Worldwide, some 600 000 to 800 000 infants die yearly from severe diarrhoea caused by a rotavirus. In the USA, each year, about 50 000 infants are admitted to hospital with rotavirus infection, and of these about 20 die. These figures make a clear case for the need for a safe and effective vaccine against rotavirus infection. Such a vaccine (Rotashield) was developed and licensed for use in the USA in 1998. About one and a half million doses were given until nine and a half months later the US Advisory Committee on Immunisation Practices advised that the use of the vaccine be suspended because of reports that the vaccine was associated with an increased incidence of intussusception in infants. Intussusception is a condition where the lining of the bowel folds in on itself and causes a blockage often needing an operation to clear the blockage.
By no means all researchers agreed on the overall magnitude of intussusception risk among infants exposed to Rotashield. Lone Simonsen and colleagues from Bethesda, MD, USA studied hospital admissions of infants for intussusception in ten US states during the period when Rotashield was in use, and an estimated 28% of the infants (304,347) received the vaccine. They report in this week’s issue of THE LANCET, “We observed an overall decrease of 4% to 10% in infant intussusception admissions during the Rotashield use period. This result was unexpected, since other studies have documented a strong association in the first post-vaccination week and had predicted significant overall increases in intussusception rates. Thus, we are facing an apparent paradox”. The authors favour the explanation that Rotashield may trigger intussusception events in susceptible infants and that this precipitation may be compensated for by a lower risk of intussusception later in infancy.
After the initial reports of apparent adverse reactions to Rotashield, the manufacturers stopped producing the vaccine. “As of March 1st, 2001, Rotashield remains licensed but unavailable, despite benefit/risk ratios presumed to weigh heavily in favour of a rotavirus vaccine in developing countries with a heavy burden of rotavirus disease mortality.”
The authors comment that the obvious strength of this study is that it documents what actually took place in terms of intussusception rates in a large subset of the US infant population when Rotashield was used. However, the main weakness is our inability to directly link intussusception and vaccination records for individual infants. In a commentary on page 1197, Andrew J Hall from the London School of Hygiene and Tropical Medicine, London, UK, agrees, “one of the weaknesses of Simonson and co-workers’ study is the lack of confirmation of the diagnosis of intussusception, which could have been done by examining medical records. Nor were vaccine records obtained, which could have yielded information about potential confounders”
Contacts : Dr Lone Simonsen, 67-B Rockledge Drive, Room 3153, Bethesda, MD 208-7630, USA. T) +1 301 402 8487; F) +1 301 402 3255; E) Lsimonsen@niaid.gov
Dr Andrew J Hall, Infectious Disease and Epidemiology Unit, London School of Hygeine and Tropical Medicine, London WC1E 7HT, UK; E) andy.hall@lshtm.ac.uk
Journal
The Lancet