© World Health Organization

WHO Technical Report, Series No. 924, 2004

Annex 3
Guidelines for the production and control of
inactivated oral cholera vaccines

This document provides information and guidance to

national regulatory authorities and vaccine manufacturers
concerning the characteristics, production and control of
inactivated oral cholera vaccines intended to facilitate
progress towards their international licensure and use. The
text is presented in the form of Guidelines instead of Rec-
ommendations because further work is still needed to
develop and standardize appropriate methods and criteria
that will assure the consistent quality, safety and stability
of these vaccines. Guidelines allow greater flexibility than
Recommendations with respect to expected future develop-
ments in the field and indicate present deficiencies.

129

ECB text 129 19/11/2004, 09:49 AM

Black
 1. Introduction 130
2. General considerations 131
2.1 The pathogen and the disease 131
2.2 Protection against the disease 132
2.3 Candidate antigens 132
2.4 Inactivated oral vaccines 134
2.5 Correlates of protection 135
2.6 Production and control of inactivated oral cholera vaccines 136
3. Manufacturing recommendations 138
3.1 Control of starting materials 139
3.2 Control of the manufacturing process 140
3.3 Control of final bulk 142
3.4 Control of final lot 143
3.5 Stability, storage and expiry date 144
3.6 Reference materials 145
Authors 145
Acknowledgements 145
References 146

1. Introduction
A parenterally administered, killed whole-cell cholera vaccine has
been widely available for many years. The WHO Requirements for
this vaccine were first adopted in 1959 and revised in 1968 (1); an
addendum was incorporated in 1973 (2). However, this vaccine offers
at best only limited protection of short duration and produces un-
pleasant side-effects in many vaccinees. In view of these limitations,
the vaccine has not been considered satisfactory for general public
health use, and in 1973 the twenty-sixth World Health Assembly
abolished the requirement in the International Health Regulations
for a certificate of vaccination against cholera.
Considerable progress has been made during the past decade in the
development of a new generation of oral vaccines against cholera.
These have already been licensed in some countries and are now
being considered for wider public health application (3). Two distinct
types of oral cholera vaccine have been developed; those consisting of
live attenuated bacteria and those consisting of killed (inactivated)
bacterial cells. In some cases, the latter are combined with the purified
recombinant DNA-derived B-subunit of the cholera toxin. These
positive developments have led to a need for international guidance
to assure the quality and safety of this new generation of cholera
vaccines. The present guidelines apply only to inactivated oral cholera
G vaccines.

130

ECB text 130 19/11/2004, 09:49 AM

Black
 Because the WHO Requirements (1) for the production and control
of the killed whole cell parenteral cholera vaccine may not be relevant
to the production and control of the new generation of cholera vac-
cines, and because such a vaccine is no longer recommended for
general public health use (although it is still produced in some coun-
tries), as well as the potential for confusion with guidelines relating
specifically to the new vaccines, the Expert Committee for Biological
Standardization, decided at its fiftieth meeting to discontinue those
requirements (4).

2. General considerations
2.1 The pathogen and the disease
Throughout history, the highly pathogenic waterborne bacterium
Vibrio cholerae has caused devastating outbreaks of diarrhoeal dis-
ease in most parts of the world. Altogether seven cholera pandemics
have been recorded, the latest of which started in 1961, and is still
continuing. An estimated 120 000 deaths worldwide are caused by
cholera each year. Humans are the only known natural host for V.
cholerae and the disease is closely linked to poor sanitation. Despite
the availability of oral rehydration treatment, small children and the
elderly are particularly susceptible to the extreme dehydration that
results from severe cholera. Although oral rehydration therapy may
often save lives it has no effect on the course of the disease or on
dissemination of the infection.
V. cholerae is a Gram-negative, rod-shaped bacterium that carries a
single polar flagellum. It is a non-invasive pathogen that colonizes the
epithelium of the small intestine after penetrating the mucus layer.
The organism causes diarrhoea through the secretion of cholera toxin,
the toxic action of which depends on a specific host receptor, the
monosialosyl ganglioside GM1.
Strains of V. cholerae are characterized by serogrouping based on the
polysaccharides of the somatic O antigen. Epidemics have almost
invariably been caused by V. Cholerae of the O1 serogroup. Three
serotypes (Ogawa, Inaba and Hikojima) and two biotypes (classical
and El Tor) have been described, although there is some debate as
to whether Hikojima is truly a separate serotype. Until recently, V.
cholerae of the O1 serogroup accounted for most cases of cholera, but
an additional V. cholerae serogroup, O139, has now emerged as a
major cause of cholera in India and Bangladesh (5). Serogroup O139
is closely related to the El Tor biotype and has now spread over a
large part of Asia. In the 1990s, cholera returned for the first time in G

131

ECB text 131 19/11/2004, 09:49 AM

Black
 100 years to Central and South America. The causative agent in Latin
America is similar, if not identical, to the agent that caused the
seventh pandemic in Asia and Africa, i.e. the El Tor biotype of V.
cholerae serogroup O1.

2.2 Protection against the disease

The available evidence suggests that protection against cholera is best
acquired through oral immunization, either through natural infection,
or by use of an oral vaccine. Data from studies in Bangladesh indicate
that natural cholera infection is about 90% effective in eliciting pro-
tection against subsequent attacks for up to 3 years. Infection with the
classical biotype of V. cholerae (Inaba or Ogawa) appears to stimulate
a more potent, or longer-lasting immunity than infection with the El
Tor biotype (6–8). The traditional killed parenteral cholera vaccine
induces only up to 50% protection for 3–6 months. The limited pro-
tection afforded by this vaccine seems to be due mainly to the route of
administration. Injected cholera vaccine gives rise to little or no local
immune response in the gut where both the pathogen and the toxin it
produces exert their action during infection. The pathogenesis of V.
cholerae involves both the colonization of the intestine and the pro-
duction of the enterotoxin, cholera toxin (CT), which acts locally to
stimulate excessive electrolyte and fluid secretion, primarily from the
crypt cells of the small intestine. Cholera toxin acts by inducing in-
creased formation of cyclic adenosine monophosphate (cAMP) and/
or cyclic guanosine monophosphate (cGMP) in the epithelial cells
resulting in the secretion of chloride and bicarbonate into the lumen
of the small intestine. Other enterotoxins, such as zonula occludens
toxin (ZOT) and accessory cholera enterotoxin (ACE) may also con-
tribute to pathogenesis, but probably play only a minor role. Protec-
tion against cholera may therefore be expected to be provided by
immune mechanisms that block colonization and multiplication of the
pathogen in the intestine inhibit the toxic activity of the toxin, or both.
The ability to stimulate local intestinal immunity is therefore now
considered critical if a cholera vaccine is to offer protection against
infection and the disease (3). In addition, antibodies to V. cholerae
have been found in breast milk and saliva and may be an indirect
measure of intestinal immunity (9).

2.3 Candidate antigens

Cholera toxin consists of five identical B-subunit peptides that spon-
taneously associate to form a ring structure into which the enzymati-
cally active A-subunit peptide is non-covalently inserted. The toxic
G activity resides in the A-subunit while the five B-subunits mediate

132

ECB text 132 19/11/2004, 09:49 AM

Black
 binding of the toxin to specific GM1 receptors on intestinal epithelial
cells and are primarily responsible for the immunogenicity of the
toxin.
The cholera toxin B-subunit elicits an effective antitoxin response
that also offers short-lived protection against disease due to the heat-
labile toxin (LT) of Escherichia coli (10–12). Furthermore, the chol-
era toxin B-subunit appears to be well-suited as an oral immunogen
because it is stable in the intestines and is capable of binding to the
intestinal epithelium, including the M-cells of the Peyer’s patches,
which is important for stimulating mucosal immunity, including local
immunological memory (13). It is believed that this is important for
protection because studies in animals have shown a direct correlation
between protection against cholera toxin-induced fluid secretion and
intestinal synthesis of secretory immunoglobulin A (sIgA) antibodies,
and also between protection and the number of antitoxin-producing
cells in the intestines. Thus, locally produced sIgA antibodies are
considered important for providing antitoxic immmunity in the gut
(14).
There are, however, other cellular components of V. cholerae that
induce potentially protective immune responses. The killed whole
cells themselves elicit an antibacterial response that is directed mainly
against the lipopolysaccharide (LPS) of the pathogen; LPS is the
predominant antigen producing immunity to cholera in an experi-
mental setting (14). There is also evidence to suggest that an immune
response to toxin-coregulated pili (TCP) may also play a role in host
protection. In classical V. cholerae O1 organisms, TCP have been
shown to play an important role in the colonization of the small
intestines (15). These pili are rarely found on the El Tor vibrios,
although an El Tor-specific type of TCP has been reported to be
expressed (16, 17). The El Tor organisms, however, express another
type of pili called mannose-sensitive haemagglutinin (MSHA) fim-
briae; these are poorly expressed on the surface of the classical vibrios
(16). There is no evidence to suggest that the MSHA fimbriae en-
hance the immunogenicity of the killed oral vaccines. However, it has
been proposed that TCP, while not an important antigen in itself, may
enhance immunity by mediating the attachment of the bacteria to the
intestinal cells. The relative importance of TCP and LPS as compo-
nents of inactivated vaccines is unclear.
The growth conditions required for maximum expression of V.
cholerae antigens in the laboratory need to be carefully determined
and may differ significantly from those expected in vivo; for example,
the conditions needed for the production of cholera toxin and TCP G

133

ECB text 133 19/11/2004, 09:49 AM

Black
 (18). Furthermore, studies have shown that V. cholerae, like other
pathogenic bacteria, express a number of antigens during growth in
vivo that are not readily produced by the organism when grown under
various conditions in vitro (19, 20). With the development of sophis-
ticated genomic-based technologies, including in vivo expression
systems to probe host environments, significant new insights into the
complexities of host–pathogen interactions are being gained. These
may lead to better control of the expression in vitro of antigens that
may be important for vaccine production and host protection. Recent
studies using in vivo expression technology have shown cholera toxin
and TCP to be expressed sequentially during infection and that full
toxin expression occurs only after, and is dependent upon, coloniza-
tion (21). There is a possibility that a quorum-dependent signal is
involved in the process. Quorum sensing is a process whereby cell–
cell communications are mediated by the synthesis, secretion and
detection of small extracellular signal molecules (22). Cell density is
likely to play a part in this process.

2.4 Inactivated oral vaccines

Two killed (inactivated) oral cholera vaccines have been developed
and clinically tested. One vaccine, developed in Sweden, consists of
inactivated whole cells of V. cholerae in combination with a purified
recombinant DNA derived B-subunit (rCTB) of the cholera toxin. In
early clinical trials of this vaccine a native B-subunit (CTB) was used.
The second vaccine, developed in Vietnam following technology
transfer from the Swedish manufacturer, consists of whole inactivated
V. cholerae cells alone. Large-scale field trials in Bangladesh and Peru
(3, 23–25) have shown that a whole-cell killed vaccine containing the
B-subunit, and a killed whole-cell preparation alone, both produced
by a Swedish company, conferred significant protection on recipients
for up to 3–5 years depending on age of the vaccines. In the field trial
in Bangladesh, three doses of the vaccine containing the B-subunit
resulted in 85% and 50% protection when assessed after 6 months
and 3 years, respectively, in all age groups, including children aged
less than 5 years. However, protection declined rapidly after the first
6 months of follow-up in children aged 2–5 years and disappeared
during the third year after vaccination. In contrast, the vaccine from
the Swedish manufacturer lacking the B-subunit, that was assessed in
Bangladesh, did not confer significant protection against El Tor chol-
era in young children. In adults, the oral vaccine lacking the B-subunit
gave a somewhat lower initial level of protection than that given by
the vaccine containing the B-subunit, but after 6 months the protec-
G tion afforded by the two vaccines was similar. The protective efficacy

134

ECB text 134 19/11/2004, 09:49 AM

Black
 of the inactivated whole-cell vaccine containing the rCTB was repro-
duced in Peru in military recruits in whom two doses gave 86% short-
term protective efficacy (25). The second vaccine for which clinical
trial results are available was produced in Vietnam. Two oral doses of
this killed whole-cell oral vaccine lacking the B-subunit were reported
to have an efficacy of 66% 8 months after immunization in all age
groups (26). A second-generation bivalent vaccine, containing the
serogroup O139 in addition to O1, but with no B-subunit component,
is being developed and evaluated (3).

2.5 Correlates of protection

A problem in the evaluation of cholera vaccines is the identification of
appropriate markers of protection. Oral vaccination promotes anti-
LPS secretory IgA responses similar to those for infection itself (14,
27) whereas parenteral immunization does not. Similarly, the B-
subunit of whereas cholera toxin also elicits high antitoxin secretory
IgA responses when given orally (14, 28). To be efficacious, cholera
vaccine must stimulate a local immune response in the gut mucosa.
Intestinal biopsies have shown that there is an increase in antibody-
secreting cells specific to the B-subunit of cholera toxin and to whole
cells following oral immunization (29). However, serum vibriocidal
antibodies may offer an indirect measure of the protective immune
response. Vibriocidal antibodies are measured by the degree of bacte-
rial lysis that occurs when serial dilutions of serum are incubated with
a large standardized inoculum of V. cholerae in the presence of
complement. Following natural infection of humans, there is a many-
fold rise in titre of serum vibriocidal antibodies. Elevated titres of
serum antibodies are correlated with protection if immunization was
by the oral route (30, 31). The killed whole-cell parenteral vaccine is
also capable of eliciting a high vibriocidal titre in immunized individu-
als, but this vaccine confers only limited protection for a short time.
Vibriocidal titre must therefore be seen only as a marker of the
stimulation of an appropriate intestinal immune response and not a
goal in itself. Serum vibriocidal antibody responses that occur follow-
ing the ingestion of live oral antigens, delivered by wild type or
attenuated V. cholerae have been shown to serve as markers for the
stimulation of a potential intestinal immunity that endures long after
the serum vibriocidal antibody titres have returned to baseline levels
(3, 8). In regions where cholera is endemic, vibriocidal antibody titres
are relatively high before vaccination, and rises in titre following oral
vaccination are modest in comparison with those obtained by vacci-
nating people in non-endemic area. The only direct predictor of pro-
tection to cholera is the local secretory IgA response in the small G

135

ECB text 135 19/11/2004, 09:49 AM

Black
 intestine, which is clearly not a practical indicator to measure in the
context of a large clinical trial. The serum vibriocidal titre is therefore
the most useful marker presently available for indicating an appropri-
ate immune response in humans.

2.6 Production and control of inactivated oral cholera vaccines

The vaccines currently produced typically contain 25–50 ¥ 109 cells
per dose of each of the strains of V. cholerae representing both Inaba
and Ogawa serotypes, as well as classical and El Tor biotypes. Some
formulations also contain inactivated V. cholerae O139 (50 ¥ 109
cells). The vaccine from Sweden also contains 1 mg per dose of
purified rDNA derived B-subunit of the cholera toxin.
The whole-cell components of the vaccines are inactivated individu-
ally, before or after washing, either by treatment with formaldehyde
or by heating. Inactivated bacterial cultures are then harvested by
centrifugation or ultrafiltration, washed, resuspended in buffer and
mixed with the B-subunit of cholera toxin, if used, to produce the final
bulk from which the final lots are produced.
There is no precedent for controlling this new type of vaccine (i.e. an
inactivated killed oral vaccine), and there is as yet no internationally
accepted direct method for measuring the potencies of such products
that guarantees that protective immunity will be elicited in the target
population. At present, there is no animal model that can meaning-
fully be used to measure or predict the potency of these vaccines
in humans. It is not known whether animal potency tests using
parenteral administration of vaccine would be a reliable indicator of
the protective effect of the same vaccine when administered orally.
Additionally, the available evidence on tests using the parenteral
administration of vaccine to rabbits suggests that the immunological
response does not follow a dose–response relationship; in mice
parenteral administration results in a large variability in antibody
titres that would necessitate the use of a large number of animals. For
this reason an animal potency assay has been omitted from these
Guidelines. Research to identify appropriate assays that better pre-
dict protective efficacy in humans is strongly encouraged. Such assays
should be able to detect sub-potent batches of vaccines.
In the light of these difficulties, it is suggested that emphasis should
be placed on the characterization and quantification in vitro of the
critical vaccine antigens and components. The characteristics of the
various antigens and components claimed to contribute to vaccine
efficacy, together with data on vaccine composition and dosage,
G consistency of production, and conformity with specifications, of the

136

ECB text 136 19/11/2004, 09:49 AM

Black
 vaccine used in clinical trials, will give some indication, though not
definitive proof, of the ability of a vaccine lot to elicit protective
immunity. These antigens and components might include LPS, TCP,
which it is suggested could act to enhance the immune response rather
than as an antigen in itself, and, where indicated, the B-subunit of the
cholera toxin. Thus the immunological, biological and biochemical
characterization of the individual components claimed to contribute
to vaccine efficacy is critical for demonstrating their structural and/or
functional integrity in vaccine production lots. Relevant tests should
be performed before any procedure such as detoxification, chemical
or heat treatment (which may modify the immunological or biological
characteristics of the component), is carried out. This would apply to
any component considered to be important to the performance of the
vaccine, but that may not easily be tested for following inactivation.
Other tests, such as that for residual activity of cholera toxin should
be undertaken routinely after detoxification, chemical or heat treat-
ment of vaccine lots, or as part of process validation.

Residual cholera toxin is a possible contaminant of inactivated whole-

cell oral vaccines. Rigorous washing of the culture and inactivation
using heat or formaldehyde treatment are features of the production
process. However, a toxicity test to confirm freedom from toxicity will
be necessary, and acceptable limits of cholera toxin activity should
be set to confirm consistency of manufacture. The amount of active
cholera toxin in a new production lot should not exceed that present
in lots shown to be safe in clinical studies. The mouse weight-gain test
currently in use to monitor the toxicity of vaccine lots is considered to
be insufficiently sensitive and of questionable relevance. A more
relevant and validated test should be sought. The potential use of the
Y-1 adrenal cell assay for cholera toxin as a more specific test for
residual toxicity should be investigated. Such a specific test could be
used on a-lot-to-lot basis or to validate the production process.

Should the use of vaccine involve administration in extra buffer to

protect against acid conditions in the stomach (as for the vaccine
containing the B-subunit) the buffer should be similar to that used in
the clinical studies and compatible with the vaccine.

The need for a preservative in multidose presentations of an oral

vaccine should be carefully evaluated and consideration given to the
use of a non-mercury-based preservative should one be thought nec-
essary. If no preservative is added to multidose containers a time-limit
of a maximum of 6 hours should be imposed on the storage of opened
containers. G

137

ECB text 137 19/11/2004, 09:49 AM

Black
 3. Manufacturing recommendations
These Guidelines apply to the production and control of liquid formu-
lations of inactivated cholera vaccine intended for oral administra-
tion. The Guidelines emphasize the importance of in-process controls
for biologicals and cover the following three areas:
— the starting materials;
— the manufacturing process; and
— the final product.
The general manufacturing recommendations contained in good
manufacturing practices for pharmaceutical (35) and biological prod-
ucts (36) should be applied at establishments manufacturing inacti-
vated oral cholera vaccine.
Production and control of the rDNA-derived B-subunit using a
genetically modified strain of V. cholerae should be according to the
guidelines for assuring the quality of pharmaceutical and biological
products prepared by recombinant DNA technology (32) and other
relevant recommendations (33, 34). The same guidelines would apply
equally to the production of rCTB in any other host organism, such as
Escherichia coli.
V. cholerae is a class 2 pathogen and represents a particular hazard to
health through infection by the oral route. It should be handled under
appropriate conditions for this class of organism (37). Standard oper-
ating procedures need to be developed for dealing with emergencies
arising from the accidental spillage, leakage or other dissemination of
cholera organisms. Personnel employed in the production and control
facilities should be adequately trained. Appropriate protective
measures including vaccination should be implemented. Adherence
to current good manufacturing practice and appropriate biosafety
measures are important to the integrity of the product, to protect
workers and to protect the environment.
Details of standard operating procedures for the preparation and
testing of inactivated oral cholera vaccines adopted by a manufac-
turer, together with evidence of appropriate validation of each pro-
duction step, should be submitted for approval to the national
regulatory authority. All assay procedures used for quality control of
the vaccine and vaccine intermediates should also be validated (38).
Proposals for modifications of the manufacturing process or control
methods should be submitted for approval to the national regulatory
authority before they are implemented.
The general recommendations for control laboratories contained in
G the guidelines for national regulatory authorities on quality assurance

138

ECB text 138 19/11/2004, 09:49 AM

Black
 for biological products (39) should be applied. A vaccine lot should be
released using a batch release procedure and only if it fulfils national
requirements.

3.1 Control of starting materials

3.1.1 Strains of V. cholerae
The current vaccines consist of classical and El Tor biotypes of Inaba
and Ogawa serotype and, in some cases, the O139 serotype may be
included. The strains used should have the appropriate morphologi-
cal, cultural, biochemical, serological and other properties appro-
priate to the strain. A strain of V. cholerae that has been genetically
modified to delete cholera toxin A-subunit genes is currently used to
produce the rDNA derived B-subunit when this is included in the
vaccine.

3.1.2 Seed-bank system

The production of V. cholerae, including strains containing the plas-
mid encoding the recombinant B-subunit should be based on a master
and working seed lot system. Cultures derived from the working seed
lot should have the same characteristics as the cultures of the strain
from which the master seed lot was derived. If materials of animal
origin are used in the medium for seed production, preservation of
strain viability for freeze-drying, or for frozen storage, they should
comply with the guidance given in the report of a WHO consultation
on medical and other products in relation to human and animal trans-
missible spongiform encephalopathies (40) and should be approved
by the national control authorities.

3.1.3 Culture media for growth of organisms

Where possible, materials of non-animal origin should be used. If
materials of animal origin are used, they should comply with the
guidance given in the report of a WHO consultation on medical and
other products in relation to human and animal transmissible
spongiform encephalopathies (40) and should be approved by the
national regulatory authorities. Human blood or reagents derived
from human blood must not be used in either the culture media used
for the production of seed banks or of vaccine. If human albumin is
used in any part of the production process, it should meet the require-
ments for the collection, processing and quality control of blood,
blood components and plasma derivatives (41) as well as current
guidelines in relation to human transmissible encephalopathies. G

139

ECB text 139 19/11/2004, 09:49 AM

Black
 3.2 Control of the manufacturing process
3.2.1 Control of production cultures
Production cultures should be shown to be consistent in respect of
growth rate, pH and yield of cells or cell products. Acceptance speci-
fications should be established.
Cultures should be checked at different stages of fermentation for
purity, identity and cell density. Unsatisfactory cultures must be dis-
carded. Where a plasmid-containing strain (see section 3.2.2) is used
for the production of the recombinant B-subunit, the cultures should
be checked for the presence and identity of appropriate genetic
markers. Numbers of plasmid copies should be checked routinely at
lot release or confirmed during process validation.
At the time of harvest and prior to detoxification, whole cell bulks
should be checked for purity, identity, opacity, pH and relevant bio-
chemical and antigenic characteristics. For assessing purity, samples
of the culture should be examined by microscopy of Gram-stained
smears, by inoculation of appropriate culture media or by another
suitable procedure.
Following killing by heat or formaldehyde treatment, the cultures
should be checked for viability, purity, opacity, identity and pH. The
inactivation process may affect cell morphology or integrity, and
opacity measurements may not be a reliable indicator of bacterial
numbers. Assays for specific antigen content should be used to deter-
mine the concentrations of the monovalent bulks used for formulat-
ing vaccines based on killed cells only. Assays for each specific LPS
should be employed.

3.2.2 Control of production of purified rDNA derived B-subunit

3.2.2.1 Strategy for cloning and expressing the gene
A full description of the host cell and expression vectors used in
production should be given. This should include:
— the source, genetic characteristics and details of maintenance of
the host strain or strains;
— the construction, genetics and structure of the expression vector;
— the origin and identification of the gene that is being cloned.
The cultural conditions used to promote and control the expression of
the cloned gene in the host cell should be described in detail. Agents
known to provoke sensitivity reactions in certain individuals, such as
penicillin or other beta-lactam antibiotics, should not be used in the
G fermentation process.

140

ECB text 140 19/11/2004, 09:49 AM

Black
 The stability of the expression system during storage and beyond the
passage level used in production should be documented and specifica-
tions set for plasmid retention during storage of seed and during
production. The stability of the host-vector system should either be
confirmed during process validation or checked routinely at the end
of fermentation. Unstable systems should not be used. The expression
system should be approved by the national regulatory authority.

3.2.2.2 Characterization of the recombinant vector

The nucleotide sequence of the gene insert and of adjacent flanking
segments of the vector, together with restriction enzyme mapping
and/or full sequencing of the vector containing the gene insert should
be provided to the national regulatory authority.

3.2.2.3 Purification procedures

The methods used to purify the rDNA B-subunit from culture har-
vests should be described in detail; the capacity of each stage of the
purification procedure to remove or inactivate substances other than
the B-subunit should also be determined. In particular, the capacity of
the purification process to assure the absence of significant quantities
of any holotoxin or other V. cholerae toxins, such as zonula occludens
toxin or accessory cholera enterotoxin, should be assessed, unless it
has been demonstrated that the cloning and expression procedures
eliminate all possibility of production of such factors. Limits should
be established for the quantities of impurities detected in the purified
B-subunit preparation and these impurities should be identified and
characterized as appropriate.

3.2.3 Characterization of rDNA derived B-subunit

Rigorous characterization of the rDNA derived B-subunit product
should be undertaken using a variety of analytical techniques exploit-
ing several different properties of the molecule, including size, charge
and amino acid composition. Techniques suitable for such purposes
include SDS-polyacrylamide gel electrophoresis (SDS-PAGE), size-
exclusion and reverse-phase chromatography. Sufficient sequence
information should be obtained by direct sequencing and by pe-
ptide mapping, or another appropriate molecular technique, for ex-
ample, mass spectrometry, in comparison with the natural material.
The identity of the product should be confirmed by at least partial
N-terminal and C-terminal amino acid sequencing. Several lots of
the product should be as fully characterized as possible. Several ap-
propriate methods should then be selected for use in routine lot
release. G

141

ECB text 141 19/11/2004, 09:49 AM

Black
 Data should be provided on the consistency of yield in terms of both
quantity and quality of product for sequential production runs. The
effects of freeze-drying should also be investigated.
The rDNA derived B-subunit should be shown to elicit antibody
responses in humans, with the antibodies shown to be functional (e.g.
toxin-neutralizing) in a suitable assay.

3.3 Control of final bulk

3.3.1 Preparation
For vaccine formulated from killed cells only, the final bulk is pre-
pared by mixing suitable quantities of each monovalent bulk sus-
pended in the appropriate buffer. For vaccines containing the rDNA
B-subunit, this component is dissolved in buffer to an appropriate
concentration and then mixed with the cell suspension final bulk to
achieve a mixture containing each component at the required concen-
tration. Preservative, if used, may be added either to individual
monovalent bulks or at the final bulk stage.

3.3.2 Antigen content

The concentration of each specific antigen (i.e. total O1 or O139 LPS,
TCP as appropriate) that is considered to play a part in protection
should be assayed in the final bulk by a suitable immunoassay ap-
proved by the national regulatory authority. Similarly, for formula-
tions containing the B-subunit, its concentration in the final bulk
should be assayed by an approved method, for example, single radial
diffusion. The final concentration of each active component should be
within limits that are consistent with those of lots shown to be safe and
efficacious in clinical trials.

3.3.3 Detoxifying agents

If formaldehyde or another detoxifying agent is used in the pre-
paration of killed cells, its residual concentration should be deter-
mined in the final bulk by a method approved by the national
regulatory authority. The final concentration should not exceed the
limits established for clinical trial lots that have been shown to be safe
and efficacious.

3.3.4 Sterility
Each final bulk should be tested for bacterial and fungal sterility in
accordance with the requirements of Part A, sections 5.1 and 5.2 of
the revised requirement for biological substances (41) or by a method
G approved by the national regulatory authority. If a preservative has

142

ECB text 142 19/11/2004, 09:49 AM

Black
 been added to the product, appropriate measures should be taken to
prevent it from interfering with the test.

3.3.5 Preservative
If a preservative has been added, its concentration may be deter-
mined at the bulk stage by a method approved by the national regu-
latory authority. The preservative, its concentration and its limits
should be approved by the national regulatory authority.

3.3.6 Potency/immunogenicity
At present there is no animal potency or immunogenicity assay that
can be recommended for use as a reliable indicator of the protective
efficacy of inactivated oral cholera vaccines in humans or for the
detection of sub-potent batches (see section 2.6).

3.3.7 Residual toxin activity

Cholera toxin should be assayed by a method approved by the na-
tional regulatory authority. Alternatively, the production process
should be validated to show that the quantities of clinically active
cholera toxin present in the product are insignificant. The inactivation
process should also be validated to assure the absence of significant
quantities of holotoxin or other V. cholerae toxins.

3.4 Control of final lot

The following tests should be performed on each final lot of vaccine
(i.e. in the final containers).

3.4.1 Appearance
The final containers should be inspected visually (manually or with
automatic inspection systems). After shaking, the vaccine should
form a uniform, turbid, white or brownish suspension free of aggre-
gates and extraneous particles. Containers showing abnormalities
must be discarded.

3.4.2 Identity
An identity test should be performed on at least one labelled con-
tainer from each final lot. The test used should identify the type of
vaccine formulated. For preparations formulated from killed cells
alone, a serological test that detects V. cholerae O1 and O139 (if
present) antigens will suffice. For preparations formulated from killed
cells and rDNA B-subunit, the identity test must be able to detect the
presence of both types of component. The procedures used should be G

143

ECB text 143 19/11/2004, 09:49 AM

Black
 approved by the national regulatory authority. The antigen-content
assays (see below) could also serve as an identity test.

3.4.3 Antigen content

The concentration of each specific antigen (i.e. total O1 or O139 LPS,
TCP as appropriate), that is considered to play a part in protection,
should be assayed by a suitable immunoassay approved by the
national regulatory authority. Similarly, for formulations containing
the B-subunit, its concentration should be assayed by an approved
method, for example, single radial diffusion. The final concentration
of each active component should be within limits that are consistent
with those of lots shown to be safe and efficacious in clinical trials.

3.4.4 Sterility
Each final lot should be tested for bacterial and fungal sterility as
indicated in section 3.3.4.

3.4.5 Preservative content

If a preservative is included, each final lot should be assayed for
preservative content unless this was done on the final bulk. The assay
method used and the preservative content permitted should be
approved by the national regulatory authority.

3.4.6 pH
The pH should be tested and shown to be within the range of values
found suitable for vaccine lots that have been shown to be safe and
effective in clinical trials and in stability studies.

3.4.7 General safety (innocuity)

No such test is recommended for an oral preparation.

3.5 Stability, storage and expiry date

The stability of the vaccine in its final container, when maintained at
the recommended temperature, should be established using real-time
studies. These should be conducted on at least three consecutive final
lots, derived from separate antigen-production lots.
The content of V. cholerae LPS and other specified antigens should
remain within specified limits for the duration of the shelf-life. If the
formulation contains the B-subunit, its content must also remain
within specified limits for the duration of the shelf-life. Accelerated
stability studies at elevated temperatures may provide additional
G evidence of vaccine stability, but cannot replace real-time studies.

144

ECB text 144 19/11/2004, 09:49 AM

Black
 When any changes that may affect the stability of the product are
made in the production process, the stability of the vaccine produced
by the new procedure should be demonstrated by additional studies.
If monovalent bulks or final bulk products are to be stored, stability
studies should be performed and an appropriate shelf-life assigned on
the basis of the data obtained.

3.6 Reference materials

No formally established international reference materials are cur-
rently available for the standardization of oral cholera vaccines, but
their development is under consideration. Manufacturers should set
aside, as reference material, a vaccine lot identical with, or demon-
strated to be equivalent to, a lot shown to give acceptable perfor-
mance in clinical trials. It is recommended that the reference lot
should be stabilized by a validated procedure, such as freeze-drying,
to maintain stability over a long period.
Other reference materials should include a stabilized preparation of
the rDNA B-subunit and holotoxin.
Manufacturers and national regulatory authorities, should establish
reference antisera against O1 and O139 LPS antigens and,
monospecific antisera or monoclonal antibodies to Inaba, Ogawa
epitopes and B-subunit.

Authors
These Guidelines were drafted by Dr M. Corbel, Division of Bacteri-
ology, National Institute of Biological Standards and Control, Potters
Bar, Herts., England; Dr D. Garcia, Quality Control of Immunologi-
cal Products, Agence Francaise de Sécurité Sanitaire des Produits de
Santé, Lyon, France, and Dr E. Griffiths, Quality Assurance
and Safety of Biologicals, World Health Organization, Geneva,
Switzerland.

Acknowledgements
Acknowledgements are due to the following experts for their comments and advice
on the issues in the standardization and control of oral cholera vaccines which
were first discussed by a WHO Working Group at a meeting held 10–11 May 1999,
in Geneva: Dr I. Feavers, National Institute for Biological Standards and Control,
Potters Bar, Herts., England; Dr D. Garcia, Agence Francaise de Sécurité Sanitaire
des Produits de Santé, Lyon, France; Dr J. Holmgren, Department of Medical G

145

ECB text 145 19/11/2004, 09:49 AM

Black
 Microbiology and Immunology, Göteborg, Sweden; Dr Huynh Anh Hong, National
Centre for Vaccines and Biologicals, Nha Trang, Viet Nam; Dr D. Kopecko, Center
for Biologics Evaluation and Research, Food and Drug Administration, Bethesda,
MD, USA; Dr Chung Keel Lee, International Vaccine Institute, Seoul, Republic of
Korea; Dr M.M. Levine, University of Maryland, Baltimore, MD, USA; Dr V.
Oeppling, Paul Ehrlich Institute, Langen, Germany; Dr S. Rijpkema, National
Institue for Biological Standards and Control, Potters Bar, Herts., England; Dr J.
Stadler, Federal Office of Public Health, Berne, Switzerland; Dr A.M. Svennerholm,
Department of Medical Microbiology and Immunology, Göteborg, Sweden; Dr R.
Winsnes, Norwegian Medicines Control Authority, Oslo, Norway; Dr P. Askeloff,
SBL Vaccin, Stockholm, Sweden; Dr U. Bjare, SBL Vaccin, Stockholm, Sweden; Dr
E. Fürer, Swiss Serum and Vaccine Institute, Berne, Switzerland; Dr J. Que, Swiss
Serum and Vaccine Institute, Berne, Switzerland; Dr M. Schroeder, Swiss Serum
and Vaccine Institute, Berne, Switzerland and Dr J-F Viret, Swiss Serum and
Vaccine Institute, Berne, Switzerland. Dr N. Dellepiane, Access to Technologies;
Dr E. Griffiths, Coordinator, Quality Assurance and Safety of Biologicals; Dr B.
Ivanoff, Vaccine Development; Dr L. Kuppens, Communicable Diseases
Surveillance and Response; Dr M. Neira, Director Communicable Diseases
Prevention, Control and Eradication.
A draft of these guidelines was reviewed at a WHO informal consultation held in
October 2001 at the International Vaccine Institute, Seoul, Republic of Korea.
Participants included: Dr Sang-Ja Ban, Division of Bacterial Products, Korea Food
and Drug Administration, Seoul, Republic of Korea; Dr Le Van Be, National Centre
for Vaccines and Biologicals, Nha Trang, Viet Nam; Dr N. Carlin, SBL Vaccin,
Stockholm, Sweden; Dr D. Garcia, Agence francaise de Sécurité Sanitaire des
Produits de Santé, Lyon, France; Dr E. Griffiths, World Health Organization,
Geneva, Switzerland; Dr M. Haase, Paul Ehrlich Institute, Langen, Germany; Dr
Hasbullah, Bio Farma, Bandung, Indonesia; Dr Lei Dianliang, National Institute for
the Control of Pharmaceutical and Biological Products, Beijing, China; Dr L.
Slamet, National Agency of Drug and Food Control, Jakarta, Indonesia; Dr Doan
Thi Tam, National Centre for Vaccines and Biologicals Control, Hanoi, Viet Nam; Dr
Nguyen Thu Van, Vabiotech, National Institute of Hygiene and Epidemiology,
Hanoi, Viet Nam; Mr M. Welin, Medical Products Agency, Uppsala, Sweden; Dr R.
Winsnes, Norwegian Medicines Control Authority, Oslo, Norway; Dr J. Clemens,
International Vaccine Institute, Seoul, Republic of Korea and Dr C.K. Lee,
International Vaccine Institute, Seoul, Republic of Korea.

References
1. Requirements for cholera vaccine. In: WHO Expert Committee on Biological
Standardization. Twenty-first report. Geneva, World Health Organization,
1969, Annex 1 (WHO Technical Report Series, No. 413).
2. Requirements for cholera vaccine. Addendum. In: WHO Expert Committee
on Biological Standardization. Twenty-fifth report. Geneva, World Health
Organization, 1973 (WHO Technical Report Series, No. 530).
3. Cholera vaccines. Weekly Epidemiological Record, 2001, 76:117–124.
4. WHO Expert Committee on Biological Standardization, Fiftieth report.
Geneva, World Health Organization, 2002 (WHO Technical Report Series,
No. 904).
5. International Centre for Diarrhoeal Disease Research, Bangladesh, Cholera
Working Group. Large epidemic of cholera like disease in Bangladesh
G caused by Vibrio cholerae O139. Lancet, 1993, 342:387–390.

146

ECB text 146 19/11/2004, 09:49 AM

Black
 6. Clemens JD et al. Biotype as determinant of natural immunizing effect of
cholera. Lancet, 1991, 337:883–884.
7. Glass RI et al. Endemic cholera in rural Bangladesh. American Journal of
Epidemiology, 1982, 116:959–970.
8. Levine MM et al. Duration of infection derived immunity to cholera. Journal
of Infectious Diseases 1981, 143:818–820.
9. Jertborn M, Svennerholm A-M, Holmgren J. Saliva, breast milk and serum
antibody responses as indirect measures of intestinal immunity after oral
cholera vaccination or natural disease. Journal of Clinical Microbiology,
1986, 24:203–209.
10. Clemens JD et al. Cross-protection by B subunit-whole cell cholera vaccine
against diarrhoea associated with heat labile toxin-producing
enterotoxigenic Escherichia coli: results of a large scale field trial. Journal of
Infectious Diseases, 1988, 158:372–377.
11. Svennerholm A-M, Holmgren J. Vaccines against enterotoxigenic
Escherichia coli. In: Sussman M, ed. Escherichia coli: mechanisms of
virulence. Cambridge, Cambridge University Press 1997: 605–626.
12. Peltola H et al. Prevention of travellers’ diarrhoea by oral B-subunit/whole
cell cholera vaccine. Lancet, 1991, 338:1258–1289.
13. Neutra MR, Kraehenbühl JP. Transepithelial transport and mucosal defence.
I: the role of the M cell. Trends in Cell Biology, 1992, 2:134–138.
14. Svennerholm, A-M, Holmgren J. Oral B-subunit whole cell vaccines against
cholera and enterotoxigenic Escherichia coli diarrhoea. In: Ala’ Aldeen DAA,
Hormaeche CE, eds. Molecular and clinical aspects of bacterial vaccine
development. Chichester John Wiley & Sons, 1995: 205–232.
15. Taylor RK et al. Use of phoA gene fusion to identify a pilus colonization
factor coordinately regulated with cholera toxin. Proceedings of the National
Academy of Sciences (USA), 1987, 84:2833–2837.
16. Jonson G, Holmgren J, Svennerholm A-M. Epitope differences in toxin-
coregulated pili produced by classical and El Tor Vibrio cholerae. Microbial
Pathogenicity, 1991, 11:179–188.
17. Kaufman MR, Taylor RK. The toxin-coregulated pilus: biogenesis and
function. In: Wachsmuth IK, Blake PA, Olsvik Q, eds. Vibrio cholerae and
cholera: Molecular to global perspectives. Washington, DC, American
Society for Microbiology, 1994:187–202.
18. Chakrabarti S, Sengupta N, Chowdhury N. Role of DnaK in in vitro and in
vivo expression of virulence factors of Vibrio cholerae. Infection and
Immunity, 1999, 67:1025–1033.
19. Jonson G, Holmgren J, Svennerholm A-M. Analysis of expression of toxin-
coregulated pili in classical and El Tor Vibrio cholerae O1 in vitro and in
vivo. Infection and Immunity, 1992, 60:4278–4284.
20. Griffiths E. Environmental regulation of bacterial virulence — implications for
vaccine design and production. Trends in Biotechnology, 1991, 9:309–315.
21. Lee SL et al. Regulation and temporal expression patterns of Vibrio
cholerae virulence genes during infection. Cell, 1999, 99:625–634. G

147

ECB text 147 19/11/2004, 09:49 AM

Black
 22. Surette MG, Bassler BL. Quorum sensing in Escherichia coli and Salmonella
typhimurium. Proceedings of the National Academy of Sciences (USA),
1998, 95:7046–7050.
23. Holmgren J et al. New cholera vaccines. Vaccine, 1989, 7:94–96.
24. Sach DA et al. Antibody responses following immunization with killed oral
cholera vaccines during the 1985 vaccine field trial in Bangladesh. Journal
of Infectious Diseases, 1991, 164:407–411.
25. Sanchez JL et al. Protective efficacy of the oral whole cell/recombinant B
subunit cholera vaccine in Peruvian military recruits. Lancet, 1994,
344:1273–1276.
26. Trach DD et al. Field trial of a locally produced, killed oral cholera vaccine
in Vietnam. Lancet, 1997, 349:231–235.
27. Svennerholm A-M et al. Mucosal antitoxic and antibacterial immunity after
cholera disease and after immunization with a combined B subunit — whole
cell vaccine. Journal of Infectious Diseases, 1984, 149:884–893.
28. Clemens JD et al. Field trials of cholera vaccines in Bangladesh: results
from a three year follow-up. Lancet, 1990, 335:270–273.
29. Quiding M et al. Oral cholera vaccination induces strong intestinal antibody
responses, gamma-interferon production and involves local immunological
memory. Journal of Clinical Investigation, 1991, 88:143–148.
30. Mosley WH et al. The relationship of vibriocidal antibody titre to
susceptibility of cholera in family contacts of cholera patients. Bulletin of the
World Health Organization, 1968, 38:777–785.
31. Clemens JD et al. Field trials of oral cholera vaccines in Bangladesh: serum
vibriocidal and antitoxic antibodies as markers of the risk of cholera. Journal
of Infectious Diseases, 1991, 163:1235–1242.
32. Guidelines for assuring the quality of pharmaceutical and biological
products prepared by recombinant DNA technology. In: WHO Expert
Committee on Biological Standardization, Forty-first report. Geneva, World
Health Organization, 1991, Annex 3 (WHO Technical Report Series,
No. 814).
33. Points to consider in the production and testing of new drugs and
biologicals produced by recombinant DNA technology. Washington, DC, US
Food and Drug Administration 1985.
34. EC Committee on Proprietary Medicinal Products: Ad hoc Working Party on
Biotechnology/Pharmacy. Notes to applicants for marketing authorizations
on the production and quality control of medicinal products derived by
recombinant DNA technology. Journal of Biological Standardization, 1989,
17:223–231.
35. Good manufacturing practices for pharmaceutical products. In: WHO Expert
Committee on Specifications for Pharmaceutical Preparations. Thirty-second
report. Geneva, World Health Organization, 1992, Annex 1 (WHO Technical
Report Series, No. 832).
36. Good manufacturing practices for biological products. In: WHO Expert
G Committee on Biological Standardization. Forty-second report. Geneva,

148

ECB text 148 19/11/2004, 09:49 AM

Black
 World Health Organization, 1992, Annex 1 (WHO Technical Report Series,
No. 822).
37. Biosafety guidelines for personnel engaged in the production of vaccines
and biological products for medicinal use. Geneva, World Health
Organization, 1995 (document WHO/BVI/95.5).
38. Validation of analytical procedures used in the examination of
pharmaceutical materials. In: WHO Expert Committee on Specifications for
Pharmaceutical materials. Thirty-second report. Geneva, World Health
Organization, 1993, Annex 5 (WHO Technical Report Series, No. 823).
39. Guidelines for national authorities on quality assurance for biological
products. In: WHO Expert Committee on Biological Standardization. Forty-
second report. Geneva, World Health Organization, 1992, Annex 2 (WHO
Technical Report Series, No. 822).
40. Guidelines on transmissible spongiform encephalopathies in relation to
biological and pharmaceutical products. Geneva, World Health
Organization, 2003 (document WHO/BCT/QSD/2003.1)
41. Requirements for the Collection, Processing and Quality Control of Blood,
Blood Components and Plasma Derivatives. In: WHO Expert Committee on
Biological Standardization. Forty-third report. Geneva, World Health
Organization, 1994, Annex2 (WHO Technical Report Series, No. 840).

149

ECB text 149 19/11/2004, 09:49 AM

Black
 G

ECBPR 18/11/2004, 07:50 PM

Black
 G

ECBPR 18/11/2004, 07:50 PM

Black
 G

ECBPR 18/11/2004, 07:50 PM

Black