An experimental malaria vaccine once thought promising is turning out to be a disappointment, with a new study showing it is only about 30 per cent effective at protecting infants from the killer disease.

That is a significant drop from a study last year done in slightly older children, which suggested the vaccine cut the malaria risk by about half — though that is still far below the protection provided from most vaccines. According to details released on Friday, the three-shot regimen reduced malaria cases by about 30 percent in infants aged 6 to 12 weeks, the target age for immunization.

It's probably not worth implementing a malaria vaccine that's 30 per cent effective in infants, an expert says. (Schalk van Zuydam/Associated Press)

Dr. Jennifer Cohn, a medical coordinator at Doctors Without Borders, described the vaccine's protection levels as "unacceptably low." She was not linked to the study.

Scientists have been working for decades to develop a malaria vaccine, a complicated endeavor since the disease is caused by five different species of parasites. There has never been an effective vaccine against a parasite. Worldwide, there are several dozen malaria vaccine candidates being researched.

In 2006, a group of experts led by the World Health Organization said a malaria vaccine should cut the risk of severe disease and death by at least half and should last longer than one year.

Malaria is spread by mosquitoes and kills more than 650,000 people every year, mostly young children and pregnant women in Africa. Without a vaccine, officials have focused on distributing insecticide-treated bed nets, spraying homes with pesticides and ensuring access to good medicines.

In the new study, scientists found babies who got three doses of the vaccine had about 30 per cent fewer cases of malaria than those who didn't get immunized. The research included more than 6,500 infants in Africa. Experts also found the vaccine reduced the amount of severe malaria by about 26 per cent, up to 14 months after the babies were immunized.

Scientists said they needed to analyze the data further to understand why the vaccine may be working differently in different regions. For example, babies born in areas with high levels of malaria might inherit some antibodies from their mothers which could interfere with any vaccination.

"Maybe we should be thinking of a first-generation vaccine that is targeted only for certain children," said Dr. Salim Abdulla of the Ifakara Health Institute in Tanzania, one of the study investigators.

Results were presented at a conference in South Africa on Friday and released online by the New England Journal of Medicine. The study is scheduled to continue until 2014 and is being paid for by GlaxoSmithKline and the PATH Malaria Vaccine Initiative.

Vaccine cost and effectiveness questioned

"The results look bad now, but they will probably be worse later," said Adrian Hill of Oxford University, who is developing a competing malaria vaccine. He noted the study showed the Glaxo vaccine lost its potency after several months. Hill said the vaccine might be a hard sell, compared to other vaccines like those for meningitis and pneumococcal disease — which are both effective and cheap.

"If it turns out to have a clear 30 per cent efficacy, it is probably not worth it to implement this in Africa on a large scale," said Genton Blaise, a malaria expert at the Swiss Tropical and Public Health Institute in Basel, who also sits on a WHO advisory board.

Eleanor Riley of the London School of Hygiene and Tropical Medicine, said the vaccine might be useful if used together with other strategies, like bed nets. She was involved in an earlier study of the vaccine and had hoped for better results. "We're all a bit frustrated that it has proven so hard to make a malaria vaccine," she said. "The question is how much money are the funders willing to keep throwing at it."

Glaxo first developed the vaccine in 1987 and has invested $300 million in it so far.

WHO said it couldn't comment on the incomplete results and would wait until the trial was finished before drawing any conclusions.