The result leaves uncertain the future of a product that Sanofi has previously said could generate more than 1 billion euros ($1.3 billion) in yearly sales.

Overall efficacy was 30.2 percent - statistically insignificant and a far cry from the 70 percent-plus protection that researchers had anticipated at the start of the company-funded trial.

The poor outcome was down to the failure of the vaccine to protect against one type of dengue virus, which turned out to be the prevalent one in Thailand at the time of the study.

The mosquito-borne disease, also known as "breakbone fever", is a threat to nearly 3 billion people and is caused by four different types of virus, none of which confers immunity from the others.

"This result knocked me off my chair," Scott Halstead, a senior scientific adviser for the non-profit Dengue Vaccine Initiative, who was not involved in the study, told Reuters.

"This is a very sobering outcome and a lot of thought and further experimentation is going to have to be done in order to understand what happened.

"You can't be very confident that this vaccine is going to significantly reduce human illness."

Hopes had been high for the closely watched Thai trial, given the fact that dengue belongs to the same virus class as yellow fever and Japanese encephalitis, both of which are controlled with highly effective existing vaccines.

But it seems that making a mixed dengue vaccine containing four different virus strains can produce uneven results.

WAITING FOR PHASE III

Sanofi had already disclosed in July that the Thai study showed efficacy against three of the four strains, without giving details. The full findings were published online on Monday night in the The Lancet medical journal.

Despite the disappointment, Jean Lang, Sanofi's head of dengue vaccine development, said the Phase IIb study involving 4,002 Thai children showed that a safe dengue shot was possible and protection was encouraging against three of the strains.

Efficacy was around 60 percent against dengue virus type 1 and 80-90 percent against types 3 and 4. Interestingly, a single dose of the vaccine proved roughly as good as three doses.

"This is a milestone, but we need to wait for the results of two large Phase III efficacy trials to have a better understanding of the vaccine," Lang said.

Sanofi is currently conducting final-stage Phase III trials with 31,000 participants in Asia and Latin America.

If the Thai trial results had been strong enough, it could have filed for early marketing approval in some countries. Instead, attention will now focus on results for the ongoing Phase III trials, with any commercial launch in 2015 or later.

Other drug companies are also working on dengue vaccines but Sanofi's product is several years ahead. In a bid to stay in the lead, the French group has already invested 350 million euros in a new French factory to make its vaccine.

MATHEMATICAL MODELS

Halstead, who wrote an accompanying commentary on the study in The Lancet, said it was possible that a vaccine against three out or four virus strains might still combat dengue effectively, since severe disease is caused by a second infection.

He suggested mathematical models could be used as a first step to work out the impact.

Dengue fever, which can cause intense joint and muscle pain, is spread by the bite of the Aedes aegypti mosquito. The insect thrives in the mega-cities of the tropics.

In the past 50 years there has been a 30-fold jump in dengue cases. The World Health Organization officially puts infections at 50-100 million a year, though many experts think this assessment from the 1990s badly underestimates the disease.

Most patients survive dengue, but it is estimated to kill about 20,000 every year, many of them children, who are less able to fight it off.

($1 = 0.7812 euros)

(Editing by Hugh Lawson)

• A Permanent Economic Emergency