Non-parenteral vaccines

Vaccination is potentially the most cost effective method of controlling infections.¹ However, as the quests for safe and effective anti-HIV and anti-malarial vaccines show, successful development of vaccines is not easy. When HIV was identified, optimists thought that a vaccine would soon follow, perhaps within two years. Two decades later, that goal seems to be as elusive as ever. Moreover, worldwide eradication of a disease—although possible, as shown with smallpox—presents logistic and economic challenges. People who need the vaccines are most often least able to pay for them or use them properly, because of educational barriers or woefully inadequate healthcare systems.

Most vaccines are administered by injection. Effective non-parenteral vaccines would be more convenient and potentially cheaper to produce and administer. Why then are so few such vaccines available? For example, in the United Kingdom, despite intense research, only the Sabin oral polio vaccine is in general use. In a survey of the field nine years ago I was optimistic and foolhardy enough to believe that soon many of the parenteral vaccines would be administered via alternative routes.²

With better understanding of the immune system, notably at mucosal portals of entry, recognition of the common mucosal immune system, rapid development of genetic engineering techniques, and the human genome project, increasing political pressure for methods to control the HIV, and the excitement of the researchers involved, one could easily be mesmerised. Yet, the debates about the relative merits of the oral and parenteral polio vaccines should have tempered our excitement.³,⁴ For non-parenteral vaccines to work, the antigen or vaccine has to breach the defensive mucosal barriers, rich in proteolytic enzymes and resistant to the passive diffusion of proteins. This can be done by identifying active carrier systems or, more practically, by damaging the barrier membranes through use of either erosive formulation additives or invasive attenuated bacteria or viruses. Both approaches are potentially toxic. Mutation of attenuated micro-organisms back to pathogenic variants, including the wild type, may have devastating effects: outbreaks of iatrogenic poliomyelitis associated with the use of the oral polio vaccine, for example, are well known.³

Recognition that traces of novel adjuvants, notably enterotoxins (cholera enterotoxin (CT) and heat labile enterotoxin (LT)) produced by Vibrio cholerae and Escherichia coli, respectively, improved the immune response to mucosal vaccines to an extent that implied clinical effectiveness, created considerable optimism.⁵,⁶ This optimism was partly justified as two non-parenteral vaccines, an oral vaccine against the rotavirus and an intranasal vaccine, with heat labile Escherichia coli enterotoxin as adjuvant against influenza were subsequently marketed.⁷,⁸ Both vaccines, however, were withdrawn soon after as a result of adverse effects—Bell's palsy in the case of the anti-influenza vaccine⁷ and intussusception in the case of the anti-diarrhoeal vaccine.⁸

A safety study of the Swiss influenza vaccine in 2000 identified Bell's palsy as a possible adverse effect (9 in 3600 subjects studied). In the early postmarketing stage (October 2000 to March 2001), with about 90 000 vaccinated subjects, five cases were reported spontaneously.⁹ By April 2001, this had increased to 46. A subsequent case-control study estimated that in contrast to the parenteral vaccine, the intranasal formulation increased the risk of Bell's palsy considerably (adjusted odds ratio 84, 95% confidence interval 20 to 352).⁷ It was estimated conservatively that 13 excess cases occurred per 10 000 subjects vaccinated. These figures show how inefficient spontaneous adverse reporting is for identifying even specific and severe adverse reactions.

Five cases of intussusception were seen among 10 054 infants who received the rotavirus vaccine, compared with one among the 4633 subjects in the placebo group.¹⁰ In a case-control study, the incidence rate ratio was estimated at 29 (95% confidence interval 16-54) within two weeks of the first dose. The second dose increased the risk more modestly. Overall, there was one case of intussusception attributable to the vaccine for every 4670 to 9474 infants vaccinated.

What lessons can then be learnt from those vaccine withdrawals? Firstly, all vaccines carry with them some risk of adverse effect that needs to be balanced against their potential benefits. Secondly, for diseases for which current safe alternatives are available, stricter validation should be required before marketing of alternative formulations, particularly when premarketing studies imply a substantial risk. Thirdly, postmarketing surveillance should be intense, given the limited premarketing testing of new vaccines. This could take the form of a mandatory register of all recipients until a sufficient number of vaccinated people has accrued to identify potentially serious adverse effects. Low reporting rates make spontaneous reporting inadequate. Manufacturers need to be asked to undertake more controlled monitoring as part of the grant of marketing authorisations. Access to case notes also needs to be improved. In the Swiss study the investigators were able to enrol only about a third of the identified cases.⁷ The crisis surrounding the vaccine for measles, mumps, and rubella (MMR) shows that consumers are intensely averse to risk, and the challenge for regulators and manufacturers is to restore confidence in vaccination. Both false alarms and postmarketing withdrawals of the vaccine make this more difficult. Fourthly, more research is required on the mechanisms by which poorly understood adverse effects of non-parenteral vaccines, such as Bell's palsy and intussusception, develop. This will require funding, probably by charities and government agencies.

Given increasing microbial resistance to antibiotics, research on vaccines needs to be increased. The once intense debate initiated over four decades ago about the relative merits of the oral and parenteral polio vaccines continues.⁴ A decade ago, Howson and Fineberg discussed the ricochet of magic bullets in their commentary on the adverse effects of pertussis and rubella vaccines.¹⁰ We have yet to achieve Paul Erhlich's dream of the magic bullet.¹¹ We have an intense dislike of friendly fire, particularly resulting from pre-emptive action about dangers that we cannot perceive clearly. Tradeoffs between risks and benefits will always be necessary.¹² Ironically, with concerns about bioterrorism, research into smallpox vaccination is restarting. Any collateral benefits from such research are to be welcome.