|dc.description.abstract||Conventional pertussis vaccine prepared from killed whole cell B. pertussis organisms has
been in widespread use since the early 1950's. Despite marked reductions in the incidence of
pertussis, the use of the vaccine has caused concern because of questions of significant
Whooping cough is not notifiable in South Africa, and there is consequently a paucity of hard
data on efficacy; in addition few cases are proven. Incidence, prevalence, severity and
transmission of the disease hence remain a matter of conjecture.
In order to provide background information and determine baseline data for undertaking further
studies, available clinical and epidemiological data on whooping cough (pertussis) in South
Africa was collated. It was intended to compare the pattern of disease seen in this country with
that known in other parts of the world.
Clinical and epidemiological findings from 1525 whooping cough admissions (diagnosed on
the basis of clinical criteria) obtained from 6 major infectious disease hospitals around the
country for the period 1980-1987/1988 are reported. The data from Durban were available in
some detail. Incidence, morbidity and mortality of the disease in South Africa, as elsewhere,
was higher in infancy. Infants and young children were predominantly affected, with 31.3% of
cases under 12 months of age and 7.2% less than 2 months. Mortality was disproportionately
higher in infancy; 53 .2% of deaths were in those younger than 12 months of age. There was a
slight female preponderance, both in respect of prevalence and mortality. Patients were
admitted with pertussis throughout the year, although there was a peak during the winter
months (May to October). The typical whoop accompanied the cough in 55.6% of patients . A
raised white cell count was recorded in 66% of patients. The most commonly detected
The average duration of hospital stay was between 1 0-13 days. Of those with vaccination
records, 26% were unvaccinated, 44% had 1 or 2 doses, and 27% had been fully vaccinated
with a whole-cell pertussis vaccine combined with diphtheria and tetanus toxoids. The picture,
incomplete though it is, reveals a pattern of pertussis similar to that described in other
The study reveals huge gaps in our knowledge of this subject in South Africa. More research
needs to be done, particularly with respect to improved diagnosis, prevention and treatment;
further pertussis should be made a notifiable disease in South Africa.
The whole-cell pertussis vaccine currently used in South Africa has not been adequately
evaluated for post-vaccination events and immunogenicity. The development of new purified
component pertussis vaccines, which appear to be safer than conventional whole-cell
preparations and of equal or almost equal efficacy (although optimal vaccine composition
remains to be defined), requires that the concept of early vaccination with this vaccine
compared with conventional whole-cell vaccine be examined in order to optimise the immune
response to these vaccines in infants; more especially since neonates do not appear to benefit
from passive immunity.
Acellular pertussis vaccine has not been evaluated previously in neonates. In order to address
the problem of high morbidity and mortality from pertussis in early infancy; and the
incorporation of the vaccine into routine vaccination schedules, a phase 11 trial of acellular and
whole-cell vaccines was undertaken in very young infants. The effect of neonatal vaccination
with acellular pertussis vaccine on subsequent immunity; and the immunogenicity and shortterm
safety and reactogenicity of routine primary vaccination with acellular vaccine compared
with conventional whole-cell preparations was investigated.
Three hundred and forty-five healthy, full-term African babies were, enrolled in the study at birth;
58% of whom were successfully followed for 9 months. Infants were assigned to 1 of 3 vaccine
groups in sequence at birth and received either acellular or whole-cell pertussis vaccine ,
combined with 0 and T (A-OTP or W-OTP) at 2, 4 and 6 months of age. Groups I and 11 received
A-OTP and Group III W-OTP. In addition, at birth, Group I received an additional dose of A-OTP
and Group 11, a saline placebo injection. No unvaccinated controls were studied for ethical
Serologic IgG responses to 3 major protective antigens of B. pertussis, filamentous
haemagglutinin (FHA), pertussis toxin (PT) and fimbrial agglutinogens 2,3 (AGG2,3), were
measured by ELlSA in sera obtained at birth, and before vaccination at 2, 4 and 6 months and at
9 months of age. The incidence and nature of post-vaccination events were recorded for 14
days after each dose.
A-OTP induced serum IgG responses to PT and FHA comparable with those reported in other
studies, with peak PT titres occurring at 6 months of age after 2 doses in babies vaccinated
according to the routine schedule (Group 11). Surprisingly, response to W-OTP was found
merely to restore levels of antibody to PT and FHA to those found in cord blood after 3 doses of
vaccine, which questions the immunogenicity of the South African product. Four doses of AOTP
(Group I) did not produce a better antibody response than the 3-dose schedule.
Incidence of all post-vaccination events to both acellular and whole-cell vaccines was low
(85.96/1000 doses, 136.36/1000 doses and 76.30/1000 doses in Groups I, 11 and III
respectively). Minor symptoms were more common in recipients of A-OTP, although no
significant differences in rates were demonstrated. Neurologic post-vaccination events
(without sequelae) occurred more frequently hi recipients of W-OTP. No infant vaccinated with
A-OTP from 2 months of age (Group 11) experienced a neurologic symptom. The risks of major
post-vaccination events cannot however be fully quantified In a study of this small size and
A-DTP vaccination commencing at birth produced final antibody titres to PT and FHA which
were superior to those of South African whole-cell vaccine but were considerably lower than
when the vaccine was incorporated into routine schedules commencing at 2 months of age.
The study findings suggest that acellular pertussis vaccines, whether given from birth or from
the age of 2 months, appeared safer and produced a better serologic response than the South
African whole-cell product which may have impaired immunogenicity.
During the course of the above study, 11 full-term infants with pertussis infection (10
subclinical) were retrospectively diagnosed on the basis of ser~logic evidence. Of the infants
with subclinical disease, all 10 had a ~ 4-fold rise in at least 2 different pertussis IgG antibodies
and nine had ~ 4-fold rise in all 3 IgG antibodies measured. Seven infants had raised IgA
antibodies to PT and FHA and 4 infants had raised IgA antibodies to AGG2,3. Subclinical
infection provoked antibody production to multiple antigens to differing degrees.
The role of various factors which may have contributed to asymptomatic infection were
analysed, viz - household contacts; type of antibody response (clinical vs. subclinical;
vaccinated vs. subclinical); maternally acquired antibody levels; vaccination status (number of
vaccine doses received); age and gender; and nutritional status.
SpeCial features of the study which require emphasis are: pertussis remained subclinical in
unvaccinated babies (most of the subjects were unvaccinated). Subclinical infection followed
incomplete primary vaccination with either acellular or whole-cell vaccines commencing at 2
months of age. Subjects were 8 months of age or younger and only 1 had completed primary
vaccination. Other infections of infancy were commonly detected; 4 infants had upper
Likelihood of subclinical infection was related to significantly lower levels of maternallyacquired
pertussis IgG-AGG2,3 antibodies but not associated with infants' nutritional status.
Subclinical pertussis is described in very young African babies at an age when the disease is
most severe, and therefore has implications for infant morbidity and mortality; it is also
relevant to disease surveillance and vaccine-efficacy studies .
Some perinatal factors influencing vaccination were also explored . In this context we looked
i. The acquisition of maternal antibodies to B. pertussis in paired mother-infant sera from
both well -nourished and SGA full-term and pre-term infants, and infants who
subsequently developed pertussis infection, and effect of these maternally-acquired
antibodies on subsequent antibody responses to acellular pertussis vaccine
administered soon after birth, and to acellular and whole-cell vaccines administered
from the age of 2 months.
if. The acquisition of maternal antibodies to diphtheria and tetanus (OT) in paired motherinfant
sera from full-term and pre-term infants, and the effect of these maternallyacquired
antibodies on serologic responses to neonatal OT vaccination followed by
whole-cell OTP vaccination at 2, 4 and 6 months.
Maternal antibodies were measured since little is known about materno-fetal transfer of
pertussis antibodies, especially in African countries where inhibition of placental transfer might
occur for a variety of reasons. Furthermore, the effect of peri-natal malnutrition and
prematurity on transplacental transfer of diphtheria, tetanus and pertussis antibodies has not
been conclusively established in these areas.
sera with levels in the latter frequently being higher, Indicating active transplacental transfer of
The significant pertussis antibody levels in maternal sera were unlikely to be due to the
currently used South African whole-cell vaccine (given the poor antibody response to PT and
FHA shown in this study). It is assumed that the presence of these antibodies are the end result
of natural infection and therefore that pertussis is widespread in the African community.
Maternal and cord IgG-PT and AGG2,3 titres were significantly lower (p <0.05) and maternal
IgG-FHA marginally lower (p ... O.05) in SGA infants compared to cohorts, although placental
transfer was equally efficient in both groups~ This study has demonstrated that the high titres of
maternally derived circulating B. pertussis antibodies do not have an inhibitory effect on the
subsequent serologic response to acellular vaccine administered in early infancy (2 months) or
with the first dose given soon after birth. Protective levels of diphtheria and tetanus antibodies
were detected in 100% and 76% of cord sera in pre-term infants and in 85% and 67% of cord
sera in full-term infants.
Although the number of infants studied was too small to make a definite comment, there did not
appear to be 'tolerance' due to neonatal diphtheria-tetanus vaccination of pre-term infants, or
to high levels of maternally-acquired antibody.
During analysis of the data from the phase 11 study of acellular pertussis vacclnes, 25 infants
with protein-energy-malnutritlon (PEM) were detected on the basis of anthropometric indices.
Seventeen infants were smaIJ-for-gestatlonal-age (SGA). of whom 9 developed PEM by the age
of 9 months. Eight other infants developed post-natal PEM before the completion of the
primary vaccination course.
and extent of immunological impairment, if any, on the serologic responses to acellular vaccine
and to conventional whole-cell OTP in SGA infants, and in infants who developed PEM before
the completion of the primary vaccination course.
The following indices were evaluated in malnourished infants; (i) anthropometric indices of
nutritional status, (ii) intercurrent illnesses including pertussis infection, (iii) post-vaccination
events, (iv) serologic responses to vaccination. Results were compared with those obtained in
well-nourished (WN) age- and vaccine-matched cohorts.
Overall, peak titres and seroconversion data of all 3 antibodies were not significantly different
in malnourished infants though final anti-AGG2,3 titres (at 9 months of age) in Group III were
significantly lower (p = 0.035).
Although peak and final PT and FHA antibody titres in many SGA and malnourished infants
were lower than in WN infants and peak responses were attained at a later age, malnutrition did
not significantly affect the response to A-OTP. Peak and final AGG2,3 antibody titres were
similar in SGA, malnourished and WN infants. Overall malnourished infants responded no less
well to pertussis vaccination than did other vaccinees.
Incidence of minor local and systemic post-vaccination symptoms was not significantly
different in PEM and WN groups although induration at injection site and irritability were more
frequently reported in the latter. No major neurologic post-vaccination symptoms to either
vaccine were reported in SGA infants or infants with PEM at the time of vaccination. No
significant differences was noted in the incidence of major symptoms in PEM as compared with
One male infant (Group I) who was malnourished at birth and who had been given 2 doses of AOTP,
developed clinical signs of pertussis infection between 2 and 4 months of age. Pertussis
antibody levels immediately prior to infection were not significantly different from those of
un infected age-matched cohorts. The percentage of infants afflicted with common childhood
illnesses were similar in PEM and WN groups (46 vs. 43.2%) although the former group incurred
significantly more illnesses at an earlier age' (s6 months) (p=O.05, chi square).
These findings, albeit preliminary given the small numbers of subjects studied, suggest that
acellular pertussis vaccine may be incorporated into routine vaccination schedules followed in
developing countries with the expectation that adequate antibody responses will be provoked
in SGA infants and in infants who develop post-natal PEM and that the incidence of vaccinerelated
adverse effects will be no higher than in WN infants.
Further and more extensive studies are indicated before the use of acellular pertussis vaccines
can be recommended for routine primary vaccination of infants in preference to whole-cell
preparations in developing countries.||en