The threat of an avian influenza pandemic.

There have been three influenza pandemics during the past century — in 1918, 1957, and 1968. Although the severity of the epidemics and the primary age groups affected varied, each was caused by a novel type A virus of avian origin. In 1957 and 1968, the new viruses had components of previous human viruses as well as avian viruses. The genome of the influenzavirus is made up of eight segments of RNA, and it was determined retrospectively that in both cases, there had been a reassortment of avian and human genes — most likely the result of the coinfection of a host by two different viruses. The origin of the lethal 1918 virus appears to have been different: recent data suggest that it resulted not from reassortment, but from the mutation of the genes of what was originally a purely avian virus. Although attention has recently been directed to the highly pathogenic avian influenzaviruses, most avian strains are of low pathogenicity. Ordinary avian influenzaviruses vary greatly, owing to their 15 hemagglutinins and 9 neuraminidases, and are widespread in migratory birds and water fowl. It was originally thought that these avian viruses could not directly infect humans, because humans have receptors for human viruses, and birds for avian viruses. Therefore, it was hypothesized that an intermediate host was required in which coinfection and reassortment could take place and that this host was the pig, which possessed receptors for both kinds of viruses. The highly pathogenic avian influenzaviruses have appeared only occasionally, causing economic losses in the form of a reduced supply of poultry; the current widespread dissemination of these avian influenzaviruses is unprecedented. Also unprecedented is disease caused by the spread of these highly pathogenic viruses from chickens directly to humans. The type A (H5N1) viruses spread in this way in Hong Kong in 1997, resulting in six deaths among 18 people with documented clinical cases; a smaller number of cases also occurred in this manner in Hong Kong and Fujian Province, China, in early 2003, and at least one death resulted. In addition, confusingly enough, there was a massive outbreak of another highly pathogenic avian influenzavirus — type A (H7N7) — in the Netherlands in 2003, in which bird-to-human transmission resulted in the death of an infected person. The antiviral drug oseltamivir was used extensively to help control and limit the spread. However, it is the A (H5N1) virus that represents an increasing global concern. Changes in the virus in 2003 resulted in the generation of what is termed the Z strain, which spread to at least nine countries in East Asia and Southeast Asia. 1 This strain was characterized by pathogenicity in a larger number of animal species than are affected by other strains and by resistance to the older class of antiviral drugs represented by amantadine and rimantadine. Bird-to-human transmission has continued, and it has now been documented as the cause of 32 deaths in 44 patients with confirmed cases. Although the direct transmission of the virus from birds to humans is itself cause for concern, the greatest worry is that human-to-human transmission may begin to occur if there is a change in the viral genome. Even without genetic change, some inefficient spread among humans may now be expected. Such events occurred in 1997 in Hong Kong as well as during the outbreak in the Netherlands, but only asymptomatic infection caused by such transmission was documented. The report by Ungchusak et al. in this issue of the Journal (pages 333–340), however, clearly suggests the occurrence of person-to-person transmission through close contact with an infected child, leading to both clinDr. Monto is a professor of epidemiology at the University of Michigan School of Public Health, Ann Arbor.