Yellow fever vaccine

Three recently published articles, from Brazil, the United States, and Australia, and three follow up letters from Europe, have challenged the reputation of the yellow fever vaccine, 17D, that for more than 50 years was almost beyond reproach.^1–6 These reports describe an illness resembling yellow fever occurring within a week of vaccination for yellow fever and leading to death within two weeks in six of 10 individuals. These adverse effects of the vaccine probably represent the delayed recognition of a very unusual outcome. The present recommendations must stand until an even safer vaccine is available. Primary 17D vaccination remains the least of several risks that any traveller to the tropics has to take.

Medical history was made in 1927 when Stokes, Bauer, and Hudson infected rhesus monkeys with blood from Asibi, a native of the Gold Coast (now Ghana) with a mild attack of yellow fever.⁷ The Asibi strain proved not to be benign and several deaths in laboratory workers were subsequently attributed to it.⁸ Nevertheless, it was the strain of yellow fever virus that between 1933 and 1937 Theiler, Lloyd, Smith and coworkers attenuated by multiple passage in mouse brain and chick embryo tissue.⁹ This empirical process seemed to remove the viscerotropic and encephalitogenic properties of the Asibi strain, as shown by monkey inoculations, trials in laboratory workers, and subsequent field trials in Brazil, Bolivia, and Colombia.¹⁰

The human trials of this 17D vaccine certainly showed that the 176 or more passages in non-primate tissues had greatly attenuated it, and once problems due to lability of the vaccine and the mistaken use of pooled human serum as a stabiliser¹¹ were ironed out it was hailed as a triumph. No serious attempt was made thereafter to develop an alternative as the 17D vaccine was regarded as safe and effective for at least 10 years, and infancy, pregnancy, egg allergy, and immunosuppression are its only present contraindications.

The current communications describe 10 incidents that mostly resemble classic yellow fever, even though the possibility of intercurrent exposure to wild yellow fever virus was out of the question in five, and very unlikely in the others. Vaccine from at least three different manufacturers was involved, which suggests that any 17D derived product carries the same risk, albeit extremely small, of an acute yellow fever-like illness with high mortality. As concurrent reversion to virulence at several different manufacturing facilities seems unlikely it must be presumed that the capacity to cause this adverse effect was always present in the 17D vaccine seed. The most likely explanation seems to be that an idiosyncratic host susceptibility allows the development of virulence associated mutations in the 17D virus during a prolonged viraemic phase.

There is almost nothing in the reports of the original trials of 17D to suggest a propensity to give rise to a short incubation illness of the sort now described,¹² and it is probably only improved surveillance during the 1990s that has brought this adverse outcome to light, perhaps aided by more primary yellow fever vaccinations of older travellers who might be especially vulnerable.¹³ If so, the sudden rash of reports is probably artefactual and the use of 17D vaccine when and where the need exists should be maintained. There are no substitute vaccines, and the incidence of the complications described, serious though they are, appears to be low, of the order of one in a million. This is so low that no new candidate vaccine could be shown to carry a smaller risk without several years of use. Nevertheless, exploration of the feasibility of a non-live vaccine is now in order, just in case these serious reactions to 17D turn out to be more common than so far seems to be to the case.

At present the risk of yellow fever to travellers to any country where the infection is enzootic or has recently been epidemic continues to exceed the known risk from 17D. Moreover, as unimmunised visitors represent a potential risk to inhabitants they are very likely to be excluded from countries where yellow fever has been reported unless they carry a valid certificate of 17D vaccination. No travellers to an area where yellow fever may occur should therefore set off unvaccinated unless they have a medical certificate stating the contraindication. A possible reservation applies to destinations in east Africa where yellow fever has not been reported and which are at a high enough altitude to be vector free—for example, cities like Nairobi and Addis Ababa. It is worth adding to this risk assessment of 17D that those having repeat 17D vaccination probably have little to fear as they can be expected to make an anamnestic response to 17D that will rapidly suppress viraemia.