The discipline of immunology developed from observations in the fields of public health and clinical medicine. In the fifth century bc, Thucydides noted that individuals who recovered from plague did not develop disease again, and similar observations of ‘immunity’ to plague were made in Europe in the fourteenth century1,2. The observation that mild smallpox infection protected against disease on subsequent exposure led to the practice of variolation — the inoculation of dried pus from smallpox pustules into the skin or nose. This was first practised in India and China, and then introduced in 1721 in England by Lady Montague, and in New England by Cotton Mather3. Jenner's clinical trial of cowpox virus vaccination, and publication of Variolae Vacciniae in 1798, gave birth to the field of immunology, but neither an understanding of the basis for its efficacy nor universal acceptance of this practice were soon to follow3,4. Instead, scientific and public scepticism and alarm were common early responses (Fig. 1).

Figure 1: The Cow-Pock.
figure 1

The Cow-Pock or the Wonderful Effects of the New Inoculation! by James Gillray was published in England in 1802 by the Anti-Vaccine Society. The etching, which shows Edward Jenner among patients in the Small Pox and Inoculation Hospital at St Pancras (London), suggests the transformation into cows of individuals vaccinated by Jenner. Reproduced with permission from The Wellcome Library, London.

The subsequent formulation of the germ theory of disease by Koch and Pasteur, and von Berhing's identification of neutralizing factors for toxins, provided a foundation for the mechanistic understanding of protective immunity5. In the ensuing years, vaccines for more than 20 infectious diseases have been developed, and in 1977, Jenner's original experiment was brought to full fruition when smallpox was eradicated worldwide6. Immunization is one of the most stunning and economically effective contributions of biomedical science to human health. So, immunologists can be proud of the fundamental biomedical insights that have arisen from the field, and of the practical application of these insights in the prevention of disease.

Advances of the last century allow us to better understand the successes (and failures) of past vaccines, and enable a more rational and diverse approach to new vaccine development. An example is the development of polysaccharide–protein conjugate vaccines against Haemophilus influenzae type b. These arose from observations showing: that antibodies to type b capsular polysaccharide are protective; that polysaccharides do not induce T-cell help and are not immunogenic in early life; and that linkage of polysaccharide to protein results in a T-cell-dependent antibody response to both components. Routine use of these vaccines has nearly eliminated meningitis and other diseases caused by H. influenzae type b6. Advances in our understanding of the determinants of protective immunity and immunological memory, of the mechanisms by which adjuvants affect the quality and magnitude of immunological responses, and of microbial genomics, offer the promise for new and more effective vaccines in the near future.

Public concerns and vaccine safety

In recent years, a vocal minority in the developed world has questioned the safety and net benefits of vaccines. Vaccines are unique among medical interventions in that they are given to healthy individuals to prevent diseases that often do not pose an immediate threat to the recipient. Many vaccine-preventable diseases are now so infrequent that the only context in which many individuals have heard of these diseases is when hypothetical adverse effects of the relevant vaccine are presented by the media as fact in an emotionally gripping story. We illustrate, through examples of real and falsely attributed adverse reactions to vaccines, the effects on vaccine use and development.

Whooping cough. Whooping cough vaccines were developed in the late 1940s, by formalin inactivation of whole Bordetella pertussis, and were later combined with diphtheria and tetanus toxoids to create the DTP vaccine. Widespread early childhood immunization led to a marked reduction in the incidence of whooping cough. Because encephalopathy and encephalitis were well-recognized complications of smallpox vaccine7, encephalopathy presenting soon after immunization with a pertussis-containing vaccine was attributed to the vaccine8. As awareness of the severity of this disease faded, public concern emerged regarding potential adverse reactions to the vaccine, including seizures, infantile spasms, encephalopathy and sudden infant death syndrome (SIDS)9,10. As a result, DTP vaccine usage declined in several countries, which was followed by a resurgence of the disease11 (Fig. 2). It was later determined that DTP does not cause SIDS12,13, infantile spasms14 or epilepsy15. In some children, DTP does cause transient fever, hypotonic-hyporesponsive episodes, protracted inconsolable crying and seizures, and might rarely cause acute encephalopathy (<1:100,000 attributable risk)16, but there is no clear evidence that any of these acute events are followed by chronic neurological disability17,18. The negative publicity that precipitated the decline in vaccine usage and resurgence of pertussis-related disease also motivated research to improve our understanding of the antigens that induce protective immunity. In the early 1980s, Japanese investigators developed acellular pertussis vaccines composed of one or more protein antigens, which are now available in combination with tetanus and diphtheria toxoids as DTaP. These vaccines have much lower rates of adverse effects and have largely replaced formalin-inactivated pertussis vaccines in developed nations, but owing to their greater cost, not in developing nations.

Figure 2: The effects of adverse public perceptions of whole-cell pertussis vaccine on pertussis disease.
figure 2

The left panel shows the incidence of pertussis disease in countries with sustained vaccine use. The right panel shows the incidence of pertussis disease in countries in which vaccine use declined in relation to public anti-vaccine movements; the timing of the anti-vaccine movements is highlighted (in yellow), and for England and Wales, vaccine coverage during the period is shown in the inset. Adapted with permission from Ref. 11.

Poliomyelitis. Decades before the development of molecular methods would allow the engineering of live-attenuated viral vaccines, an attenuated poliovirus vaccine for oral administration (oral polio vaccine, or OPV) was developed by empirical, sequential passage of virulent viruses in heterologous cell-culture systems. Shortly after its introduction, sporadic cases of vaccine-associated paralytic poliomyelitis (VAPP) were noted (1:750,000 recipients)19. The continued occurrence of VAPP following the virtual elimination of wild-type poliomyelitis from the Western Hemisphere in 1991 prompted a change from OPV to inactivated poliovirus vaccine (Box 1) in the United States, as was already the practice in many countries of the developed world. OPV remains the vaccine of choice in the drive to eradicate poliomyelitis worldwide, and has worked successfully in regions where vaccine coverage is high. In two regions where OPV coverage was low, paralytic polio due to transmission of vaccine-derived polio viruses has recently been described, perhaps reflecting the accumulation of genetic changes in vaccine-strain virus sufficient to restore neurovirulence and transmissibility following sequential passage between susceptible individuals20.

Curiously, these rare, but real, adverse effects of OPV have not received media attention comparable with that prompted by a false adverse effect. In 1992, an article in Rolling Stone magazine suggested that a candidate poliovirus vaccine derived from virus cultured in kidney cells from African green monkeys (which are known to carry simian immunodeficiency virus; SIVAGM) and tested in the Congo was the source from which human immunodeficiency virus (HIV) arose21. Because SIVAGM is genetically distant from HIV-1, the hypothesis proposed in Rolling Stone was implausible. In 1999, a book entitled The River: A Journey to the Source of HIV and AIDS22 renewed public concern by proposing that HIV was introduced into the Congo by poliomyelitis vaccine prepared in chimpanzee kidney cells contaminated by a simian immunodeficiency virus (SIVCPZ) that evolved into HIV-1. This hypothesis was plausible, as SIVCPZ is closely related to HIV-1; however, several groups have shown conclusively that the vaccines used in the Congo contained no chimpanzee DNA21. These and other data have closed the door on the OPV–AIDS hypothesis in scientific circles, but not all public sceptics have been converted.

Current vaccine controversies

Whether real, alleged but unproven, or incorrect, a perceived risk of a vaccine might outweigh concerns about the disease it is designed to prevent. Perceptions, be they true or false, drive behaviour. Although this might seem to be scientifically irrational behaviour, several truths contribute to such decisions: the safety of biological products is not absolute; rare adverse effects of vaccines might not be apparent in pre-licensing studies; public (and scientific) concerns about adverse effects have contributed to the development of safer vaccines; and if you or someone dear to you is affected, it matters little to you that the adverse event is rare.

In the past 15 years, vaccines against H. influenzae type b, hepatitis A and B, rotavirus (Box 2), Streptococcus pneumoniae, varicella, meningococcus C and Lyme disease have been introduced. In the United States, children are now required to receive 23 (or more) vaccine doses by the age of six. These vaccines have led to a marked reduction in the incidence of many of these diseases. At the same time, there has been a striking increase in the apparent prevalence of some chronic and neurodevelopmental disorders, including attention-deficit hyperactivity disorder (ADHD), autism, allergy and autoimmune diseases23,24,25,26. The temporal association between these events has led some to propose that the vaccines are causally related to an increase in the incidence of these disorders.

Reports by Wakefield and colleagues27,28 in the United Kingdom indicated that, following the administration of the measles, mumps and rubella (MMR) virus vaccine, vaccine-strain measles virus persists in the gut of certain at-risk children, and, through a complex hypothetical pathway, contributes to the development of autism. This hypothesis received widespread media attention in the United Kingdom and Ireland, and was followed by substantial public concern and decreased use of MMR in some regions29,30,31. This concern has crossed the Atlantic, prompting media concern and congressional hearings in the United States. Studies and expert panels in the United Kingdom and United States32,33,34 concluded that the apparent increase in the rate of autism is not causally related to MMR vaccine, although data are, at present, inadequate to fully exclude the possibility that MMR might be a contributing factor in a rare child34. In France, concern has revolved around the potential induction of multiple sclerosis by hepatitis B vaccination, prompting its withdrawal from routine use in some groups, and widespread public concern (Box 3).

The increased prevalence of type I diabetes and asthma in many countries in the developed world has been linked by some to the prevention of infectious diseases and microbial contact through better hygiene and vaccination23,25,35,36,37. There is some evidence that children with more frequent infections in early life have a lower rate of allergic diseases, including asthma, in later life23,25,35,36,37, but the epidemiological evidence is conflicting38,39. The hypothesis, which states that frequent infections protect against atopic disease, fits with the notion that microbes trigger the production of cytokines by cells of the innate immune system, for example, interleukin-12 (IL-12) and type 1 interferons, which favour the development of T helper (TH)1 T-cell-mediated responses. Some propose that, in the absence of such signals, the infant's immune response to non-microbial environmental antigens, particularly in children with a genetic predisposition, is more likely to deviate to an allergic, TH2 response23,36,37. Because infection is known to decrease the risk for type I diabetes in genetically prone rat and mouse models26, if these principles apply in at-risk humans, reduced contact with infectious agents might increase their risk. However, even if this ‘hygiene hypothesis’ has merit, the contribution of vaccine-induced prevention is likely to be minor26,40, as relatively few infections are prevented by immunization.

Non-antigen-specific effects have also been proposed. Many vaccines contain alum, an adjuvant that favours the development of TH2 responses41. Whether or not the frequency and amounts of alum administered are sufficient to influence the nature of ongoing immune responses to intercurrent infections, or to do so in a subset of children with a genetic predisposition to allergic disease, is uncertain. Comparative data from West Africa showed that immunization with DTP vaccine, which contains alum, increased the risk of allergy in children, whereas immunization with BCG (Bacillus Calmette–Guérin) decreased this risk and overall mortality35. However, potential confounding variables might account for these differences. Finally, almost 25% of US parents are concerned that the number of vaccines administered ‘overload’ the capacity of the infant's immature immune system, perhaps impairing immunity to other infections or altering the body's tolerance to self-antigens42 — a concern for which there is no clear scientific basis. Concerns have also been raised about the potential for thimerosal43, which was included in a number of vaccine formulations as a preservative, to contribute to the development of neurodevelopmental disorders, including autism. The active ingredient in thimerosal is ethylmercury, and mercury is a known neurotoxin44. The increased number of routine childhood immunizations in the United States resulted in an increase in the total amount of ethylmercury received by young children, at the same time that the apparent incidence of autism has increased. Although the amounts of ethylmercury received are below those shown to cause neurotoxicity, theoretical concerns led to the removal of thimerosal from vaccines manufactured in the United States44. Further study is needed to determine the risk, if any, of these doses of thimerosal43. Thimerosal has not been removed from vaccine formulations provided by the World Health Organization, owing to concerns that the resultant increase in cost would limit vaccine availability. Similarly, the presence of bovine serum in vaccine preparations has raised theoretical concerns about the transmission of bovine spongiform encephalopathy (BSE); although unlikely, a finite (<1:109) risk cannot be excluded.

At the same time that public trust in vaccines is threatened, the impetus for the development of new vaccines is great. Earlier predictions that infectious diseases would cease to be an important threat to human health have been followed by a more sobering reality. Infection currently accounts for 25% of all deaths worldwide, new infectious diseases have emerged as global threats, and there has been a resurgence of many diseases for which we lack effective vaccines45. We, in the developed world, are no longer safe from infections that plague the developing world, as shown by the HIV epidemic. The pressure to bring prototype HIV vaccines to early clinical trials, particularly in Africa, is great46,47. This has led some public figures to dismiss scientific concerns that hypothetical negative effects of a candidate vaccine on protection might go unnoticed in initial clinical trials47. Some in the developed world have used such issues as fuel for their own concerns about vaccine safety. They ask whether the enthusiasm for the triumphs achieved through vaccination will lead to the introduction of new vaccines, including vaccines against HIV, the full safety of which will not be known, and that they and their loved ones will be the unwitting experimental subjects in whom adverse effects will first become apparent.

Call to action

What must we do to allay these fears and to restore and sustain the public's trust? First, advances in immunology must be exploited to develop more effective, safer vaccines, particularly for diseases of global importance. We must carefully assess the safety and efficacy of each vaccine, when administered alone or in combination with other vaccines. We must ask if vaccines alter the risk of allergic and autoimmune diseases, and understand genetic factors that influence immunogenicity or predispose individuals to adverse effects. Finally, immunologists, along with scientists from other fields, public health officials and physicians, must participate in a frank public dialogue with those to whom vaccines will be given, stating clearly what we know for certain, what we suspect, but for which we lack compelling proof, and what we do not know48,49. The debate about vaccine safety takes place in three courts: the court of medicine and science, the court of public opinion and the legal courtrooms. The rules of evidence — what constitutes proof of causation — differ in these venues, and therefore so might the judgements rendered. Science must be the final arbiter, but we must explain the science more clearly and effectively to a public whose scepticism reflects their struggle to understand information from diverse sources of varying validity, including the web (see the web sites listed in the links box below). Both our individual credibility and the credibility — and therefore the potential — of the entire collaborative enterprise of immunization depend ultimately on both the rigour of our scientific practice and our ability to communicate this to the public.