Customize your JAMA Network experience by selecting one or more topics from the list below.
Letvin NL, Bloom BR, Hoffman SL. Prospects for Vaccines to Protect Against AIDS, Tuberculosis, and Malaria. JAMA. 2001;285(5):606–611. doi:10.1001/jama.285.5.606
Author Affiliations: Department of Medicine, Harvard Medical School, Beth Israel Deaconess Medical Center, Boston, Mass (Dr Letvin); Office of the Dean, Harvard School of Public Health, Boston, Mass (Dr Bloom); and Malaria Department, Naval Medical Research Center, Silver Spring, Md (Dr Hoffman).
Given the scope of the worldwide health problems caused by the acquired
immunodeficiency syndrome, tuberculosis, and malaria, it is imperative that
vaccines be developed to prevent these infections. Recent advances in the
understanding of these diseases suggest that T-lymphocyte–mediated immunity
is important in containing these infections. The application of novel vaccine
technologies for eliciting this type of immunity promises to provide successful
vaccines for controlling the spread of these deadly infections.
Vaccination is perhaps the most powerful of all medical interventions.
During the past 25 years, vaccination has eliminated smallpox worldwide, polio
from the western hemisphere, and Haemophilus influenzae as a cause of life-threatening disease in North America and Europe.
Prevention of infection by vaccination has improved the health status of human
populations throughout the world. Despite this record of accomplishment, a
virus (human immunodeficiency virus [HIV]), a bacterium (Mycobacterium tuberculosis), and a parasite (Plasmodium
falciparum) now kill more than 5 million people annually.1
HIV and the acquired immunodeficiency syndrome (AIDS) have only become an
epidemic in the recent past. Emergence of multidrug-resistant M tuberculosis and the immunosuppression caused by HIV have dramatically
increased the incidence of tuberculosis (TB) and death caused by TB throughout
the world. Development of drug resistance by parasites and insecticide resistance
in the Anopheles species mosquito, the deterioration
of political and health care infrastructures in many parts of the world, movement
of nonimmune refugee populations into malarious areas, and the population
explosion in sub-Saharan Africa have all contributed to the increased incidence
AIDS, TB, and malaria are more severe problems in poor countries than
in affluent ones, and these diseases have not received an investment in research
dollars commensurate with their importance. There is a growing realization
of the impact of these diseases and an increasing belief that these diseases
can be controlled by vaccination.
In this article, we describe the global impact of AIDS, TB, and malaria;
summarize the current understanding of how the human immune system can be
mobilized to contain the pathogens responsible for these diseases; and describe
the approaches for vaccines that promise to eliminate these diseases.
According to the Joint United Nations Programme on HIV/AIDS,2 more than 18.8 million people worldwide have died
of AIDS and 34.3 million are infected with HIV, with 5.4 million people newly
infected with HIV in 1999 alone. More than 13.2 million children have been
orphaned by AIDS, and the disease has had a profound impact on rates of infant,
child, and maternal mortality, life expectancy, and economic growth. In 16
countries, more than one tenth of the adult population aged 15 to 49 years
is infected with HIV. In Botswana, 35.8% of adults are infected with HIV,
and in South Africa, 19.9% are infected.2
Anti-HIV drug therapy will have a limited effect on containing the AIDS
epidemic. Although available drugs decrease virus replication in infected
individuals, the virus persists, even in those in whom anti-HIV therapy has
eliminated measurable plasma virus. Ongoing viral replication will allow the
emergence of drug-resistant HIV variants. More important, the real costs of
providing antiretroviral drugs to the millions of individuals in need of treatment
is beyond the financial resources of populations in the developing world.
HIV can only be controlled worldwide by development of an effective vaccine.
HIV infection occurs most commonly worldwide as a result of venereal
spread. While rare individuals have a relative resistance to HIV infection,
all people are ultimately susceptible to HIV infection. Preventing infection
and the pathologic consequences of infection can be accomplished only through
Although HIV causes immune dysfunction by several mechanisms, the central
immune abnormality is a loss of CD4 T lymphocytes.3
Antibodies that recognize the envelope glycoproteins of HIV can neutralize
the virus in vitro and block infection in a nonhuman primate. However, because
of the extreme sequence variability of the envelope glycoproteins, such neutralizing
antibodies are usually isolate-specific. The central role that CD8 cytotoxic
T lymphocytes (CTLs) and CD4 T lymphocytes play in containing human HIV infections
and monkey simian immunodeficiency virus (SIV) infections has been the subject
of intense interest. During primary infection, HIV and SIV replication is
contained by CTL responses.4,5
Potent CTL responses are associated with low virus loads and quiescence of
clinical disease.6 CD8 lymphocyte–depleted
monkeys are unable to contain SIV replication during primary or chronic phases
of infection,7 and after virus challenge, SIV
replication is contained in monkeys with vaccine-elicited CTL.8
Preservation of HIV-specific CD4 T-lymphocyte function correlates with containment
of HIV replication in infected individuals.9
Taken together, these observations suggest that an effective HIV vaccine should
elicit potent virus-specific CTL and CD4 T lymphocytes.
Several vaccine technologies elicit high-frequency CD4 T-cell and CD8
CTL responses.10 Live recombinant organisms
are being explored as potential vaccines. In this approach, a gene encoding
a protein of a pathogen can be inserted into an organism that infects humans
but does not cause disease. This recombinant organism will then express the
product of the inserted gene, and immunity to that gene product will be elicited
as part of the immune response to the organism. Genes of HIV, M tuberculosis, and malaria have been inserted into such diverse vectors
as pox viruses, bacille Calmette-Guérin (BCG), adenoviruses, and enteric
bacteria. These approaches offer the possibility of eliciting the same long-lasting,
potent immunity that is induced by infection with a live organism but without
delivering a potentially pathogenic organism to the vaccinee.
Another approach for inducing T-cell responses is the direct injection
of plasmid DNA expressing a gene encoding a protein antigen. After intramuscular
or intradermal injection, DNA plasmids are taken up by cells and the encoded
protein antigens are expressed. The proteins are processed by immune cells
and generate strong, persistent cellular immune responses.
Other strategies under investigation combine 2 different vaccine technologies.
The rationale for such "prime-boost" strategies is that the combination of
2 different immunization modalities that induce a cellular immune response
by different mechanisms can be combined and synergize elicitation of T-cell
immunity. "Prime-boost" approaches can also include 1 vaccine modality that
elicits CTL and another that elicits neutralizing antibody responses. These
approaches to vaccination are likely to provide the means of inducing meaningful
immunity to these pathogens.
All of these strategies are under investigation as potential vaccines
to prevent HIV infection. Because all potential vaccines cannot be evaluated
for immunogenicity and efficacy in human populations, potential vaccines are
evaluated in nonhuman primates, and the data from these studies are used to
select the more promising approaches for early phase human testing.
Protection against challenge with highly pathogenic AIDS virus isolates
has not yet been achieved in nonhuman primate studies. Nevertheless, data
from nonhuman primate investigations suggest that eliciting T-cell immunity
by vaccine prior to virus infection can alter the pathogenic consequences
of infection.8 For example, previously vaccinated
monkeys have no measurable virus in the plasma and no loss of CD4 T lymphocytes
following infection. This raises the possibility that prior vaccination can
reduce viral replication in humans subsequently infected with HIV. Such individuals
are predicted to manifest decreased disease burden and decreased HIV transmission
Plasmid DNA and recombinant pox strategies are being assessed in human
clinical trials, and additional HIV vaccine strategies will be assessed in
humans in the near term. However, until a vaccine can elicit antibodies that
neutralize a diversity of HIV primary patient isolates, true protection against
an infection might not be achieved. Improved virus-specific T-cell responses
should contain viral replication and reduce disease manifestations among those
who become infected.
Creation of an HIV vaccine is feasible, but a worldwide commitment is
needed to develop such a vaccine. The industrialized nations of the world
must commit the resources to develop this vaccine, and developing nations
must create and provide the infrastructure to facilitate the testing of vaccine
Each year, TB is responsible for 8 million new cases, 2 million deaths
worldwide, and contributes to the deaths of an additional 900 000 people
with AIDS.11 Thus, TB and AIDS are the largest
causes of mortality from infectious diseases. Tuberculosis affects about 16
million people worldwide, with a case-fatality rate of 50% in untreated disease.
In some countries with the highest prevalence of HIV coinfection, the fatality
rate is about 23%. In an autopsy study in Africa, TB was the cause of death
in 32% of AIDS deaths and a contributory cause in an additional 15% to 25%.12 The peak age of incidence of TB is 15 to 25 years
of age, but infection can persist, often for a lifetime, in a silent form,
and can reactivate with HIV infection.13 Only
1 in 10 individuals infected with M tuberculosis,
determined by tuberculin skin testing, develops disease within a lifetime,
whereas the risk is about 8% per year in immunodeficient individuals. This
suggests that most infected individuals are protected by an immune response
and that enhancement of the natural immune responses could increase resistance
to disease. Epidemiological models predict that even a 50% effective TB vaccine
would have a major impact on the disease and would save perhaps 40 million
lives over a decade.14,15
Perhaps due to its unusual waxy coat, the tubercle bacillus is impervious
to most antibiotics. Resistance develops quickly to single drugs, and a complex
regimen known as directly observed treatment, short course (DOTS) requires
administration of 3 to 4 drugs for the initial 2 months of treatment, followed
by 2 drugs for 4 to 7 months.16,17
The complete DOTS regimen is accessible to only 15% of patients worldwide,
and multidrug-resistant TB has emerged and is extremely difficult and expensive
to control. Globalization and migration represent threats for transmission
of drug-resistant TB to the United States. Drug-resistant TB is a major threat
and emphasizes the need for new tools for prevention and treatment.
The BCG vaccine, an attenuated strain of Mycobacterium
bovis discovered in 1908 and first used in humans in 1921, is the most
widely used vaccine in the world, being administered currently to about 104
million children. The BCG vaccine clearly prevents death caused by disseminated
TB and TB meningitis in children, but its effectiveness in adults varies.18 For example, in a large prospective trial in the
United Kingdom, BCG resulted in 77% protection for teenagers,19
whereas a similar trial in India was not protective in any age group20; other trials have efficacies between those extremes.
The reasons for this wide variability are not fully clear. Since children
represent only 10% of TB cases, the impact of BCG on the epidemiology of TB
in adults has not been great.
The lower lung is the target of primary infection by M tuberculosis, but the bacilli spread hematogenously to the apex of
the lungs or to other organs where disease becomes manifest. In animals, BCG
does not block infection, but limits hematogenous spread and disease. The
immune mechanisms that provide protection against TB have not been identified.
Most evidence indicates that cell-mediated immunity is essential for protection
and that antibodies play little role. Mice with targeted disruptions of immunologically
important functions, eg, deficient in CD4 T cells and in the ability to produce
interferon-gamma, become highly susceptible to TB, indicating that lymphokines
and macrophage activation are essential for protection. Cytotoxic T lymphocytes
may also be required.
More than 100 vaccine candidates have been tested in animal models.
With the genome sequence of M tuberculosis complete,
and that of BCG and nonpathogenic mycobacteria in process, additional targets
for vaccines are certain to emerge. Potential vaccine concepts include the
Subunit vaccines, consisting of mycobacterial
protein, lipid, and carbohydrate antigens in various formulations, have the
potential to be specific, defined, and safe. Their disadvantage is limited
persistence in vivo and the nature and duration of the immune responses they
DNA vaccines that encode several M tuberculosis antigens are protective in mice. These are easy to produce,
relatively inexpensive, and induce long-lasting cell-mediated immune responses.
Formulation of the DNA in adjuvants and alteration of the composition of DNA
increase immunogenicity. Safety and duration of protection are not yet defined.
One promising report indicates that a DNA vaccine encoding the 65-kd antigen
was able to abrogate persistence or latency in a mouse model, which live BCG
was unable to do.24
Nonmycobacterial microbial vectors such as
engineered Salmonella and vaccinia organisms that
express mycobacterial antigens are being tested. The former has the potential
to induce mucosal immunity, and the latter can induce CTLs.
Live attenuated mycobacterial vaccines, including
nonpathogenic mycobacterial species and genetically engineered BCG expressing
immunodominant antigens of M tuberculosis, are under
development. In a UK trial, Mycobacterium microti
was as effective as BCG in protection, even though it produced positive skin
test conversion in only a fraction of the vaccinees. In addition, genetically
attenuated strains of M tuberculosis, including auxotrophic
mutants and mutations in genes relevant to persistence and virulence are all
under development and some appear to be effective in mice. They have the advantage
of containing a wide range of antigens, the adjuvanticity associated with
mycobacteria and persistence, but safety in immunodeficient individuals has
not been established.
There are 4 general strategies to test safe and immunogenic vaccines
for efficacy in humans:
Infants at high risk for early infection and disseminated disease or
meningitis would be vaccinated close to birth, and the impact on prevention
of infection, acute disease, or persistence would be assessed, perhaps over
a 3- to 5-year period. It is not known whether protection against acute disease
in young children would predict efficacy in teenagers and adults.
The target population is tuberculin-positive healthy adults, in areas
with a high reactivation rate, eg, 3% per year. Diminution of reactivation
would be assessed over 3 to 5 years. It is unclear whether results of trials
in individuals already immunized by natural infection will predict protection
in unimmunized individuals.
From experience with BCG trials, this design would require large numbers
of individuals who would have to be followed up for 15 to 20 years, but would
provide the most unambiguous resolution of vaccine efficacy.
An epidemiologically characterized population would be immunized. Those
already tuberculin positive would be analyzed for protection in 3 to 5 years,
and the efficacy among uninfected individuals could be ascertained in the
same trial over a longer period of time.
Since efficacy trials are long, complex, and expensive, surrogate markers
that correlate with immunologic protection are needed. These markers, whether
lymphokines, CTLs, or mycobacteriocidal activities, if measurable rapidly
and quantitatively, would accelerate vaccine development and testing.
Although TB occurs in every country of the world, more than 80% of cases
occur in developing countries. This limits the market for TB vaccines greatly
and is a major disincentive to the pharmaceutical industry to develop them.
To make the investment by industry in vaccine development feasible, there
is a need for public-private collaboration, in which public sector investments
move the research forward ("push") and ensure markets or purchase of vaccines
found to be effective ("pull").
Finally, the collaboration of developing countries where the disease
is most prevalent will be essential to evaluate the safety and efficacy of
The mortality rate of children with severe malaria in primary care hospitals
in the developing nations has not been reduced in 25 years. In many parts
of the world, malaria is more common today than 25 years ago. Malaria accounts
for an estimated 300 million to 500 million new infections and 1 million to
3 million deaths annually and is thought to reduce the annual gross domestic
product in sub-Saharan Africa by 1% to 4%.25
There is currently no malaria vaccine, and there is little prospect for deployment
of an effective vaccine in the next 5 years against any of the malaria parasites
that infect humans.
Parasites that cause malaria are more complex than the viruses and bacteria
for which vaccines are available, making vaccine development difficult. They
have approximately 6000 genes, and a multistage life cycle with stage-specific
expression of many proteins at each stage. This means that antibodies against
a protein on sporozoites (the stage inoculated by mosquitoes) will generally
not recognize the major protein on the surface of erythrocytic stage merozoites.
A single individual can be infected with more than 5 different strains of P falciparum (allelic variation), and one of the proteins
expressed on the surface of the infected erythrocytes, a protein that is important
in disease pathogenesis, can evade antibody responses by expressing 50 to
100 antigenically different variants.
In contrast to vaccines for other diseases, several types of malaria
vaccines can be envisioned. One for nonimmune travelers would prevent infection
of erythrocytes, thereby preventing all clinical manifestations (type 1).
A second type for young children in sub-Saharan Africa would limit replication
at the erythrocytic stage without preventing infection, thereby preventing
the 1 million to 3 million deaths caused by malaria annually (type 2).26,27 There is also interest in developing
a malaria vaccine directed to the sexual phase of parasite reproduction that
would not protect the individual, but would reduce transmission of infection
within a community, thereby reducing disease burden.
Development of malaria vaccines is believed to be feasible because human
models are available for the approaches. Immunization of volunteers by exposure
to the bites of more than 1000 irradiated, infected Anopheles species mosquitoes provides more than 95% protection for up to 9 months
against experimental challenge with multiple strains of P falciparum. The irradiated sporozoite vaccine would be an ideal type
1 vaccine, but is impractical. Children in areas of sub-Saharan Africa with
intense malaria transmission who live until the age of 8 to 10 years do not
develop severe malaria, but do become infected and develop febrile illness.
These children have an immune response that does not prevent infection but
limits the pathological and clinical effects of the infection. A vaccine that
essentially turned infants and young children into 10-year-olds from an immunological
standpoint would be a candidate type 2 vaccine.
The challenge has been to characterize immune responses that provide
protection, to define which parasite antigens/epitopes are targets of these
protective immune responses, and to develop vaccine delivery systems that
induce the appropriate immune responses. Irradiated sporozoite-induced protection
against malaria in mice is primarily mediated by CD8 T cells that recognize
8 to 10 amino acid peptides derived from the parasite on the surface of infected
hepatocytes. This immunity is complemented by antibodies that prevent sporozoites
from invading hepatocytes and by CD4 T-cell responses against the proteins/epitopes
expressed at the liver stage. Other evidence suggests that naturally acquired
immunity in residents of malaria endemic areas is primarily mediated by antibodies
that recognize parasite proteins expressed on the surface of erythrocytic
stage merozoites or on the surface of infected erythrocytes. T-cell responses
at the erythrocytic stage, antibodies against sporozoites, and T-cell responses
against infected hepatocytes are also thought to play a role.
Three major approaches are being pursued to attempt to create a multi-antigen,
multistage malaria vaccine.
The first is designed to enhance antibody and CD4 T-cell responses against
a few key proteins, on sporozoites, merozoites, and sexual stages using purified
proteins or peptides administered in a strong adjuvant. The most progress
has been made with the major surface protein of sporozoites, the P falciparum circumsporozoite protein.28
A vaccine formulation known as RTS,S/SBAS229
consistently protects 50% of volunteers against experimental malaria challenge
for 2 to 3 weeks, but not for 6 months,30 and
among Gambian semi-immune adults in The Gambia, provided approximately 65%
protection for 2 months and no protection at 6 months. This reproducible but
unsustained protection is a step forward, but is not adequate and this vaccine
alone most likely will not reduce mortality in infants and young children
in the developing world. A protein on the surface of erythrocytic stage merozoites
(merozoite surface protein 1 [MSP1])31 is also
a component of the experimental vaccine, SPf66,32
which despite early promise has not been effective in field trials in The
Gambia, Thailand, Tanzania, and Brazil; MSP1 is also part of a trivalent vaccine
recently shown to be promising in Papua, New Guinea. Recombinant P falciparum MSP1 will be tested in volunteers, both alone and in combination
with other proteins like the P falciparum circumsporozoite
protein, and at least 5 other asexual and sexual (transmission blocking) erythrocytic
stage–derived recombinant proteins will be tested in the near future.27
The second approach is designed to induce antibody and CD8 and CD4 T-cell
responses against proteins expressed by irradiated sporozoites in hepatocytes
(n = 5) and proteins expressed on or near the surface of erythrocytic stage
merozoites (n = 10). This approach did not seem feasible using purified recombinant
proteins, but DNA vaccines make it possible to devise and evaluate multistage,
multiantigen vaccines. Early studies indicate that CTLs33
and interferon gamma–producing CD8 T lymphocytes can be elicited in
human volunteers. However, while DNA vaccines appear to prime the immune system,
boosting the immune response with recombinant viruses and recombinant proteins
appear to provide more protective immunity than DNA immunization alone.34 Clinical trials are under way with multiple liver
stage genes as DNA vaccines alone and in a prime-boost strategy.35,36
The third approach uses data from the Malaria Genome Sequencing Project.
Irradiated sporozoite and natural immunity is induced by exposure to the entire
parasite and could reflect immune responses against many proteins encoded
by the estimated 6000 genes in the P falciparum genome.
The genomic sequence of P falciparum will be completed
by the end of 2002,37,38 and a
variety of techniques are being used to identify new targets for vaccine development.
Using the genetic sequence for development of successful vaccines will require
new approaches to constructing DNA-based and polyepitope vaccines.39 Clinical trials of such vaccines could begin within
3 to 5 years.
The human models (irradiated sporozoite and naturally acquired immunity)
demonstrate the feasibility of a malaria vaccine, and developments in genomics,
proteomics, molecular immunology, vaccinology, population genetics, and quantitative
epidemiology have created great expectations for development of effective
malaria vaccines. It will be a formidable task to determine which antigens/epitopes
from which stages of the life cycle of the malaria parasite are required for
sustainable protection, which immune responses predict protection, which vaccine
delivery systems are optimal, who and when in life to immunize, and the true
impact of a malaria vaccine. However, the next 10 to 25 years should see the
development of effective malaria vaccines that will mitigate the effects of
the disease worldwide and, combined with other interventions, will eradicate
malaria in many areas. (Figure 1)
AIDS, TB, and malaria are caused by different pathogens that differ
in many ways. Nevertheless, each elicits a profound immune response and all
3 diseases could be successfully contained and perhaps eliminated through
vaccination.Cell-mediated immunity can play a central role in controlling
these infections. Several novel vaccine strategies elicit durable cell-mediated
immune responses that appear to be able to contain these infections. These
diseases are important public health problems in the developing world, and
the affected nations must provide the health care infrastructure to deal with
them. Advances in understanding of these diseases and immunization technology
suggest that vaccine protection against HIV, M tuberculosis, and malaria parasites are achievable in the coming decades. Attaining
these goals requires a concerted commitment of scientific and economic resources
to these public health problems.
Create a personal account or sign in to: