LONDON, (Reuters) - It's a showpiece drug that has the potential to end a disease that kills half a million African children a year. Yet even before it wins a license, the world's first malaria vaccine has lost some of its sheen.

Backed by billionaire philanthropist Bill Gates and developed by GlaxoSmithKline, the vaccine -- called RTS,S or Mosquirix -- is being assessed by regulators and global health authorities.

Granting it a license and recommending it for rollout in sub-Saharan Africa, where malaria kills one child almost every minute, ought to be a no-brainer.

But Mosquirix is hampered by caveats, complexities and cost implications that threaten to make its arrival on the global health stage more of a problem than a solution, possibly not just for malaria but for vaccines in general.

"There's a lot of excitement for a malaria vaccine. But it's a very complicated vaccine, so the recommendation is presumably going to be complicated too," says Seth Berkley, chief executive of the GAVI global vaccine group.

Malaria is caused by a parasite carried in the saliva of mosquitoes. GSK's vaccine is designed to go to work at the point the parasite enters the human bloodstream after a mosquito bite.

By stimulating an immune response, it can prevent the parasite from multiplying in the liver. Without that response, the parasite re-enters the bloodstream and infects red blood cells, leading to fever, body aches and sometimes death.

One big problem with Mosquirix is that while it's the best malaria vaccine so far, it still doesn't work very well.

Unlike polio or smallpox vaccines, which offer life-long high-level protection from the diseases they are designed to prevent, Mosquirix gives only partial protection against malaria, and even that dwindles within a few years.

Data from clinical trials which ran across seven countries in Africa show that at best, in children aged 5-17 months, it offers 50 percent protection. In babies aged around 3 months, that drops to 30 percent.

Dosing is also a problem. Pedro Alonso, director of the World Health Organization's Global Malaria Program, explained in a briefing last week that even to get that efficacy, children would need for four doses over 18 months.

"In the absence of four doses, the efficacy disappears and no significant protection is documented," he said. "It's challenging in terms of understanding how it would best add value."

The WHO has promised to make a decision on whether and how to recommend use of Mosquirix by the end of 2015.

European Medicines Agency drugs regulators, who have been wading through a quarter of a million pages of evidence submitted by GSK, are expected to decide soon, likely later this month, on whether it should be licensed.

COSTS, AND DELIVERY

Sources close to the approval process suggest Mosquirix is likely to get both EMA and WHO backing, partly thanks to the weight of history pushing for the world's first malaria vaccine to get to market.

But there will be conditions attached, all of which carry cost implications that may make the reality of delivering Mosquirix prohibitively expensive.

GSK hasn't yet put a price on the vaccine, but the firm's charismatic chief executive Andrew Witty has promised it won't be expensive, with a profit margin of 5 percent over cost of manufacture which he promises to reinvest in research on malaria and other neglected diseases.

Sources involved in planning for Mosquirix's potential future use told Reuters they have been advised to work with a price tag of around $5 per dose.

That would make a dose of Mosquirix about the same as the cost of an insecticide-treated bednet. And while a bednet can protect two people for three years before it needs replacing, with Mosquirix, the likelihood is that one child would need four doses -- around $20 in medicine costs alone -- to get an extra 30 percent protection from malaria for a shorter time.

Adrian Hill, a vaccine expert at Oxford University's Jenner Institute, notes the numbers look even less attractive if Mosquirix is recommended for slightly older babies.

If the shot were to be licensed for babies of 3 months old, when it could be given as part of the routine so-called Expanded Program on Immunization, the cost of delivery could be relatively low.

"The real problem arises if it's given at 6-12 months. The cost of delivery will be greatly increased and we just don't know how feasible extra immunization time points will be," Hill said.

"Sadly, the data show the protection in younger infants, who we immunize routinely, is very modest, but it is better in older infants."

REPUTATIONAL RISK

As the world first human vaccine against a parasitic disease, Mosquirix is a historic milestone and close to the hearts of its key backers, the philanthropic Bill & Melinda Gates Foundation and GSK.

If global health authorities push ahead despite the complexities, experts say there could be substantial reputational risks for Gates, Witty, the WHO and even for vaccines in general.

After all, if children vaccinated against malaria continue to get the disease, why should mothers trust other vaccines, developed by drugmakers, backed, promoted and recommended by the WHO, against diseases like pneumonia, measles and polio?

The Gates Foundation is keen not to voice an opinion on Mosquirix at this sensitive time, when regulators and the WHO are assessing it, said foundation director Alan Magill.

He stressed this is a first generation vaccine, "no silver bullet", and only one of a range of weapons against malaria.

A GSK spokesperson also emphasized Mosquirix is designed for use alongside other malaria control measures such as bednets. "Given the huge burden of malaria in sub-Saharan Africa, we believe this could have a significant public health impact," she said.

Despite widespread acknowledgment of the complexities and costs surrounding Mosquirix, none of the experts Reuters spoke to believes regulators and the WHO will say no to the world's first malaria vaccine.

"It's going to be a first time in history situation," said Alonso. "The challenges are around where this vaccine could provide additional benefits to the tools we already have."

(Reporting by Kate Kelland; Additional reporting by Tom Miles in Geneva; Editing by Sonya Hepinstall)