity of the indicator reflectance across multiple labels.
The test of VVM reaction rate at 37°C involves testing 100 samples representing 3 label lots
for an interval corresponding to 75% of the nominal time period for the VVM in question. For
example, for a VVM7, the 75% interval would be after incubation for 5.25 days, the “target”
interval would be 6.12 days, and the maximum 7 days. No more than 5% of samples should
convert before the earlier 5.25-day timepoint, and more than 95% should have converted by
the final 7-day timepoint.
The reversion test measures whether the VVM central indicator lightens after being returned
for 20 days to +2°C-to-+8°C storage after a +37°C incubation period until its target time
(e.g., 7 days for a VVM7).
Dr Felnerova then showed examples of actual reports of such validation testing for VVM7 and
VVM14 labels. One graph showed the distribution of O.D. differences (indicator minus
reference) at 75% of the time, for which none had reached endpoint, and at the full time, by
which all had converted.
She explained that lot acceptance involves review by the vaccine manufacturer of the
Certificate of Analysis from the VVM manufacturer that arrives with each lot of labels. She
then showed an Inspection Card that records the reflectance observations made soon after
receipt of the labels. Photographs were shown from a Crucell manufacturing plant in South
Korea its (minus) -25°C storage and temperature monitoring and recording facilities for VVMs.
The discussion of Dr Felnerova’s presentation involved questions on how VVMs are stored
while incubated for validation; the voluntary or mandatory nature of such testing; the shelf life
of the VVMs upon arrival at the manufacturer; the effect of humidity on VVM performance;
health worker reading of VVMs; and the need for a wider variety of VVMs.
Discussion of 3.3.4: VVM validation and lot acceptance
During the question period, Prof Colton asked Dr Felnerova about Dr Dellepiane of WHO responded that such testing was required as
the storage bags used to hold the VVMs when subjected to water-bath part of each company’s validation process for VVMs, and it was
incubation. She clarified that these are aluminium pouches which are formerly required for routine acceptance of all label batches (lots). But
heat-sealed to keep out the water. Dr Weniger complimented the clear that requirement was subsequently relaxed by WHO, so that now they
and precise presentation, and then asked specifically about the only need to do the initial measurements of OD when the batches
“voluntary” nature of the water-bath testing over time that Dr Felnerova arrive. Measuring endpoints of incubated labels at the three time points
described, as distinct from taking the beginning O.D. readings on all is no longer required for lot acceptance, although some companies do
label lots as they are received, which he understood all vaccine it voluntarily, perhaps based on their NRA’s advice.
manufacturers do. Dr Kartoğlu amplified her answer that VVMs are treated like any
He directed the specific question to WHO about whether the [water other raw material entering a manufacturing unit, which requires initial
bath incubation] reaction rate assay just described is voluntary or validation. When it comes to such [real-time water-bath] testing for lot
mandatory on the part of every vaccine manufacturer on all lots of acceptance, however, whether it continues to be performed would
labels received from Temptime. If it is required for all, he also asked depend on the company’s individual standard operating procedures
whether that mandate comes from each country’s national regulatory (SOPs), quality controls (QC), and quality assurance (QA) policies.
authority, or from WHO’s Quality, Safety and Standards (QSS) team. Some companies do and some don’t. Some may audit the supplier

Technologies & Logistics Advisory Committee - Meeting Report

Geneva, 11-12 March 2009 Page 15 of 30
 instead, Dr Dellepiane added. darker?”. Dr Prusik replied that this could be problematic because of
Dr Weniger lamented that it was unfortunate the heparin inevitable variation around whatever final and fixed darkness was
manufacturers in the U.S. and Europe had not similarly audited their engineered into the indicator’s chemistry. If some VVMs they
suppliers and the original sources in China for quality controls over happened to “stop” just short of the reference ring’s darkness, the VVM
their raw materials. It might have avoided the recent terrible scandal of would never convert.
illness and death from contaminated heparin. The lesson is that Prof Colton asked whether there is some kind of “eye chart” that they
sometimes the system fails to work in terms of being able to audit back can use to compare what they are seeing on the active VVM with a
to the original source. reference standard, which might address the human factors around
Dr Benes asked about shelf-life of the VVMs themselves, while reading the VVMs. Dr Prusik replied that this could be done, but there
sitting at the vaccine manufacturer’s plant, waiting to be applied to the is some variability between the colours on different VVM types, which
vaccine containers. Dr Prusik replied that the VVMs have a 4-year would make it a “moving target” [to come up with such a single
shelf life [at the recommended below-freezing <-24°C storage] that is reference card]. Regarding human factors, Dr Prusik said studies of
printed on the label rolls. This term results from the adhesive properties colour reference cards caused confusion among subjects when the
of the label sticker as the limiting factor, and not from the VVMs, which hue of the VVM was slightly different from that of the reference card.
themselves could remain unaffected for many years longer while in This was avoided by using a colourless gray-scale card, so that the
cold storage. Temptime guarantees a minimum of 2-years shelf life in subjects focused solely on lighter vs. darker, rather than exact colour
cold storage for arriving labels, but often it is 3 years. matching. As a result, VVM examples in the new instruction materials
A vaccine industry representative commented that humidity also Temptime is producing will be in gray-scale only.
needs to be monitored and can affect VVM performance. He went on Prof Colton then directed a question at Dr Felnerova: whether her
to say that the fourth stage of VVM conversion [when the inner square company felt there was a need for more VVM types [beyond VVMs “2”,
is darker than the outer indicator ring] is problematic for some health “7”, “14” and “30”] for some of its vaccines, as well as how her
workers in his country, where they confuse it with the second stage company picked the VVM to use on a particular product. She said, “of
[when the inner square is lighter than the outer ring]. He wondered if course” additional VVM types would be desirable, as among many
one could make the final appearance be stage three [when both products it’s not easy to fit the stability of some to one of these four
square and ring are the same and the square cannot be discerned]. types. In regard to picking the VVM to use, she said they present their
That way, one could instruct to use the vaccine only when you can see stability data to WHO and ask for advice.
the inner square. Mr Steinglass commented that in the field, training materials were
Dr Prusik explained that VVM validation testing occurs at specified often poor-quality copies of a copy of a copy of the original document,
relative humidities (RH) of 30% (pretty dry) and 73% (pretty moist), and and then asked Dr Prusik whether the new gray-scale instructions
there is no measurable effect of RH. With regard to how the indicator would solve that problem. He answered that such poor quality copies
changes, he said it was preferable to have a gradually changing would not be solved.
indicator, rather than a “gotcha” one that jumps quickly from still Mr Davis asked Dr Prusik how hard would it be to produce a fifth
looking pretty fresh to one that is converted. Otherwise, health workers type of VVM if “WHO in its infinite wisdom decided that it needed five
would have no warning at the beginning of a long trek that VVM versions of the VVM”. The Temptime representative said that as long
expiration was near. as the desired parameters of the new VVM fit within the lines of their
Dr Kartoğlu pointed out that there is no official recognition of the chemistry’s activation energy, they could do it. In fact, they have
“stages” of VVM darkening, as was mentioned. It is a gradual process already done so and could provide samples of prototype indicators
for which the only criterion is whether the indicator is lighter or is the other than the existing VVMs 2, 7, 14, and 30.
same or darker as the outer ring. Dr Eggers pointed out that the Ms Phillips recalled that when the first VVM2 for OPV came out, the
question was somewhat different than what was answered; the South Africans requested posters to go on the wall to guide health
question was: “can a VVM be produced in which the indicator square workers in reading the VVMs, and there was not problem using them.
can become identical in darkness to the outer ring, but never gets
3.3.5 Alternative Time-Temperature Indicator (TTI) technologies:
mechanisms, advantages, and disadvantages
Prof Ted Labuza, Professor at the University of Minnesota and an expert on degradation of
foods and biologicals and time-temperature integrators (TTIs), has been involved with,
consulted for, or followed closely TTI research involving Temptime, the 3M Corporation, the
U.S. Department of Defense, and other organizations in the development and testing of many
products in the decades-long history of developing TTIs for foods, drugs, and vaccines. He
provided an encyclopaedic historical and technical review of shelf live dating, showed many
examples of different types of TTIs, and illustrated their various applications for food and
pharmaceutical products like vaccines.
Prof Labuza explained the kinetics of the rate of degradation at different temperatures and the
key factor that determines it – the “activation energy” – which is a variable in the Arrhenius
equation for those reactions in which it applies. A simpler way of looking at this variable,
which he called the “Q 10 ”, is that it represents how much faster a reaction goes for each 10°C
increase in temperature. He said this factor “varies all over the map” for enzymatic reactions,
despite what some biochemistry books claim is a fixed value of 2, and is one of the reasons
why it is so hard to design [one] TTI to be used for differing products.
He pointed out that it is probably unknown what are the exact mechanisms involved in the
deterioration of vaccine potency over time. The Arrhenius relationship is not a plot of the rate
of degradation, but rather of the rate constant. However, he said that if one uses some
mathematics to cancel out some terms in the equation, and assuming the rate is constant

Technologies & Logistics Advisory Committee - Meeting Report

Geneva, 11-12 March 2009 Page 16 of 30
 along different sections of the temperature range, and degradation is a zero-order reaction,
one can plot it simply as a straight line of units against time. In the case of foods, when the
“units” of quality reach 50% compared to when fresh, that is usually considered its end point.
The line can then be extrapolated to predict shelf life at future times when the quality units
will reach zero.
First-order reactions, he explained, can be plotted on semi-log paper to produce a straight line.
But he warned that many forget – including authors of published mistakes in prestigious
journals – that the actual equation is to the power of e, so that the slope from the log 10 graph
must be multiplied by 2.3.
One major principle he pointed out was that the decline in the quality of a food as it degrades
[or the potency loss of a vaccine] should correspond with the changes in its TTI. Thus, on a
graph of time to degradation versus storage temperature [a “shelf-life plot”], the line of an
ideal indicator would be parallel to [with same slope] and occur just slightly “earlier” than the
line of the product it was labelling.
Prof Labuza illustrated different types of such mismatches, such as when the two slopes differ
and thus their lines diverge or converge, or the TTI reaches its endpoint after the product has
degraded. Prof Labuza indicated that for complex biological products it was a major challenge
to get these two degradation lines to match closely.
As an example, he circulated samples of the 3M diffusion tag developed for measles-mumps-
rubella (MMR) vaccine as part of a WHO effort. Its mechanism is wax that melts and diffuses
from its original pool down the tag. Its activation energy was from 5-to-9 kCal per mole, which
is below what is typically found for food degradation reactions, and he presumed it would be
the same for vaccine degradation. When working at 3M, he superimposed the plot of the 3M
MMR marker against potency data to find, for example, that at +5°C the tag would convert at
25 days, while the vaccine potency would not be lost until 1,500 days, illustrating its potential
for wastage, and the reason why it was not used.
He showed examples of newer TTIs applied now on food products. One uses a chemistry
capable of activation energies ranging from 5 to 50 kCal per mole, and is currently used for
meat and fish; some versions he said might be suitable for vaccines. Another, the Vitsab, uses
pouches with separate compartments. Another, Fresh-Check from Temptime (when it was
known as Lifelines), uses a similar chemistry as for its VVMs. He pointed out that this company
responded to WHO’s call that led to the VVMs in use today on vaccines. Prof Labuza is aware
of several other companies with chemical TTIs, but had not seen their performance in any
publications.
Prof Labuza discussed several products and their different mechanisms of actions for
temperature limit indicators, another technology relevant to concerns about freeze damage.
These detect either freezing or any excursion above a certain temperature. For use with
vaccines, he said the key need for applying a freeze indicator would be to know the true
freezing point of the vaccine, which will vary among different products.
He illustrated many electronic temperature-logging systems, which are currently used in
distribution systems, but not yet individual packaging. Such temperature loggers had the
advantage - if used correctly and linked up with a reporting system – to pinpoint when, and
thus where, undesirable temperature excursions had occurred, and thus lead to corrective
action. But they do not integrate the time-temperature exposure of the product in order to
predict its remaining shelf life.
Another type of electronic device, he said, could be programmed to follow a complex
degradation pattern of a product, but these existing “tags” have the disadvantage of higher
cost, which makes it impractical to directly attach one to every vial. Other TTI products are on
the horizon, including both electronic (such as radio-frequency identification [RFID]) and
chemical ones, that would further enhance the ability to monitor the exposure to temperature
over time. He showed prototype examples of a small, film-mounted freeze indicator his group
has been developing that could be applied to small packaging like vials.
The final section of Prof Labuza’s presentation was devoted to his belief that “what is really
needed is a good marker for end of shelf life” for both foods, and certainly vaccines. The first
step is to collect good data on how a tag functions, as is now available for the VVMs. Then the

Technologies & Logistics Advisory Committee - Meeting Report

Geneva, 11-12 March 2009 Page 17 of 30
 kinetics must be matched at some point with the vaccine data. Neither the tag nor the food or
vaccine product should have a “history” effect, which is when a product does not follow the
Arrhenius prediction when held at different temperatures for consecutive periods of time. This
effect is most likely the result of multiple, simultaneous [degradation] reactions, and not just a
single one modelled by the Arrhenius equation. Finally, the ideal tag should have an easy-to-
read and abrupt endpoint.
The discussion of Prof Labuza’s presentation covered the use of RFID; design of trials to
assess vaccines OCC; the effect of light exposure on VVMs; use of TTIs for cold-chain
monitoring; other inventions relevant to vaccines; temperature spike indicators; high
temperatures in the tropics; interpretation of policies in the field; a “fast chain” for vaccine
handling; and flexibility in the cold chain.
Discussion of 3.3.5: Alternative TTI technologies
The first question posed to Prof Labuza involved whether RFID don’t tell us very much about problems. Prof Labuza mentioned for 5 to
scanning might affect biologicals as a result of the radiofrequency 6 years there has been a cold-chain management group at a university
energy to identify them. Prof Labuza doubted that the short 2- in Germany, with an annual symposium attended by various
microsecond interval for an RFID read would harm the vaccine, and companies, and that someone from this group might attend these.
had not heard of anyone presenting this subject to the FDA for review. Mr Zaffran recalled efforts a few years ago to find competitors to
Mr Zaffran said WHO has no formal position on the question, although Temptime for alternative VVMs, and that 3M was one of these. He
some discussion had occurred regarding their use with drugs. He asked if there were any other products not yet brought to market that
understood that liquids interfere with the reading, so there is a might work. Prof Labuza indicated he held part of the patent on the 3M
theoretical problem with vaccines. Also, batch scanning sometimes device, and that the company had a figurative warehouse of TTI
misses some items, so although RFID is a promising technology much technologies sitting on the shelf unpursued. He described a simple
remains to be looked into about them. Prof Labuza added that DHL wax-impregnated filter paper heat indicator that the 3M Corporation
puts their liquid products into aluminium coolers that prevent the waves invented to ensure that frozen dinners were sufficiently cooked by
from entering, and the RFID tags are put on the outside. consumers to avoid salmonella. The problem is that food
Given Prof Labuza’s departure before the next day’s discussion manufacturers want something that costs less than US$0.01 each,
about the OCC, Dr Weniger asked him for his thoughts on what would while 3M was hoping to charge $0.07 for it. He said one has to sell
be the pivotal trial design to effect a regulatory change for taking billions to make a profit at a selling price of just $0.01 each. He said the
vaccines out of the cold chain at possibly high ambient temperatures. value of the product needs to be high enough to justify spending
Prof Labuza responded that he would first choose the temperatures, at proportionately more on a TTI to indicate its quality for consumption.
least 5 of them, commenting that in some PhD studies of bacterial Dr Weniger commented about a point made in an earlier discussion
growth up to 10 or even 15 might be looked at. He also tries to have at that day around the OCC initiative that a brief spike to a very high
least 15 time points to examine for each of the temperatures, but temperature, such as +55°C for 15 minutes, might “cook” the vaccine
certainly a minimum of 5. He said this would provide a much better entirely, but its VVM might not detect it because the vial returned
estimate of the rate constant. He said then one would need to discuss quickly to a lower temperature. He commented on a device that Prof
whether to calculate the 90% or the 95% confidence limit for the results. Labuza had shown about a high-temperature detector that could
In regard to the number of separate “tags” [TTIs] to study [at each indicate if such an event happened, but it might cost another five cents.
temperature], they should be “more than 10 and less than 100”, but Prof Labuza confirmed that it would be more expensive. Dr Weniger
perhaps the higher number is not needed now that reliability of the returned to the issue of designing a study for the OCC initiative at high
TTIs is now greater than in earlier days. ambient temperatures, referring to the minimum of 5 temperatures that
You need data on the vaccines also, as well, he said, and it should Prof Labuza had suggested. He asked what those temperature might
be done at the same time as the TTIs, because there is some be. Prof Labuza indicated the importance of knowing the denaturation
insulating effect from the bags, and one wants to ensure both the TTIs temperature, which for many proteins is around +55°C to +58°C.
and vaccine share the same conditions. He said a problem was getting Commenting on the subject, Dr Kartoğlu said that there were an
tight temperature control in an incubator chamber with air as the unlimited number of hypothetical situations, such as temperatures of
medium. Water baths could achieve this at much lower cost. +55°C that might occur, for which there is no knowledge. He urged
A followup question, remarking on the bright conditions in some realism in thinking about these remote possibilities. He said they
outdoor immunization sessions shown in the movie, asked about the already conducted stress tests of cold boxes at temperatures of up to
effect of light and whether there should be a separate study arm +43°C for 48 hours, for which there is no actual situation in the world.
exposed to light, and how much light, because it is known to accelerate Dr Prusik asked about his understanding that when proteins in liquid
a VVM reaction. Prof Labuza thought that would be worth doing, are denatured, they coagulate and turn white. If so, for hepatitis B
particularly with special UV bulbs that simulate the spectrum of sunlight, vaccine, for example, such a change of appearance might serve as an
as he has not seen any published data on that question. indicator of exposure to damaging high temperature. Prof Labuza
Mr Zaffran commented that the VVM-anniversary movie showed only answered that some proteins might coagulate and settle while others
3 vials that were in use for immunization were exposed to ambient light, would remain in solution, citing the example of whey protein, which
and that the rest would be kept in the usual darkness of cold boxes. when boiled for hours still remains crystal clear.
Any acceleration of the VVM reaction as a result would simply Mr Steinglass admitted that during his work in the smallpox
encourage one to use up the vial more quickly. eradication campaign, he kept vaccine in his pocket in the desert of
Prof Colton asked about any new technologies in development in Ethiopia, one of the hottest places on earth,. But smallpox still got
regard to TTIs or freezing detectors that would be suitable for what is eradicated. Prof Colton commented that his body may have provided a
needed for vaccines. Prof Labuza said no [although later showed cooling effect. Getting serious, Mr Steinglass pointed out that he has
prototypes of a small, simple freeze indicating sticker his group was worked in places in which the temperature is over +40°C, for example
developing that was intended to be attached to products]. in Oman, where he worked for some years for WHO, it is over +45°C
Mr Evans commented that many of the devices Prof Labuza showed for three months of the year. Thus there would be places where a
would be very relevant to cold-chain monitoring, for which the VVMs

Technologies & Logistics Advisory Committee - Meeting Report

Geneva, 11-12 March 2009 Page 18 of 30
 vaccine [OCC] would be seriously stressed, he said, including in a car, exposed to temperature spikes, which of course would be totally
even within a cold box. unacceptable.
His primary point was, he said, that any “out of the cold chain” policy Prof Labuza mentioned his work with the transport chain for fresh
may be misinterpreted, similarly to how the existing MDVP has been flowers, in which each person among the many handlers from nursery
reinterpreted over the years in ways that were not intended. This to consumer is allocated a maximum amount of shelf life that they may
resulted from well-meaning simplifications to make the policy easier to use up. If they use more than the expected time, there is a penalty.
understand, but which may have been wrong. He used the example of DHL has put such a system into effect.
an elegant explanation -- which happened to be wrong: “if it’s liquid, Mr Davis urged a balanced view of the OCC strategy, which is not at
you do this, if it’s a frozen vaccine you do that”. Mr Steinglass warned all new. He cited a “fast chain” program introduced in South Sudan by,
that rules and policies often get misinterpreted and diluted as they are he believed, Hans Everts and colleagues about 10 years ago. So there
passed down the system. He also feared that the OCC could be is already a decade of experience with this methodology. Mr Davis said
interpreted to mean “all rules are off, you can keep it in your pocket”. he recently learned that the Northern Kenya Turkana has also been
Even though OCC proponents have reassured that “they won’t do that”, using this “fast chain” method for just as long. It’s not new, even if not
he countered that “they may do that”. the same idea as the proposed OCC. So he advised against being
For this reason, Mr Steinglass said he liked the suggestions in Mr scared of OCC. He asked what were the maximum temperatures and
Evans’ paper on MDVP that there needs to be a lot of improvement in expected durations of exposure in those areas where it was proposed
vaccine handling. He repeated his fear that an OCC policy will be that OCC might be used for hepatitis B vaccine.
“dumbed down to mean anything goes”, so it needs to be discussed Mr Munck weighed in by saying that nobody is talking about
carefully. Prof Labuza suggested it might be useful to identify some “anything goes”. By the end of this year, he said we would know how
field sites where people are handling new vaccines and get some real good the VVM is [at elevated temperature] and then we can start
data on what is going on. thinking about how to make the cold chain more flexible. He dismissed
Mr Zaffran clarified that the intention of the OCC initiative is not to some of the worries expressed about doing away with the cold chain.
eradicate the cold chain, nor to get rid of its rules, but meant to provide He said even sceptics would agree that we can go up from +2°C to
some flexibility that will, of course, need management controls and +8°C for a cooler chain, and that a pilot project would be useful. He
training. We need to know whether we can advise some flexibility for said there was already knowledge about what many vaccines can
countries. Even just keeping the vaccines at +25°C in cool rooms tolerate and there were new methods to be pursued. If such pilot
would be a major gain. We want simply to explore this strategy as a studies could be started soon, for once, he said, “we would be ahead
possibility, while avoiding any risk and without implying any liability for of events and not behind them”. There has not been any fresh thinking
the vaccine companies. We would avoid the risk of vaccines being about cold chain for the past 30 years, he added.

3.3.6 Summary of TLAC discussion of the use of VVMs in the MDVP revision
and the OCC strategy, including needs for further clarification or
research
Prof Colton summarized the several sessions and subsequent discussions devoted to TLAC’s
education in the area of VVMs. It was thus prepared for enlightened discussions on using
vaccines out of the cold chain and on revisions to the multi-dose vaccine policy.
TLAC identified a number of open issues:
• TLAC would like to see more raw data on vaccine potency and VVM response. This
would allow TLAC to see the scatter, mean values, standard deviations, and confidence
intervals. Also, TLAC would like to see more plots of vaccine potency versus VVM
response, such as Dr Kartoğlu presented. This would allow one to determine how well
vaccines are matched to their VVMs.
• TLAC considered it useful to develop a life-cycle, economic model of VVM use. This
would allow one to see the impact that VVMs have, as well as to perform economic
trade-off studies for various scenarios.
• TLAC considered freeze monitors that could be applied directly to vials would be useful
in monitoring the cold chain. One issue with their use is that they should be a stock
management tool, similar to VVMs, rather than an “oops” monitor, the only purpose of
which is to tell a health workers to dispose of a vial.

4. Use of vaccines out of the cold chain (OCC)

The session objectives were to update the TLAC on the progress, protocols, and plans for
studies on the use of hepatitis B vaccines out of the cold chain (OCC) and brief TLAC on new
ideas for solving challenges to the cold chain through “cool rooms” and domestic refrigerators.

4.1 The collaborative methodology used for OCC studies

Dr David Wood of WHO explained that the WHO collaborative study for the Out-of-the-Cold-
Chain (OCC) initiative is a scientific study designed to provide evidence for decision-making,
defined at the outset, in consultation with WHO and participants. He pointed out that a WHO
study usually provides information that cannot be obtained from studies in laboratories

Technologies & Logistics Advisory Committee - Meeting Report

Geneva, 11-12 March 2009 Page 19 of 30
 working by themselves, and provides scientific information about the materials studied and the
assay systems in current use.
He listed three principles used to outline the methodology to be used for WHO studies. These
are, firstly, that WHO, through Expert Advisory Bodies or working groups, makes broad
outlines of the studies to be pursued; such Expert Advisory Bodies do not normally contribute
to the specific details of a collaborative study design. Collaborative studies should be
organized by one or more scientist(s) familiar with the appropriate biological field, working
closely with an experienced biometrician.
Secondly, the planning and design phase of the study requires up-to-date knowledge about
the structure and function of the biological material, the nature and limitations of assays
currently available, the availability of potential study materials, and the availability of potential
participants. The planning phase will address the suitability of the proposed assay methods,
the number of materials that can be tested in each assay, the number of assays that can be
performed, and the timeframe for the study.
Thirdly, the report of the collaborative study will be prepared by the study leader which is then
reviewed with the study participants and, where appropriate, the working group. The
individual samples are identified by code and their identity is not disclosed. The final report
will be provided to WHO, and shared with the relevant advisory body and scientific community
for comment. The study leader (and his/her institution) is responsible for archiving the raw
data from the study.
Dr Wood suggested that two members of the TLAC subgroup on OCC might join the advisory
group set up by WHO for the OCC study and thus increase the understanding for TLAC of the
activities and deliberations of that other group. He concluded by referring TLAC for further
information on this process for collaborative studies to Annex 2 (pages 107 to 114) of the
WHO Technical Report Series number 932, published in 2004. o
Discussion of 4.1: The collaborative methodology for OCC studies
During questions and answers, Dr Wood was asked to explain the that transgenic mice can be used instead of monkeys for polio vaccine
process for reviewing protocols, and to give examples from his own testing. He said that study, which lasted many years, used a process
experience for how advisory bodies have been involved with protocol similar to that set up between for the Hepatitis B and OCC
design. He explained that advisory committee involvement is collaborative study, where a study working group oversaw the work
determined on a case-by-case basis that varies by protocol. and the committee relied on the working group to run the study while
He gave the example of his involvement in a collaborative study run they focused on using that work to build on their core responsibility of
by the National Institute of Biological Standards and Control (NIBSC) of bringing about a policy change. This model allowed for smooth
the United Kingdom when he was working there on a study to show operations of the study and resulted in worldwide policy change.

4.2 Vaccines Out of the Cold Chain: An update on the laboratory studies
Dr Julie Milstien, consultant to WHO, updated TLAC on the laboratory studies comprising the
OCC initiative. She began by reviewing the goals for Project Optimize’s investigation of taking
some vaccines out of the cold chain (OCC) for specified periods of time: (1) to increase the
possibilities to deliver vaccines to hard-to-reach populations, (2) to prevent wastage of
vaccines due to freeze damage, and (3) to liberate cold-chain space now at a premium as new
vaccines enter immunization programs. Vaccines candidates considered for OCC use were
evaluated on the basis of temperature stability, use in a specific population outside of the
usual EPI range, potential for freeze damage, and manufacturer interest. Hepatitis B vaccine
met all these criteria; another potential candidate is human papillomavirus vaccine, for which
manufacturer interest is being explored.
As part of the process to define OCC use, she said WHO proposed to investigate the stability
of hepatitis B virus (HBV) vaccines to exposure at +45°C for 7 days, and the ability of vaccine
vial monitors (VVMs) to correlate with the vaccine’s performance. She indicated that the
VVM30, the monitor generally supplied on hepatitis B vaccines based on its stability
characteristics, is already known to follow Arrhenius kinetics up to +90°C and to show a color
change at approximately 8 days after exposure to +45°C. The time-temperature range of 7
days at +45°C was chosen after discussions with field logisticians as sufficient to allow OCC
use for in-country transport of HBV vaccine and for storage at the periphery. These were the

o. WHO Expert Committee on Biological Standardization. Fifty-fifth report. WHO Technical Report Series
(http://www.who.int/biologicals), no. 932, 2004 (http://whqlibdoc.who.int/trs/WHO_TRS_932_eng.pdf).

Technologies & Logistics Advisory Committee - Meeting Report

Geneva, 11-12 March 2009 Page 20 of 30
 situations most likely to incur freeze damage and to inhibit optimal coverage of the birth dose,
which is the primary use of this vaccine in its monovalent form. An initial laboratory study,
overseen by a panel of national regulatory experts (NRA Advisory Group) with responsibility
for oversight of hepatitis B vaccine testing and regulation, was designed to show the
correlation between hepatitis B vaccine behavior at this temperature range and that of the
VVM. Samples are being received from 6 vaccine manufacturers now supplying this product
through UNICEF to national immunization programs, and an analysis of results by the NRA
Advisory Group is expected by the summer of 2009.
The discussion of Dr Milstien’s presentation covered the terminology for OCC, the scientific
design and goals of the planned studies, the possibility of VVM45s, and the other advisory
group overseeing it, among related topics.
Discussion of 4.1: Vaccines OCC: Update on the laboratory studies
At the beginning of discussion following Dr Milstien’s presentation, example, vaccines could be labelled as “store up to two years at +2°C
several TLAC members commented about the ambiguous meaning of to +8°C, then up to 3 months at +37°C”. This would give countries
the term "out of the cold chain" and suggested that a clearer term be flexibility, as approved by their own regulatory agencies, to store the
used to describe this initiative better. There were also several products in alternative settings, such as air-conditioned rooms. “It
questions and concerns expressed regarding the methodology of the would not be a license to store vaccines at +55°C”, he said. Mr Zaffran
study. suggested that perhaps a TLAC sub-group needed to be formed to
Ms Diana Chang-Blanc of UNICEF asked if the study would examine look at these programmatic implications.
the effect of oscillating temperatures, i.e., holding the vaccine at Dr Milstien re-iterated that the goal of the study is to develop a
different temperatures for different intervals. Dr Milstien replied that the model licensing pathway for certain vaccines and certain immunization
study intends to correlate vaccine potency and the VVMs. As the strategies to see if a relaxed cold chain can be used in certain
VVMs have been shown to account for such temperature cycling, the situations. A primary objective, in addition to identifying that regulatory
study will not specifically look at oscillating temperatures. Dr Benes pathway, is to increase coverage for a specific delivery option – the
asked if in vivo potency assays would be performed. It was explained birth dose of monovalent HBV vaccine.
that these would be conducted during a third stage of the study for Dr Benes expressed his concerns for vaccine use out of the cold
confirmation purposes by manufacturers and/or regulatory agencies. chain, among which was realism for what could be expected to change
TLAC was asked whether +45°C for 7 days was the right metric for at a programmatic level as a result of these OCC studies. He pointed
this study. [After the meeting it was clarified that the study would out that routine immunization services that deliver HBV vaccine will
actually test the vaccine for potency at the time when its attached VVM likely not change their practices, even if the vaccine is licensed for use
converted, “which was likely to be longer than 7 days.”] Although out of the cold chain use. This is because they use it in combination
+45°C was considered a valid temperature to look at, concerns were with other vaccines that will still need to remain in the cold chain.
expressed that the proposed time interval (i.e., 7 days) might not be Dr Osman Mansoor of UNICEF stated there has been good
sufficient to reach remote areas. Mr Davis pointed out that multi-dose experience using HBV vaccines OCC for up to a month in locations
vials might not be used up by 7 days. Dr Milstien explained that the where temperatures are moderate. He wondered if some HBV vaccine
initiative aims to issue a uniform recommendation that all hepatitis B products could be assigned a VVM45, and tested for this, which might
vaccines can meet the metric of +45°C for 7 days. She explained the give countries the option to keep such vaccines OCC for longer than 7
importance of using +45°C as the upper limit to ensure vaccines stay days.
potent and safe for use. As all vaccines in study already meet the Mr Zaffran requested that TLAC members, in addition to asking
criteria of 30 days at +37°C, they should all meet the proposed metric. specific questions on the study and its protocol, should also evaluate
It was pointed out that different HBV vaccines may have different whether this pathway was worth exploring as a means to relax the cold
temperature-stability profiles, thus a blanket OCC recommendation for chain. He asked if this process – a collaborative study with one vaccine,
all of them was not appropriate. Mr Davis asked about the geographic HBV – was the way forward, or are there other avenues that should be
areas for which eventual OCC recommendations by WHO would apply, explored? Dr Deeks agreed that the process, as proposed, is a good
given the greatly varying temperatures in climates around the world. Dr methodology to pursue.
Milstien expressed that this was a matter for the TLAC to review once Dr Weniger stated his concern that these studies had been
the study is completed and results presented. described to TLAC only in several conflicting 1-page or 2-page
Mr Spanner requested clarification for the -20°C temperature used summaries, but without a detailed protocol. Thus, they might not be of
as a negative control in the study. Dr Milstien explained it was to sufficient scientific rigour that would be required by North American or
ensure the vaccine was damaged completely and lost all potency. European NRAs for such a significant change in labelling. He agreed
Several members of the TLAC group asked about raising the that the overall goals and objectives were worthy, but they require
temperature(s) to be studied to +50°C or higher. Dr Milstien replied studies of appropriate design, with sufficient temperatures and
that +45°C was chosen to give flexibility while still protecting the samples to be studied. But as TLAC has not been able to review any
vaccine. She said the limited supply of vaccine samples precluded detailed protocol, he could not say that these particular initial studies
testing at two temperatures. But the study could potentially look at a are suitable.
slightly higher temperature point, if TLAC felt it was important. Dr Milstien reminded TLAC members that the study is being
Mr Zaffran explained that some vaccines are very stable. More overseen by an advisory group of experts – both within and outside of
flexible rules that still uphold safety standards could be applied for WHO – to assure its technical structure and rigour. She said this
them in countries where the standard +2°C to +8°C cold chain cannot advisory group includes representatives of the NRAs of the United
be implemented in every location. He added that there are other Kingdom, Germany, South Korea, and Thailand, who all guarantee the
products, OPV for example, that do need to be kept in the cold chain. study has enough rigour to satisfy their policies. Dr Weniger
However, many other products might be kept at controlled acknowledged the existence of this other advisory group, but
temperatures higher than the range of +2°C to +8°C. He indicated the wondered if it was as knowledgeable as TLAC regarding VVMs as a
purpose of the OCC work was to assure that vaccines which are result of its review over several months and the prior full-day’s session.
indeed stable be licensed in a way that reflects this stability. For For example, whether it was aware that high but brief temperature

Technologies & Logistics Advisory Committee - Meeting Report

Geneva, 11-12 March 2009 Page 21 of 30
 spikes destroying the vaccine might not convert the VVM, and that the first question. TLAC should remember, he said, that the technical
Arrhenius kinetics in this temperature range cannot be proven without rigour of the study was already being overseen by what he believes is
studying at least three temperatures. another dedicated advisory committee that is doing a great job. He
Dr Eggers reminded attendees that TLAC has been asked to look – asked TLAC to consider the very important programmatic
through the work of its OCC subgroup – at the regulatory pathway for consequences of these changes, and echoed the suggestion made by
taking vaccines OCC and the programmatic implications of doing so. Mr Zaffran that perhaps another TLAC sub-group would be needed to
He expressed concern that the current discussion was only looking at focus on those programmatic issues around the OCC initiative.

4.3 Rethinking the cold chain

Two related but independent proposals for addressing increasing challenges to the cold chain
were presented by two TLAC members with long experience and expertise in the immunization
cold chain.

4.3.1 Moving vaccines from cold chain to “cool room”

Mr Mogens Munck presented a draft version of a monograph that will be presented in final
version at the next TLAC meeting. It makes his case to overcome many problems facing the
existing cold chain by moving suitable vaccines at the health centre level from refrigerators at
standard +2°C to +8°C temperatures to rooms held in the range of +20°C to +25°C
temperatures by thick walls and conventional air conditioners.
This idea to “relax the cold chain” would avoid the problems of vaccine freezing in existing
refrigerators and lack of current cold-chain space for many new and bulky vaccines. A basis for
it can be found in WHO publications on temperature stability of vaccines, which indicate that
some in the EPI can be kept safely in +20°C to +25°C temperatures for up to one month. p,q
Mr Munck argued that this “cool chain” at the health centres could avoid the need for
refrigerators. In addition, the cool rooms would offer almost unlimited storage capacity
compared to refrigerator volumes, thus solving the looming problem posed by single-dose vial
presentations and prefilled syringes, to be used for many new vaccines. He said such
packaging requires up to 15 times more storage volume at the health-centre level. Existing
store rooms could be upgraded by installing air conditioners, removing any heat-generating
equipment such as refrigerators and freezers, and increasing insulation of the walls.

4.3.2 Increasing cold-chain flexibility at health centre level

Mr Søren Spanner presented his alternative or complementary idea to Mr Munck’s: that of
using conventional home refrigerators set to higher temperatures well away from freezing to
store vaccines up to +15°C, instead of at +2°C to +8°C in typical cold-chain refrigerators. He
said that such domestic refrigerators are usually upright and have two doors with separate
compartments for freezing and refrigerating.
The freezer sections could be used for making ice packs for reconstituted vaccines and for
transporting OPV, as well as for storing OPV itself. The refrigerator sections could be used for
freeze-sensitive vaccines and for making cool [water] packs to transport such vaccines, which
could also be made in existing cold-chain refrigerators.
He said that the introduction of new vaccines and the use of cool packs to prevent freezing
during transport is requiring major and difficult expansions of the existing cold chain, which
needs up to a ten-fold increase in capacity. He argued that his proposal could help achieve
that economically.
The discussion of the presentations by Mr Munck and Mr Spanner focused on challenges to
space in the cold chain, relabeling vaccines that could be stored outside it, the lack of data to
support these novel ideas, and the use of eutectics.
Discussion of 4.2: Rethinking the cold chain
Questions and answers involved both Mr Munck’s and Mr Spanner’s new vaccines and anticipated introduction of additional ones in the
proposals. Several participants expressed their appreciation for the “pipeline” will put increased pressure on the cold chain. In that context,
presentations. There was consensus that the recent introduction of there is a real need to study and fully evaluate such alternative

p. WHO. Thermostability of vaccines. Document WHO/GPV/98.07, 1998, pp. 1-64

(http://whqlibdoc.who.int/hq/1998/WHO_GPV_98.07.pdf).
q. WHO. Temperature sensitivity of vaccines. Document WHO/IVB/06.10, 2006 (http://www.who.int/vaccines-
documents/DocsPDF06/847.pdf).

Technologies & Logistics Advisory Committee - Meeting Report

Geneva, 11-12 March 2009 Page 22 of 30
 temperature storage options and technologies such as those presented. could be presented at the next meeting for review by TLAC in order to
Dr Evans expressed the view that if TLAC wants to have a major provide guidance to WHO.
impact, it should focus WHO’s efforts on working with manufacturers to The use of eutectic materials was also discussed. These are specific
get vaccines labelled for their true stabilities. Achieving that goal would mixtures of salts in a solvent, usually water, which have a
be critical, he said, in adopting any of the alternative options presented. freezing/melting point at a temperature above 0°C, maintaining a
Mr Steinglass noted that the projections shown for future vaccine stable temperature higher than freezing as they absorb their heat of
storage needs may be underestimates because they do not account for fusion in melting. Several participants noted their key advantages: they
the surge capacity needed for special campaigns. can be frozen in normal refrigerators (e.g., at +5°C) without the need
Dr Piyanit stressed that the current lack of data supporting these for a freezer; when frozen they can keep vaccines cool without the risk
novel proposals is a barrier to their adoption, and suggested that of freezing them. It was pointed out, however, that once the vials for
consultants be engaged to document existing evidence of some vaccines are opened, a regular ice pack or water-based ice is
effectiveness of using the higher temperatures discussed. Such results required to keep them cold.

4.4 Summary of OCC session and TLAC discussion and recommendations

for OCC pending working group deliberations
The session served to update TLAC on the methodology used by the WHO collaborative study
of hepatitis B vaccine, including the existence of a separate advisory group for it. WHO asked
TLAC specifically whether studying these vaccines at +45°C until their VVM changes colour was
acceptable for the first phase of this study. Notwithstanding discussion about the need by
TLAC to better understand the methodology, and differing points of view expressed on the
suitability of its design, TLAC agreed by a vote of 8 members to 3 members that this would
suffice.
Secondly, TLAC discussed the many implications for the OCC study and its application. A
potential need was pointed out to consider these separately, i.e. the regulatory pathway to
label changes versus programmatic aspects for implementing OCC. Finally, it was determined
that two members of the OCC subgroup – Dr Piyanit T. and Dr Deeks – will be invited to sit on
the advisory group for the hepatitis B study to facilitate understanding of the study and allow
the TLAC subgroup to fulfil its terms of reference.
The second part of the session was devoted to a discussion document by Mogens and Soren
on two visions for relaxing the cold chain. This item was for information. It was well received
and TLAC hoped it would remain as a future agenda item to allow more discussion. It will also
be discussed at the OCC subgroup level.

5. Revision of the multi-dose vial policy (MDVP)

The objectives of this session were to inform an ongoing effort to revise the existing WHO
policy on how health workers should handle multi-dose vaccine vials once opened, which is
challenged by new vaccine formulations and presentations which will render the existing MDVP
obsolete. The current policy applies two distinct rules: (1) Discard at the end of the current
vaccination session, or by 6 hours after opening, whichever occurs first. (2) Vials may be used
for vaccination in subsequent sessions up to 28 days after opening, as long as they are kept
under proper refrigeration and their VVMs have not reached endpoint.
The session previewed a draft document prepared for the TLAC to outline key issues to
address in a revised MDVP. The document identified additional issues to be considered, and to
develop a process for information-gathering, documentation, discussions, and other work to
be able to complete the effort for a revised recommendation by the end of 2009. r

5.1 Analysis of issues and revision of the MDVP and proposed way
forward
WHO consultant Dr Peter Evans was engaged to review the landscape surrounding the existing
MDVP and the factors requiring its updating. He indicated the existing policy must be
expanded to take into account the introduction of new vaccines and to become part of a
comprehensive vaccine handling approach. Such a policy needs to be broad and include
situations currently considered unusual, such as (1) extracting multiple mini-doses for dose-
sparing purposes from vials labelled for fewer nominal doses, (2) mixing separate vaccines at
the point of use to make a larger combination product, (3) taking vaccines beyond the cold
chain as per the OCC initiative, etc.

r. TLAC requested that related document drafts for review and consideration should be distributed to TLAC members
at least 2 weeks prior to a meeting.

Technologies & Logistics Advisory Committee - Meeting Report

Geneva, 11-12 March 2009 Page 23 of 30
 The effort was complicated, he said, by many intertwined issues and tradeoffs that will have to
be made. He advised that any new policy must deal with both the short term and long-term
use of vaccines and must be easy to keep current. Finally, it must ensure that its statements
and recommendations have a scientific basis and will further full compliance by all parts of the
EPI system.
Discussion of Dr Evans’ topic covered the growing need for an MDVP revision and its features,
packaging trends, signals to industry, and the involvement of related advisory groups and
bodies.
Discussion of 5.1: Analysis of the MDVP and proposed way forward
During extensive comment following the presentation by Dr Evans, it It was mentioned that one precedent for healthy dialogue between
was pointed out that the range of vaccines available on the market has the public and private sectors has been the Vaccine Presentation and
widened considerably since the policy was originally established some Packaging Advisory Group (VPPAG), composed of representatives
years ago, so a revision is imperative to reflect the newer vaccines. from UNICEF, WHO, industry, NGOs, and other stakeholders in global
The discussants observed that the trend in vaccine packaging immunization. A member of the VPPAG in attendance at the TLAC
among vaccine producers in the International Federation of meeting noted that it has been a useful forum for exchange and
Pharmaceutical Manufacturers & Associations (IFPMA) is towards negotiation among the sectors it represents, and that industry has been
single-dose vials which do not require antimicrobial preservatives, and receptive in responding to the needs of developing countries. One
towards innovative presentations that serve industrialized market recent example cited was the agreement by GlaxoSmithKline to
needs (e.g., the Prevnar®/Prevenar® pneumococcal conjugate vaccine change the packaging of oral Rotarix® rotavirus vaccine from one
in pre-filled glass syringes). consuming 100 cubic centimeters (cm3) of volume per dose in the cold
TLAC members agreed that if this market evolution would not meet chain to another requiring only 70 cm3 per dose.
the needs of developing countries, the public health sector would need It was pointed out that the perspective of manufacturers in the
to signal clearly to industry what characteristics and packagings are Developing Country Vaccine Manufacturers Network (DCVMN) is likely
preferred. This signal should be early in the development phase. The to be different from that of members of the IFPMA, serving as they do
public sector also must understand that its needs have a cost, for they different markets with different needs and distinct regulatory
would require manufacturers to alter their production processes. frameworks. Nevertheless, the public sector was warned to expect that
Another commenter said that private industry does not respond solely manufacturers will all be under pressure eventually to move away from
to “needs”, but responds to actual demand that is backed up with mercury-based preservatives in vaccines, which have generated
financing to buy what is produced. controversy in some developed countries.

TLAC Consensus
TLAC members indicated their support for the initiative to revise the multi-dose vial policy.
TLAC members reiterated that the new MDVP should be based on data for both current
vaccines, and the expected ones of the future. There was general agreement that the policy
must be easy for health workers to comprehend and apply, so they are reassured in their
decision-making. It should include simple measures, such as clear labelling of which vaccine
vials should be discarded promptly and which may be kept for another day.

5.2 Current challenges and interim solutions for the MDVP

In consecutive presentations, Dr Rudi Eggers and Dr Nora Dellepiane, both of WHO, honed in
on approaches for revising the MDVP, building upon the information gathered by Dr Evans.
They focused on three distinctive areas of work to be addressed in completing the MDVP
revision:
• Providing a “visual cue” to the health worker
• Information to set the policy on which vaccines “to keep” and which “to throw away”
• The preferred presentations of vaccines
They reiterated an earlier point that a new MDVP for when to discard an opened vial needs to
be part of a larger, more comprehensive vaccine handling policy. As suggested in the analysis
by Dr Evans, it should clarify with vaccine manufacturers if a third discard category after
opening is required beyond the existing “discard at end of current session or 6 hours,
whichever occurs earlier” and “discard after 28 days” (if stored under specific condition. A new
category might require discard much earlier than 6 hours, as for very labile varicella virus
vaccines.
The presentations by Drs Eggers and Dellepiane strongly suggested that the vaccinator in the
field should be able to know which discard policy to follow simply by looking at the vial, which
would contain some clear indicator, or “visual cue”. They said this would likely require some
feature on the label or vial, would require cooperation by vaccine manufacturers, and might
become a future label specification for pre-qualification of vaccines.

Technologies & Logistics Advisory Committee - Meeting Report

Geneva, 11-12 March 2009 Page 24 of 30
 Possible visual cues and rules they listed were:
• For EPI vaccines, the absence of any marking requires the earliest discard category
• Coloring of the label background (e.g., orange = discard at end of session; blue = keep
for subsequent sessions)
• An icon, symbol, or logo
• Placement of the VVM on the vial cap or ampoule top versus on the vial label
The presenters urged that development and agreement on such a visual cue should occur
rapidly, and that a decision should be made by the September 2009 TLAC meeting, following
appropriate consultations and focus group assessments. Once preliminary agreement is
reached, the WHO staffers indicated that the visual cue should be pilot tested in a country.
Changing the discussion to the broader MDVP revision, the speakers indicated that further
information and data will be needed to demonstrate the scientific bases of the new policy.
They suggested several studies supporting the MDVP should be commissioned to allow
evidence-based thresholds and criteria to be set. These studies may include, inter alia, (1)
monitoring the introduction of visual-cue-equipped vaccines in the field, (2) assessment of the
contamination potential of a pierced vial septum upon submersion in water, (3) confirmation
that 6 hours is indeed the appropriate time to discard vaccines heretofore following that rule,
and (4) studying alternative preservatives to mercury.
Drs Eggers and Dellepiane indicated that in the interim while such new data was being
pursued, new vaccines submitted to the WHO pre-qualification team will be evaluated on a
case-by-case basis and assigned to the appropriate MDVP existing category. As information
and data become available, and as pre-qualification decisions are made, this temporary
process can be changed to incorporate a more long-term, rules-based approach.
During the current, interim process, Dr Dellepiane advised vaccine manufacturers to provide
adequate information for pre-qualification to assess the following questions:
• If vaccines containing standard amounts of thiomersal as preservative (50 to100 µg/mL)
can be kept beyond the end of the session?
• If vaccines containing no preservative must be discarded at the end of the session or
within 6 hours, whichever comes first?
• What policy discard should apply for vaccines containing reduced amounts of
thiomersal or alternatives to this preservative?
The presenters said that vaccine manufacturers should be provided with guidance on what
vaccine presentations would be preferred for pre-qualified vaccines. Dr Eggers cited the
considerable work VPPAG has already done to establish the preferred "generic" vaccine
presentations, considering both programmatic and regulatory considerations. In addition, it
was stated that WHO needs to establish the relationship between such "preferred
presentations" and the pre-qualification requirements. A key issue to be faced, they said, was
whether WHO should pre-qualify vaccines that substantially depart from a preferred
presentation, or should WHO grant pre-qualification but declare them programmatically
difficult to use.
Discussion of the presentation by Drs Eggers and Dellepiane included the example of WHO
recommendations for off-label use of rabies vaccine, applicability of an MDVP to unusual
situations, the process for developing visual cues to prompt the correct discard policy, and a
WHO suggestion for a potential visual cue.
Discussion of 5.2: Challenges and interim solutions for the MDVP
The two presentations by WHO staff on the MDVP generated much One discussant pointed out that field realities may make it
comment and discussion by TLAC on many disparate but MDVP- impractical for the planned MDVP revision to include new rules and
related topics. One discussant pointed out that two WHO expert means to cover the use of fractional doses and a third category of
committees already recommended a current practice in many vaccines requiring discard much earlier than six hours. It was
developing countries of using rabies vaccine in “fractional” dose- suggested that health workers should not be involved in “developing
sparing amounts from nominal single-dose vials without thiomersal products” and that careful consideration must be given to which
preservative, which are then stored opened under refrigeration over a vaccines can safely be included into the MDVP.
period of weeks. It was suggested that TLAC should review how it Regarding the process for inventing, testing, developing, and
might interface with and be of relevance to such other WHO expert implementing a visual cue, there was debate among TLAC members
advisory bodies. on the appropriate balance between speed and prudence. Dr Weniger

Technologies & Logistics Advisory Committee - Meeting Report

Geneva, 11-12 March 2009 Page 25 of 30
 said it would be important to engage professionals in iconography and position on the vials where the VVM is applied could serve to prompt
their recognition for such an important process. Experts from the for the correct discard policy. This policy is already in place as a result
advertising and related fields, he said, might be able to dream up of their prequalification review: All VVMs on vaccines to be discarded
various candidates as visual cues, or suggest modifications to or the same day as opening are placed on the cap or the top of the
weaknesses of existing ideas, and then carefully test the most ampoule, so that they are no longer connected to the container once it
promising ones in various cultures and language groups on different has been opened. VVMs are placed on the labels for other vaccines
continents for their ability to prompt the correct discard policy. On the that can be kept for sessions on subsequent days (under certain
other hand, there was some pressure expressed that the MDVP conditions). They indicated however, that if this current practice is
needed to be revised as quickly as feasible, since new vaccines that formalized in a new MDVP, the increased responsibility would strain
undermine existing rules-of-thumb are coming soon onto the market. the Prequalification team, which is already stretched in responding to
Staff of the WHO office involved with prequalification of vaccines the many dossiers of vaccine products submitted for prequalification.
proposed that the VVMs – already widely recognized – alone might be TLAC members emphasized the importance of ensuring adequate
adapted to serve as the desired visual cue. Their idea was that the support and resources for this essential prequalification process.

TLAC Consensus
There was general agreement by TLAC members for an overriding principle of simplicity
regarding a revised MDVP, to minimize confusion, with a maximum of only one or two visual
cues to be developed. In addition, the default position for an MDVP ought to be that pre-
qualified EPI vaccines follow the earliest discard policy, unless otherwise indicated by a visual
cue that informs health workers they may deviate from such default rule. Concern was raised,
however, for confusion and inadvertent wastage of privately-procured vaccines in “mixed”
formularies that also contain vaccines procured outside the UNICEF system. Such vaccines
without a visual cue or a VVM might be unnecessarily discarded, even though still safe to keep
opened for weeks under refrigeration.

5.3 Review of activities for delivery of reduced doses of vaccines and

related programmatic issues
On a tangential topic relevant to opened vaccine, Dr Martin Friede of the WHO Initiative for
Vaccine Research reported on several research groups and vaccine companies which are
investigating the potential of reduced-dose intradermal (ID) delivery to permit cost-savings
and extend population coverage for scarce vaccines. He said this approach was already used in
many parts of the world for pre- and post-exposure prophylaxis against rabies, in which
reduced 0.1 mL ID doses are removed from the 1.0 mL contents of a mono-dose vaccine that
is labeled for intramuscular (IM) delivery. This approach has been shown to induce
seroconversion with several types of rabies vaccine.
He said such removal of multiple fractional doses from mono-dose vials raised two
programmatic issues to be considered in assessing such dose-sparing. First, the practice
essentially converts a single-dose vial into a multi-dose one. As most ID doses are 0.1 mL, 5
or 6 doses could be taken from a standard 0.5 mL single-dose vial of vaccine.
If the vaccine does not contain a preservative, the second and subsequent withdrawals risk
being contaminated by overgrowth of bacteria introduced by the first needle entry. Dr Friede
suggested that the new policy on multi-dose vials should apply to this situation. For example,
it might recommend that preservative-free vaccines be discarded within a certain numbers of
hours after its seal is first punctured.
The second programmatic issue for MDVP that Dr Friede raised was that the rubbery stoppers
on vaccine vials may not be safe if penetrated by needles more than a certain number of times.
Consecutive cuts by the needle at the same spot may by a coring effect cut off pieces of
stopper which enter the vaccine and perhaps the patient. A general rule is to penetrate
stoppers no more than 20 times. But taking 0.1 mL ID doses from a common 10-dose, 5 mL
vial could require up to 50 stopper penetrations. He said the use of blunt 'boring' spikes into
vials by single-dose adaptors for filling ID delivery devices might also exacerbate
disintegration of the stopper and contaminate the vaccine with stopper fragments. He noted
that the use of vial adaptors that remain on the vial to fill multiple ID cartridges could possibly
eliminate this risk.
As this presentation was primarily for information purposes, there was no discussion by the
committee.

Technologies & Logistics Advisory Committee - Meeting Report

Geneva, 11-12 March 2009 Page 26 of 30
 5.4 Summary of MDVP session and TLAC discussion and preliminary
recommendations on MDVP pending working group deliberations
Consultant Peter Evans reviewed the existing MDVP and related guidelines and framed the
issues for TLAC consideration. TLAC members agreed that an updated MDVP is needed to
accommodate new vaccines, which have formulation and presentation characteristics that
render obsolete the existing MDVP policy. The typical interpretation of MDVP by vaccination
workers in the field, according to whether the vaccine is freeze-dried or liquid, will no longer
work with some of the newer vaccines.
The availability of these new products has outpaced the reformulation of policies to properly
handle them. The trend in developed countries is to move away from preservatives. Yet multi-
dose vials without preservatives are at risk for bacterial contamination in areas with poor
vaccine handling and injection practices and therefore pose safety concerns. A new
comprehensive vaccine handling policy is needed, of which revised rules for how long opened
vials may be used is just one component.
Mr Evans proposed a way forward along technical, scientific and political lines. The annex to
the review document he initiated will list the properties and presentations of available and
upcoming vaccines that was not ready for review by TLAC at the time of this meeting.
Drs Dellepiane and Eggers of WHO outlined a methodology with stakeholder roles defined to
address policy changes and interim solutions in the area of visual cues for health workers.
Information is still needed to set the proper policy and determine the preferred presentations
of vaccines. The WHO staff reinforced the consultant's point that the new MDVP needs to be
part of a larger, more comprehensive policy for handling vaccines, although it was still
amenable to addressing in a stepwise fashion. WHO indicated that both short-term and long-
term actions are needed.
WHO will use a case-by-case approach in the interim, to be replaced later by a rules-based
policy, as additional needed information becomes available. TLAC endorsed such a stepwise
process, as was outlined, and anticipates reviewing WHO’s timeline for moving forward. TLAC
also looks to WHO to clarify the future relationship between TLAC and the separate Vaccine
Presentation and Packaging Advisory Group (VPPAG).
There was general TLAC consensus that one or two icons would be needed to indicate clearly
to health workers whether opened, pre-qualified vaccine vials should be discarded at the end
of each vaccination session, or could under certain conditions be used on subsequent days.
The physical location itself of a visual cue on the vial (e.g., where the VVM is located ) was not
deemed sufficient on its own to signal correct handling of the opened vial, particularly when
the discard policy may no longer correlates with old rules-of-thumb for lyophilized or liquid
vaccines.
TLAC suggested that any such icon(s) be professionally designed and tested in diverse cultures
to ensure correct and easy interpretation by health workers. TLAC members have developed
some prototype icons as examples. TLAC welcomes the opportunity to review WHO’s proposed
methodology for design and testing of the icon(s).

6. Injection Safety
This session was for informational purposes to inform TLAC on some activities and issues
regarding the safety of injections. No general policy recommendations were sought.

6.1 Safe Injection Global Network (SIGN): current and future activities
of importance to TLAC
Dr Selma Khamassi, representing the WHO/SIGN Secretariat, reported on the activities of this
initiative (http://www.who.int/injection_safety/sign/en/) to reduce the many dangers and
drawbacks of conventional needles and syringes, among other goals. These included producing
a variety of assessment tools, publications and policy recommendations, participating in the
development of international standards for autodisabling syringes, s and assisting in the WHO

s. ISO. Sterile hypodermic syringes for single use - Part 3: Auto-disable syringes for fixed-dose immunization.
Number 7886-3:2005(E)
(http://www.iso.org/iso/iso_catalogue/catalogue_tc/catalogue_detail.htm?csnumber=36742).

Technologies & Logistics Advisory Committee - Meeting Report

Geneva, 11-12 March 2009 Page 27 of 30
 prequalification of these and other relevant technologies for safety such as sharps boxes,
needle removers, and cold-chain equipment. Some of these represented major achievements
in assisting WHO Member States to assess their injection practices, identify safety breaches,
and plan and implement national programmes and strategies for safer injection.
She said there was increasing recognition that interventions to improve the safety of injections
represented a very cost-effective means to prevent the transmission of blood-borne pathogens,
which are mainly hepatitis B, hepatitis C, and HIV. Transmission occurs to patients through
unsterile reuse of injection equipment and mishandling of multi-dose vials. Healthcare workers
become infected through needlestick injuries. The community at large is at risk through
improper management and disposal of sharps waste.
The second part of her presentation reviewed and discussed issues around various devices for
safer injection and their applications and market trends. A major component of these are
auto-disabling (AD) syringes intended to prevent inadvertent or intentional reuse of unsterile
used syringes. She said that such syringes have been categorized by whether they are
“passive” or “active”.
Passive devices have internal features that do not require the health worker actively to
cooperate with preventing reuse; they automatically prevent the syringe plunger from being
pulled back a second time to permit a second, improper use of the device. Thus, they are
described as “passive”.
In contrast, “active” devices require a health worker to carry out a separate, additional step to
disable the syringe for any reuse. An example of these are syringes that use an internal spring
to propel the needle back into the barrel of the syringe. These require a final push to activate
the spring during the last couple of millimetres of plunger stroke. She indicated this is a
problem when the health worker intentionally avoids that final push, permitting him or her to
reuse that same syringe for many injections on the same or different patients.
Other terms used are used in describing needle-syringe devices. One is “Re-use Prevention”
(RUP), which is not exactly synonymous with auto-disabling, as it applies to any method to
prevent re-use, either passive or active. “Anti needle stick” (ANS) describes devices or
methods to shield the needle after use to avoid accidental poking of patients or health workers
with used needles. Some shield automatically and do not require health care workers to do
anything. Others are “manual” shielding and this step must be initiated by the health worker,
such as folding down a needle cover, or activating the spring as described above. Dr Khamassi
showed the relative costs of various syringe types.
The last part of the presentation reviewed SIGN activities of interest to TLAC. For example, the
new funding opportunities for safer immunization through GAVI and for therapeutic injections
through the Global Fund for HIV, TB, and Malaria. SIGN is advising on the safety aspects of a
trial of BCG vaccination with needle-free jet injectors, and collaborating with EPI on preventing
re-use of glass syringes of the new pneumococcal conjugate vaccines. It is working with the
WHO Performance, Quality, and Safety (PQS) unit to develop specifications for safety boxes
and needle cutters, and reviewing policy on which specific AD syringes should be used in the
EPI and for other injections, particularly for their reuse and needlestick prevention features.
The discussion of the Dr Khamassi’s presentations dealt with the ISO standards for AD
syringes, their WHO prequalification status and future, and medical waste.
Discussion of 6.1: Safe Injection Global Network activities
Dr Weniger thanked Dr Khamassi for her presentation, and began barrel to compensate for not pushing that same distance at the end of
the discussion by indicating his familiarity with the problem she the stroke, thus permitting an irresponsible health worker to give full
described, that some spring-loaded syringes were not truly auto- doses to consecutive patients with a single reused, unsterile syringe.
disabling because they required the health worker intentionally to During the discussion, Prof Colton pulled up on his computer the
activate the re-use prevention feature (which also activated the applicable international standard. s It stated “The syringe and needle
needlestick prevention feature as well). He wondered how these shall be passively and automatically rendered unusable by the delivery
spring-loaded needle-syringe devices had been able to achieve of the intended fixed dose. No secondary or additional action on the
prequalification as AD syringes for the UNICEF procurement, because part of the user shall be required.”
they did not seem to satisfy the spirit, if not the letter, of the Dr Kartoğlu of the WHO PQS unit explained that all AD fixed-dose
International Organization for Standardization (ISO) standard for auto- syringes have "passive" disabling mechanisms by nature to comply
disabling syringes, which requires “passive” activation of the AD with the required ISO 7886-3 standard. He said activation of the AD
function. Dr Weniger said it was easy to defeat the disabling action by mechanism may occur at three different times during the injection: (1)
overfilling such syringes by 2-3 millimeters past the fiducial line on the when the injection is commenced; (2) when the injection has delivered

Technologies & Logistics Advisory Committee - Meeting Report

Geneva, 11-12 March 2009 Page 28 of 30
 50% of the intended dose; and (3) when all the intended dose has beginning of delivery (type 1).
been delivered. Other points raised during the discussion include that such injection
Dr Kartoğlu further explained that the UNICEF Supply Division had safety issues did not logically fall into the domains of the existing TLAC
been questioning such risks of reuse for those syringes in the type 2 subgroups focusing on the OCC and MDVP topic. But this area is
(disabled only at 50% of delivery) and type 3 (disabled only at 100% of important and should be included in TLAC’s broader discussions on
delivery) categories for when re-use became prevented. UNICEF had vaccine handling policy. Another observation was that retractable-
been requesting the elimination of such devices from prequalification. needle syringes, when used properly, can prevent both reuse and
The response by the PQS unit to this query from UNICEF was that needlestick injuries. However, they must still be discarded with
unless data from the field could document such actual reuse (abuse), standard medical waste and later processed for final disposal, which
PQS could not remove them from its prequalification list. However, Dr can be problematic. Finally, immunization campaigns generate much
Kartoğlu pointed out that it might be possible to prepare a road map such medical waste, and better options for disposal should be pursued.
with their manufacturers to try to convert such devices to activate at the

6.2 Preliminary TLAC recommendations on injection safety

The appropriate units at WHO, including PQS and SIGN and perhaps others involved with
curative medicine, were requested to prepare a report to TLAC on how such spring-loaded
needle-syringe devices came to be prequalified as auto-disabling syringes, and to suggest
possible measures to resolve this risk to immunization and curative programs by their reliance
on them to prevent intentional reuse.
Also requested was a more general discussion outlining the main issues related to injection
safety for immunizations that would be relevant for consideration as part of a major
recommendation on vaccine handling policy.

7. General comment
Throughout the meeting, observers from industry, other international agencies, non-
governmental organizations, and other entities contributed information and comments.
Among these, Dr Christopher Nelson of Merck, a member of the IFPMA, thanked the
organizers for including industry representatives. He said the issues discussed are timely and
relevant to policy development and, as such, have an impact on the WHO/QSS vaccine
prequalification process. This activity, he continued, helps define “the regulatory pathway” for
WHO prequalification and thus UN procurement and availability for public sector programs in
developing countries. He encouraged filling of the vacant “regulatory position” in TLAC.

8. Administration and interactions

During the 2-½ days in Geneva, informal discussions were held among TLAC members, and
with WHO staff, to improve the interactions, sharing of information, and constraints between
the advisory committee and the agency. The goals were to better meet the needs and purpose
of WHO in creating an independent, outside advisory body, as well as to satisfy TLAC members
that they would have the necessary data and scientific evidence on which to apply their
diverse experience and knowledge for recommendations affecting global immunization.
As part of those discussions, Dr Gregory Hammond crafted a framework to guide the work
interactions between TLAC and WHO:
The Technology and Logistics Advisory Committee (TLAC) recognizes its advisory role to
the WHO to support and work with it on policy, programs and operational issues. TLAC
aims for its work to have the following positive outcomes:
- to be important in having a positive impact on heath
- to meet a relevant need
- to be useful in field operations

To permit TLAC to achieve these outcomes, it requires processes that assure optimal work
interactions with WHO, which will:
- encourage anticipation of upcoming issues
- assure clarity of roles, issues, and requests
- provide access to relevant background and context information
- assure transparency
- stimulate creativity
- develop goodwill

Technologies & Logistics Advisory Committee - Meeting Report

Geneva, 11-12 March 2009 Page 29 of 30
 Because of the complexity of the issues, collaboration between TLAC and WHO will
benefit from the following organizational methods by both parties:
- a systems approach to issues, analysis and potential solutions
- agenda that recognize short, medium and long term issues and priorities
- small working groups to lead the response to issues and requests
- acquire and use evidence in its deliberations and recommendations
- obtain the best possible expertise and information to educate TLAC for its work
- seek input and assistance from stakeholders
- use communications technology appropriate to the needs
- seek regular feedback to assure optimal interactions

9. Adjournment
The TLAC meeting concluded in the late afternoon of 12 March 2009, with the next face-to-
face meeting of the advisory committee scheduled in Geneva on 22-24 September 2009. In
the intervening period, the full TLAC membership, as well as its two working groups would
confer by email exchanges and teleconferencing.

10. Disclaimer
This report reflects the discussions and recommendations made by the Technologies and
Logistics Advisory Committee to WHO, but does not necessarily reflect WHO position, unless
specifically stated.

Technologies & Logistics Advisory Committee - Meeting Report

Geneva, 11-12 March 2009 Page 30 of 30