Return to July 2002 contents page
Call for intensified research after pneumococcus trial surprises
AS a key component of the plan to accelerate vaccines against
pneumococcus, partners and scientists are calling for research to
be urgently stepped up to provide better data on the burden of the
disease and the efficacy of current candidate vaccines. Their
calls come after trials of the most advanced candidate vaccine
against pneumococcus produced unexpected results, answering some
important questions but raising as many new ones.
Around 2 million children under five years old die from pneumonia
each year in developing countries, according to the latest
estimates (1) . The bacterium
Streptococcus pneumoniae, known commonly as pneumococcus, is
thought to cause 50% or more of these deaths and a similar
proportion of severe cases. A pneumococcal conjugate vaccine is
licensed in the US and Europe and has proven efficacy against
invasive pneumococcal disease (infections of the bloodstream). But
trials are still ongoing to find out whether the vaccine can
prevent pneumonia in children in developing countries: if the
answer is yes, it could sharply reduce child deaths, alongside
vaccines against the other major microbe responsible for
pneumonia, Haemophilus influenzae type b (Hib).
Waiting in line: pneumococcal vaccines could save many lives but
better data are urgently needed on just how many
Assuming that around half of all cases of severe pneumonia are caused by
pneumococcus, and assuming that the vaccine has less than 100% efficacy,
researchers had hoped that the vaccine could reduce the total number of
pneumonia cases by around 30%. Some even hoped that the figure could be
as high as 40%. But at the Third International Symposium on Pneumococci
and Pneumococcal Diseases in Alaska in May, Professor Keith Klugman of
Emory University, Atlanta, presented the results of a trial
(2) from Soweto, South Africa, involving 40,000 children in which the
vaccine reduced total pneumonia cases by around 22%. This figure is
lower than expected and only marginally statistically significant.
However, the trial did confirm that the vaccine, made by Wyeth, reduced
the incidence of invasive pneumococcal disease by more than 80%. Even in
children infected with HIV, for whom invasive pneumococcal disease is a
serious threat, the vaccine halved the incidence. In addition, separate
data from studies in the US, also presented in Alaska, showed that the
vaccine may help to reduce the spread of pneumococcal infections, as
well as protecting those who are vaccinated. When infants in the US are
vaccinated, the number of infections in people aged 20-39 and over 60
also drops suggesting that parents and grandparents benefit.
Klugman is upbeat. If the Soweto findings are borne out elsewhere, then
combined use of pneumococcal vaccine and Hib vaccine could cut the
overall burden of pneumonia in children by some 40%, as well as offering
specific benefits to children with HIV, he says. However, like other
researchers, he believes that the pneumococcal vaccines
lower-than-expected efficacy against pneumonia needs to be better
"These results are forcing researchers to rethink their expectations
about the vaccine, and strongly reinforce the need for continuing with
other efficacy trials," says Dr Orin Levine, of the US National
Institutes of Health, one of a team charged by GAVI with the task of
developing an agenda to rapidly evaluate and introduce pneumococcal
vaccines into developing countries.
At present, researchers are missing key pieces of information. First, it
is still not clear exactly how big the burden of pneumococcal pneumonia
is. Doctors rely on chest x-rays rather than laboratory cultures to
diagnose pneumonia, and chest x-rays cannot distinguish between
pneumonia caused by Hib, pneumonia caused by pneumococcus or other
microbes. The estimate that pneumococcus causes half of all severe cases
of pneumonia is based on a handful of studies from developing countries
where bacteria have been cultured from patients lung fluid or blood,
but those studies may not be representative. One purpose of vaccine
trials is to get a better estimate of the burden.
Another problem is that, to measure a vaccines efficacy, there must be
clearly defined "endpoints" to the trial, such as comparing the number
of cases of pneumonia in those children who have been immunized with the
number of cases in those who have not. However, with x-ray as the main
tool for diagnosing pneumonia, doctors in different settings may
disagree over whether some individual cases should be defined as
pneumonia or not. This may affect the numbers, and so the estimated
efficacy of the vaccine.
The Soweto trial was the first to use standardised criteria for x-ray
confirmed pneumonia, developed by WHO, says Klugman. "It may be that the
criteria need to be reworked," he says.
Given these problems, says Levine, it is difficult to know what the
impact of the vaccine is. The results from Soweto could be interpreted
in a range of different ways, he says. At one extreme, we might
hypothesise that pneumococcus is as big a problem as we expected but we
need better vaccines. At the other extreme, the hypothesis would be that
the vaccines are highly effective, but pneumococcus is not as big a
problem as we thought. Its essential to find out where between these
two extremes the truth lies, says Levine. "As long as vaccines remain
expensive, we are going to have to have very convincing data of their
"This opens up a whole bunch of questions," agrees Professor Kim
Mulholland, a paediatrician specialising in international health at the
University of Melbourne, Australia. First, what will other trials show?
All eyes are now a major trial of the same vaccine in The Gambia, which
is due to end in late 2004. Conditions there may be more representative
of Sub-Saharan Africa as a whole than Soweto. Although children in
Soweto are disadvantaged in many respects, they have better access to
hospital care than in most of rural Africa.
Key research questions
The Soweto trial was not designed to measure the impact of the vaccine
on mortality, so no one knows how many deaths it could prevent. "It is
entirely possible that the vaccine may have a higher impact on the more
severe forms of pneumonia," says Dr Thomas Cherian, of the Christian
Medical College, Vellore, India, who has been at WHO coordinating
pneumococcal vaccine research.
Attention is also turning to other candidates. GlaxoSmithKline has
developed a conjugate vaccine that is intended to protect against
11 different strains, or serotypes, of pneumococcus, compared with
the 9 serotypes in the Wyeth product. Walter Vandersmissen of GSK
told Immunization Focus that, following some technical
delays, the vaccine is now due to start Phase II clinical trials
in Europe and Latin America before the end of the year.
Soweto: despite these tough living conditions, children are more
likely to get hospital treatment here than in most of rural Africa
Another candidate pneumococcal conjugate vaccine was developed by
Aventis Pasteur. The company recently decided to abandon the vaccine in
order to develop a protein pneumococcal vaccine instead (see
"This time, a vaccine for everyone?", Immunization Focus,
March 2002 ), but trials of the conjugate vaccine are still
continuing in the Philippines. Even if the vaccine produces promising
results, there are no known plans to develop it commercially. But
scientists say the results of the trial will still be very important in
providing information about the efficacy of this type of vaccine.
Another important question is to find out whether vaccines against
pneumococcus simply move the goalposts for the microbe. Scientists have
been concerned that, in theory, even if a vaccine protects children
against the serotypes of pneumococcus included in it, other serotypes
may simply take their place and cause disease. The actual findings on
this phenomenon, known as serotype replacement, have been mixed. In most
studies, there is no evidence that it has happened. In Finland, however,
researchers have found that vaccinated children do develop ear
infections with different serotypes but the most severe infections are
still prevented and the number of children needing to have ear tubes
inserted has been reduced.
Levine says it will be important to find out whether the more severe
infections are prevented in pneumonia too, and to monitor carefully for
serotype replacement, in the remaining clinical trials.
Dr Tore Godal, executive secretary of GAVI, says the surprise results
from Soweto are good for the field. "They force us to answer the
important questions about disease burden and vaccine efficacy," he says.
Researchers are hopeful that the GAVI decision to back an accelerated
development and introduction plan for pneumococcal vaccines (see above)
will now help to kick-start precisely the kind of studies needed to
answer these questions.
1. Williams, B.G. et al. Estimates of worldwide
distribution of child deaths from acute respiratory diseases. Lancet
Infectious Diseases 2, January 2002.
2. Klugman, K. Presentation to 3rd International
Symposium on Pneumococci and Pneumococcal Diseases, Anchorage, Alaska,
Return to July 2002 contents page