¿ìè¶ÌÊÓƵ

Imagine you are the health minister of a poor country where AIDS has
affected every family you know. One in three adults in your city is infected
with HIV and growing numbers of children are left without parents. Everyone
knows that a vaccine is urgently needed. Next, suppose that a promising
candidate vaccine emerges-almost certainly from the West. It is possible
that your country could agree to participate in trials of that vaccine to
see whether it works. However hopeful that might seem, it would also entail
enormous ethical and scientific problems.

How, for example, would your government ensure that any vaccine proven
to work would continue to be available to your country, regardless of its
cost, after a trial ended? How would the organisers of the trial take steps
to explain the risks, as well as the potential benefits, to participants?
How confident could you be that the strain or strains of HIV on which the
vaccine was based would protect you against the strains in your region?

These questions may be hypothetical now, but they will not remain so
for long. Trials of potential vaccines against HIV could begin in parts
of Africa, Southeast Asia and Latin America in as little as five years.
No one has firm plans yet, but a number of developing countries want trials,
and teams in American and European institutions are already working with
some of them.

As yet, there is no vaccine ready to be tested for its ability to protect
against HIV infection. Nevertheless, several candidate vaccines are now
being put through their paces in healthy volunteers, mainly in the US, to
evaluate their safety-phase 1-and the kind of immune response they produce-phase
2 . While researchers know they still face important obstacles in the design
of a vaccine, they will face even greater ones in future if they do not
start preparing the ground now for phase 3, the test of whether a vaccine
works. So they are preparing.

In Uganda, for example, Britain’s Medical Research Council is working
at the request of the Ugandan government with scientists in Entebbe at the
Uganda Virus Research Institute and others, to build up data on infection
in the population and the strains of HIV that are present. Geoffrey Schild,
head of the MRC’s AIDS Directed Programme, has recently said that Uganda
‘could be one of the appropriate sites for some of the eventual trials of
a vaccine’s effectiveness’. And a dozen universities in the US, supported
by the National Institutes of Health, are training researchers from developing
countries with possible vaccine trials in mind. The participating countries
in the NIH’s scheme include Haiti, Brazil, Peru, Thailand, Uganda and Zimbabwe.
The US Department of Defense also has collaborative projects with researchers
in many countries, through the Walter Reed Army Institute of Research in
Rockville.

But is it right that the world’s poorest populations should be the testing
grounds for vaccines against a disease that affects people everywhere? At
its worst, might this approach not look like ‘safari research’-the exploitation
of the vulnerable by the unscrupulous? ‘That’s exactly what we are trying
to avoid,’ says David Heymann, head of research at the World Health Organization’s
Global Programme on AIDS. In fact the WHO, the NIH and others are so sensitive
to this issue that they are holding meeting after meeting, some of them
closed, to discuss it. Next month in Geneva, the WHO’s Vaccines Development
Steering Committee, which represents member states, will finalise new guidelines
for any country doing research on people, particularly in vaccine trials.

Many epidemiologists think that trials will have to happen in the developing
world, for example in Africa, because it is there that the virus is spreading
most rapidly through the adult populations of many cities. In order to test
whether a vaccine protects against HIV, scientists need to compare the rate
of new infections in vaccinated individuals with the rate in others who
have received a placebo, over a given period, in an area where the risk
of being infected is high. The higher the incidence of new infections each
year, the fewer the participants and the shorter a trial need be before
the vaccine is either proven or found wanting.

The effect of the incidence level on the size of a trial is not trivial.
For example, in a population where ‘only’ two people per thousand become
infected in a year, a trial would need to involve between 8000 and 40,000
people, depending on how well a vaccine was thought to work and how stringent
the statisticians wanted to be with their analyses. By contrast, in a population
where one in three people become infected each year, a trial could involve
only hundreds of individuals.

But despite the relatively low incidence of infection in the rich countries
overall, some individuals are at high risk of infection, for example intravenous
drug users and the uninfected partners of HIV-positive people. Last year,
a working party on HIV vaccines for the European Community said that such
groups might be possible participants in a trial in Europe (AIDS, vol 4,
p 1043). And privately at least, some scientists think it should be equally
possible to find such a group in the cities of North America.

Jaap Goudsmit, director of the Human Retroviruses Laboratory at the
University of Amsterdam, and a member of the EC working party, says there
are advantages in doing a trial in the West. First, the researchers will
already have done long-term studies on the group and will have baseline
information on the rate of new infections. Secondly, they will probably
know more already about the prevalent strains of HIV in that population.
In Africa, by contrast, much of this information has yet to be collected.

‘I think it could be done in Amsterdam among drug users,’ says Goudsmit.
Although the incidence of HIV in this group appears to have dropped sharply
with the advent of needle-exchange schemes and intensive counselling, people
still take risks and the rate has reached a plateau at around 5 per cent.
Yet even with several cities combined, there might be ‘room’ for only one
such trial. The MRC believes that no more than one phase 3 trial, if any,
would be feasible in Britain.

Some scientists think it would be difficult to keep track of intravenous
drug users. Ironically, the very behaviour that puts such people at risk
of infection, and therefore makes them eligible for a trial, is illegal.
Would drug users be deterred from returning for checkups and tests through
fear that their participation might bring them into contact with the law?
In medical jargon, this translates into ‘problems of compliance and follow-up’.
As the EC working party put it: ‘Structured cohorts of individuals who are
both at risk for infection and reliable for continuous follow-up might be
difficult to establish.’

Mary Lou Clements, director of the Center for Immunisation Research
at Johns Hopkins University, agrees that there might be difficulties in
monitoring drug users. (Johns Hopkins is one of the NIH’s units for phase
1 and 2 trials, and also trains researchers from developing countries.)
But, she says, studies by researchers in Baltimore have shown that proper
cooperation and follow-up with this group is possible. ‘And they have got
around some of the confidentiality issues,’ she says. ‘But we really haven’t
planned trials in this population in the US as yet.’

Some scientists question the implicit assumption that participants in
trials in developing countries would somehow be more cooperative and easier
to follow up. Bruce Forrest, clinical trials coordinator for AIDS vaccines
at the MRC, says that the problems of follow-up could be equally serious
wherever a trial was done. For example, people not used to the West’s medical
ideas might see little point in visiting the clinic for regular checkups
if they felt well.

Mike Bailey, HIV/AIDS adviser to Save the Children, says: ‘How can anyone
suggest that an African study population would be more compliant? That must
mean you are going to place all sorts of impositions on what they are allowed
to do. People in Africa expect to move about, between cities and rural areas,
to markets, to work, to visit relatives.’ Clements agrees that ‘it won’t
be any easier’ to do trials in Africa than in the US.

Meanwhile, developing countries want to get on with trials as soon as
they can. The WHO is supportive, says Heymann: first, because the problem
is at least as serious, if not more so, for them as anyone else, and secondly,
because researchers need to make sure they can protect against the less-known
strains of HIV that are prevalent in Africa. Staff from the WHO are already
inspecting a number of possible test sites in countries that have requested
trials, says Heymann. Until specific sites are chosen, no one is naming
names, but a good guess might include Bangkok as well as several central
African cities. ‘Finally six or seven sites will be selected by the steering
committee,’ says Heymann. Once that is done, and it will probably be within
months, the WHO will hold another planning meeting involving the other main
research bodies, such as the NIH and the MRC, to agree training programmes
for local researchers and other ways in which the Western collaborators
could strengthen the infrastructure of the country involved. For example,
laboratory facilities will almost certainly need improving.

Money and ethics

The WHO is short of money, however, and it is unlikely to be the only
sponsor of trials. This year, the Global Programme on AIDS is running on
funds only three-quarters of its projected budget, despite the rapidly worsening
pandemic (This Week, 8 December 1990). Schild at the MRC is convinced that
the WHO should be the sole umbrella organisation to coordinate trials, and
the sole body to agree criteria for vaccines. But other, more wealthy agencies,
particularly the NIH, are also keen to take an active role that they see
as complementary to the WHO’s efforts. Dale Lawrence, senior scientific
adviser for the vaccines branch at NIH says: ‘We are encouraging a thorough
review of the international possibilities and how we might best use our
resources in collaboration with others to find safe, ethical and meaningful
ways to test vaccines.’ Heymann says the organisation sees the participation
of other bodies positively. ‘We’re certainly not trying to exclude other
people,’ he says.

The WHO’s main ethical concerns, says Heymann, are that vaccines should
meet internationally agreed safety standards; that any country conducting
a trial must have a national policy for AIDS and a system of healthcare
that will provide advice, counselling and medical treatment as well as the
test vaccine; that phases 1 and 2 should already have been done in the vaccine’s
country of origin, and then, probably, repeated in the country involved;
and that the country should have been able to collect and isolate the prevalent
strains of HIV in its population.

All countries will be encouraged to use an ethical ‘check list’ and
contracts between the Western collaborators and the host country’s scientists.
‘We are working right now to develop the check list,’ says Heymann. This
will ask, for example, whether the country has a proper procedure for obtaining
informed consent from individuals before they join a trial. What is understood
by informed consent is, of course, the nub of the whole problem. For example,
someone who cannot read will have to have the potential risks and benefits
of participating explained to them only verbally.

In vaccine trials anywhere in the world, people will have to be told,
for example, that there is a fifty-fifty chance they will get a placebo,
and that the vaccine is unproven and therefore affords no licence to let
down their guard against risky behaviour such as unprotected intercourse.

‘One would have to use the appropriate means of informing people,’ says
Clements. ‘It would probably be verbal consent, and then you need some way
of verifying that they do understand.’ Staff would also need to be educated
to a higher standard. ‘The counselling’s far greater than with any other
type of vaccine,’ says Clements.

There is a risk that private foundations or eminent scientists could
finance trials and bypass the international guidelines. ‘It’s always a real
possibility,’ admits Heymann. But, he says, ‘We are not a police organisation.’
Nzilambi Nzila, joint director of Project SIDA in Kinshasa, Zaire, is no
stranger to this risk. ‘Everywhere when we go to AIDS conferences people
are pushing for trials; and we say that we are not ready,’ he says. No one
can enforce guidelines, agrees Clements. Would pharmaceuticals companies
be tempted to rush trials of their candidate vaccines? Clements thinks that
any company that did would be ‘very unwise’ ; the FDA has warned all American
companies in no uncertain terms against rushing into poorly designed trials,
she says. But she admits that it could happen elsewhere.

Just as the rich countries are sensitive to suggestions that they are
about to embark on safari research, the poor countries are equally sensitive
to suggestions that they might allow it. In Zaire, for example, the chairman
of the national ethical committee, Bila Kapita at Mama Yemo Hospital, Kinshasa,
has made it clear that his committee will scrutinise any proposal for a
trial. Zaire is sensitive about vaccine trials after the controversy created
by Daniel Zagury, a French researcher who injected himself and some Zairean
children with an experimental preparation (see This Week, 16 March 1991).
AIDS researchers in Kinshasa are loath to discuss Zagury, but privately
they distance themselves from his work. Instead, Projet SIDA is discussing
possibilities for future trials with US researchers (see ‘Africa’s Growing
AIDS Crisis’, ¿ìè¶ÌÊÓƵ, 17 November 1990).

In fact it is a Zairean who has just written a hypothetical protocol
for a phase 1 and 2 trial in Zaire. Kayembe Kalambayi from the University
Clinic in Kinshasa has just returned home after a spell studying at Johns
Hopkins in Clements’s team, funded by NIH money. He worked with the Americans
on the safety trials in the US of one candidate vaccine, a recombinant form
of gp160. ‘We’re encouraging him to get on as quickly as possible,’ says
Clements.

Under a separate scheme, the NIH’s international arm, the Fogarty International
Center, is paying for the training of epidemiologists and others, from participating
countries in the developing world, in AIDS research in the US. For example,
Case Western Reserve University in Cleveland, Ohio, is training researchers
from Uganda; the University of Washington in Seattle trains Kenyans, Thais,
Senegalese and Zimbabweans.

One of the most urgent priorities is to ensure that the world’s widely
variable strains of HIV are collected. No one knows how much the variation
between strains within and between different geographical areas will matter
for an effective vaccine, because no one knows exactly what part of the
immune response to HIV is correlated with protection from disease . But
in the meantime, some information on what strains are where is long overdue.

With this in mind, the WHO has recently begun to set up a network of
primary laboratories in developing countries to collect samples of HIV.
Where facilities are limited, the isolating of the strains may have to be
done by a laboratory in the West. Then the strains will be stored as an
international repository with existing ‘strain banks’, partly in the US
at the NIH and partly in Britain. The strains can then be sequenced and
scientists can find out whether antibodies to one strain can neutralise
another.

In Amsterdam, Goudsmit is coordinating a project with Rafael Najera
from the Charles III Institute, Madrid, sponsored by the EC, to collect
and test strains from across Europe and the Soviet Union. ‘We want to see
if you can actually cross-neutralise virus from Portugal with serum from
Denmark, for example,’ says Goudsmit. So far, he and his colleagues have
collected 50 strains, but they hope to get about 200. This June in Madrid,
the Europeans will meet with colleagues from the Walter Reed Army Institute
for Research and other researchers from the US to discuss their work so
far, then the American institute will exchange strains with the Europeans.
Walter Reed’s laboratories all over the world have been collecting strains
for several years. Finally, the WHO will receive the strains.

This is a crucial step for vaccine developers. Until very recently,
most companies making vaccines had been restricted to working with American
or European isolates, such as LAV/IIIB, MN and SF2. With access to a wider
range, they may be able to find out more about the necessary components
of broadly protective vaccines. At Chiron in California, researchers have
tested recombinant gp120 vaccine based on strain SF2 in the laboratory against
a range of strains, including all the usual North American ones but also
two from the Caribbean and two from Zaire. One of the Zairean strains, NDK,
is highly virulent and kills cells 10,000 times more efficiently than IIIB
(see This Week, 20 October 1990).

A vaccine for all

Larry Kurtz of Chiron says antibodies to the gp120 were capable of neutralising
all these strains. ‘We are not making a vaccine just for San Franciscans,
we are making a vaccine hopefully for the whole world,’ he says. Just suppose
the Chiron vaccine turned out to be effective. How would the company deal
with the issue of its cost for the world’s poor? ‘We would acknowledge that
there should be a mechanism in place whereby people can afford to buy it,’
says Kurtz. So far, only one realistic mechanism has been proposed (see
This Week, 15 September 1990).

This issue will not go away in the coming years. But in the meantime,
trial planners will face many other practical problems. There is the question
of exactly what we mean when we talk of protecting people with a vaccine.
Do we mean protection from infection, or do we just mean protection from
disease? The difference is a big one, in the case of HIV. Most vaccines
seem to work by limiting the invading organism’s attack to the immediate
site of entry, then overwhelming infection. In the case of HIV, this could
be a disaster. If any virus survives anywhere in the body, the person may
remain healthy but will technically be infectious to others. So vaccinated
individuals who continued to practise risky behaviour might be safe themselves,
but would be infecting others.

How should researchers measure the success or failure of a vaccine?
If they wait to use progression to AIDS as a measure of failure, they will
have to keep a trial running for many years, because of the slow course
of the disease. If they use other markers of success or failure, such as
declining levels of T4 cells, they will be able to end a trial at an earlier
stage but they will still need to follow up the participants for several
years, says Forrest, to be certain of the outcome. It would also be necessary
to have an independent body to watch the trial’s progress. This body, unlike
the researchers or the vaccinees, would have access to the data during the
trial and could advise its coordinators to discontinue the trial immediately
if, for example, the vaccine was shown to be successful or had adverse effects.

In the end, everyone is hoping for a vaccine, or several vaccines, that
are affordable to everyone and effective all the time. Whether it takes
five years or 20, most scientists are now much more confident that they
can design the product. Whether the world can devise a fair way to distribute
it is another matter.

* * *

1: Putting candidate vaccines through their paces

¿ìè¶ÌÊÓƵs are already testing candidate vaccines in healthy, HIV-negative
volunteers, mainly in the US. These trials seek to determine whether a vaccine
is safe, what dose is best and what sort of immune response it produces-but
not yet whether it protects against infection.

The main coordinator of such trials is the NIH, which has trial units
at five universities: Rochester, Vanderbilt, Johns Hopkins, St Louis, and
the University of Washington, Seattle. The US Department of Defense is also
conducting trials based at the Walter Reed Army Institute of Research. Britain’s
MRC currently has one study.

Most of the candidate vaccines are based on HIV’s protein coat (gp120
or gp160). The main ones currently on trial in the US are: recombinant gp160
made in insect cells by MicroGeneSys; recombinant gp160 made in mammal cells
by Immuno; recombinant gp160 expressed in a live vector, the vaccinia virus,
and made by Oncogen/Bristol-Myers Squibb; recombinant gp120 made in yeast
cells by Chiron/CibaGeigy, and recombinant gp120 made in mammal cells by
Genentech. In Britain, British Biotechnology and the MRC are testing virus-like
particles, which are produced by yeast cells and carry the viral core protein
p24, as a component of a possible vaccine.

* * *

2: The search for an immune response that counts

No vaccine designer has ever faced a challenge quite like HIV. Not only
does the virus attack the very cells that normally defend the body against
infection; it is also capable of remaining hidden in its host’s cells, out
of reach of the immune system. What is more, HIV’s structure can vary considerably:
its protein coat changes rapidly and unrelated strains may be as much as
30 per cent different from each other at the most variable region of the
coat, the V3 loop. Faced with these problems, researchers have been encouraged
in the past 18 months by a series of successful animal experiments.

The only way to prove that a vaccine protects against infection is to
challenge animals with a dose of virus after they have been immunised, and
compare them for signs of infection with unvaccinated animals challenged
at the same time.

However, there is no perfect animal model for AIDS. Chimpanzees can
be infected with HIV, but the virus does not appear to make them ill as
it does humans. In any case, research on chimps is highly controversial,
as well as expensive. So far, the animals that seem most appropriate for
vaccine research are macaques. These primates have their own family of naturally-occurring
simian immunodeficiency viruses (SIVs), which resemble HIV and make the
animals ill with an AIDS-like disease.

Late in 1989, Michael Murphey-Corb and her colleagues at the Delta Primate
Center announced that they had protected macaques that had been immunised
with whole, inactivated SIV against a challenge of the same strain of virus
(¿ìè¶ÌÊÓƵ, Science, 16 December 1989). The finding is not directly
applicable to people: most scientists would not consider vaccinating healthy
humans with whole inactivated HIV, because there remains a risk that the
virus could become active in the body. However, the result marked an important
first step.

Then, last June, Philip Berman and his colleagues at the company Genentech
in California, and others at the Southwest Foundation for Biomedical Research
in San Antonio, came up with another surprise. Berman and colleagues had
succeeded in protecting two chimpanzees against challenge with HIV. The
team had immunised the animals with a recombinant form of gp120, the outer
part of HIV’s protein coat. (The entire protein, gp160, cleaves to form
two parts, gp120 on the outside of the virus and gp41 spanning its membrane.)
Two other chimps had been inoculated with a second protein-a recombinant
version of gp160-but showed no signs of protection. Why this should be so
remains unclear, and some researchers question whether we should conclude
much from a result involving only four immunised animals.

Meanwhile, no one knows whether the protection offered by gp 120 will
turn out to be permanent. At a meeting of AIDS researchers in Paris last
October, Berman confirmed that the two chimps were still free of signs of
infection more than a year after challenge. But another researcher at the
meeting, Marc Girard at the Pasteur Institute, told a different story. He
had immunised three animals, each with different combinations of viral proteins.
After six months, all three animals appeared to be protected against challenge,
but at 32 weeks Girard isolated virus from one of them. This, he said, underscored
the ‘worrisome’ fact that HIV can remain hidden in cells, defying detection
by all the standard methods.

Two more recent experiments have, however, given researchers a more
upbeat mood. A team of MRC scientists for the AIDS Directed Programme has
shown that macaques can be protected against infection with SIV by a vaccine
consisting of cells infected with whole inactivated SIV (The Lancet, vol
336 p 1538). Not only did the vaccine produce a strong response of antibodies
and T cells in the animals, but it also worked after a shorter dose schedule
than the Americans’ experiments-four months compared with 10.

Now, in a second experiment the British team has protected macaques
against a different strain of SIV from the one used in the vaccine. While
it is dangerous to conclude too much about HIV from studies with SIV, this
finding at least increases optimism about the possibilities of protecting
people against divergent strains of HIV-1. The results, reported at recent
meetings in New Orleans and Florida, have yet to be published, but the team
says the difference between the two strains of SIV is ‘consistent with’
the average degree of variation between strains of HIV-1 found in the West,
which is around 13 per cent.

Despite all the optimism, scientists concede that they still do not
know precisely which components of the immune response to HIV are the crucial
parts that would provide protection. ¿ìè¶ÌÊÓƵs believe that both ‘arms’
of the immune system, antibodies and T and B cells, will have to be stimulated.
To get a handle on the problem, researchers are testing HIV’s envelope protein
bit by bit to find the most effective antigens-that is, those parts of the
protein that stimulate large numbers of neutralising antibodies and specific,
cell-killing T cells.

However, it is too simplistic to hope that there might be one crucial
portion of the protein that everybody’s T cells would instantly recognise,
says Robert Siliciano of the Johns Hopkins University. Everyone has a slightly
different set of genes for the HLA (human leukocyte antigen) complex, which
governs the way foreign proteins in the body are presented to the immune
system. Some people will respond to one part of the protein, others to another.
Nevertheless, says Siliciano, ‘most people can probably respond in some
way to the envelope protein’.

A potential pitfall with recombinant vaccines is that genetically engineered
forms of HIV’s envelope protein may differ in shape from the natural form.
This is because of the way sugar molecules are attached to proteins inside
cells. In theory at least, a protein’s shape may strongly influence the
way it ‘looks’ to the immune system.

Even if we can pinpoint the crucial antigens, we are still worryingly
ignorant about the way the immune system works when it is challenged by
infection with HIV through the usual route-that is, across the mucus membranes.
In other words the exact process of infection through sexual intercourse,
through the lining of the vagina or rectum, is still not properly understood.

Bruce Forrest, clinical trials coordinator for AIDS vaccines at the
MRC, stresses the need for more research into this mechanism of immunity,
particularly in women. There is evidence, for example, that immunoglobulins,
or antibodies, in the cervical mucus fall to low levels during ovulation.
How might this affect infection and protection against it?

‘In comparison with other areas of immunology, there really has been
very little work published so far on aspects of mucosal immunity in the
female genital tract,’ says Forrest.

One important problem is that no one knows whether the animal experiments
performed so far can tell us anything about sexual infection. Until now,
all the vaccinated animals that experimental vaccines have protected were
challenged by intravenous injection of the virus directly into the bloodstream.
But would they be equally protected against mucosal infection? Just before
Easter, the MRC team challenged macaques with SIV across the mucous membranes.
The full results of the experiments will not be known for some months.

Another area of ignorance is how human cells infected with virus, rather
than virus on its own, transmit infection. For example, does HIV spread
from person to person mainly through infected lymphocytes, or is, say, semen
more infectious than we had previously thought? ‘There is a considerable
gap in our knowledge,’ says Forrest. Until these and other questions have
been answered, the path to an effective vaccine still looks like an obstacle
race.