* Early signs pipelines improving -Deloitte, Thomson Reuters
* Drug approvals increasing but sales value lower
* R&D returns at top-12 firms 7.2 pct vs 7.7 pct year ago
* Average cost of developing a new drug steady at $1.1 bln
LONDON, Dec 4 (Reuters) - Drug companies are becoming more efficient in hunting for new treatments, though this has yet to be reflected in improved investment returns, according to a report on Tuesday.
Low productivity in research labs is the biggest single challenge facing the global pharmaceutical industry, which is struggling to replenish its medicine chest after a wave of patient expiries that peaked this year.
While companies are getting more compounds into late-stage development - reflecting a smarter and more targeted approach to research and development (R&D) - turning those new products into big commercial winners is an uphill struggle.
That reflects growing caution among governments about buying costly new drugs, as well as the arrival of more specialist products that address relatively small patient populations.
The latest annual study of R&D productivity by Deloitte and Thomson Reuters found that the number of new drug approvals increased by around 30 percent, yet the expected revenue from these medicines actually fell by a similar amount.
In total, the world's 12 top pharmaceutical companies had 41 new drugs approved, with combined forecast revenues of $211 billion, while the year-earlier tally was 32 products with expected revenues of $309 billion.
In effect, the industry is treading water in the fight to deliver better returns on the billions of dollars ploughed into the hunt for new drugs each year.
With an average internal rate of return (IRR) from R&D in 2012 of 7.2 percent - against 7.7 percent and 10.5 percent in the two preceding years - Big Pharma is barely covering its average cost of capital, estimated at around 7 percent.
Nonetheless, there are some encouraging signs. In particular, 10 of the 12 companies tracked in the report showed an improvement in replenishing their stock of late-stage experimental drugs.
"We've seen returns stabilising and there are signs on the horizon that the situation might turn around, depending on how successful the industry is at commercialising new assets as they come through," said Julian Remnant, head of Deloitte's European R&D advisory practice.
At $1.1 billion, the average cost of developing a new medicine has remained fairly constant, although it varies hugely between companies, since this figure includes money spent on drugs that ultimately fail.
For the most successful company in the group, the average cost of developing a drug was just $315 million, while at the other extreme one firm spent $2.8 billion.
The companies analysed in the study were Pfizer, Roche, Novartis, Sanofi, GlaxoSmithKline, Johnson & Johnson, AstraZeneca , Merck & Co, Eli Lilly, Bristol-Myers Squibb, Takeda and Amgen.
The study calculated IRRs for these companies by estimating the future value of sales from products in final-stage Phase III clinical trials, or those submitted for regulatory approval, using standard industry benchmarks for success rates. (Editing by Louise Heavens)