Television cameras may be few and far between in rural areas of Bangladesh, south Asia's poorest nation. But killer human viruses are recurrent there, and are quietly wreaking havoc.

An outbreak of the emerging Nipah virus in the Faridpur district of the country earlier this month infected 30 people and killed 18. That epidemic is one of several that have hit Bangladesh, Malaysia and Singapore since the virus was first discovered in 1998.

The Nipah virus and the related Australian Hendra virus form the Henipavirus genus within the paramyxovirus family. They cause high mortality — two in every five infected people die — but pose less of a global risk than more notorious viruses such as SARS (severe acute respiratory syndrome). That is because infection requires close contact with the animal host, the Pteropus fruit bat, and does not seem to pass from human to human.

The virus was discovered in Malaysia, during a 1998 outbreak in the village of Nipah that killed 105 people. The transmission route of the virus in Bangladesh seems different from outbreaks elsewhere, however, says Marie-Claude Georges-Courbot, an expert on the disease at the Pasteur Institute in Lyon.

The exact mechanism of transmission is poorly understood, says Georges-Courbot. The Malaysian epidemic arose in humans following an outbreak in pigs that had come into contact with bat urine. But the outbreaks in Bangladesh — one in February killed 17 people — appear to have been caused by children who had direct contact with bat-contaminated fruit, she says.

The symptoms of the disease in Bangladesh are also different, being largely neurological; the Malaysian outbreak caused pulmonary complications.

Georges-Courbot is part of a team of French and Malaysian researchers who earlier this year reported that golden hamsters could be protected from Nipah by vaccinia viruses. The team used vaccinia to express two proteins that the Nipah virus uses to enter human cells. The research also showed that serum from vaccinated hamsters prevented onset of the disease in others. This suggests the possibility of an immunotherapeutic approach to treating Nipah, and perhaps its Hendra cousin.