Last checked at 
* Standalone strategy hinges on new drugs delivering
* AstraZeneca has had more than fair share of flops in past
* Experiences with Exanta, other failures haunt company
By Ben Hirschler
LONDON, May 20 (Reuters) - AstraZeneca's confidencein its promising new drugs may have seen off Pfizer's $118 billion takeover bid for now, but investors are leftcontemplating the lessons from 2004, when it was also trumpetingits product pipeline.
AstraZeneca became renowned a decade ago for talking up theprospects for its new treatments under former chief executiveTom McKillop, only to see key experimental medicines tumble likeninepins in a series of development setbacks.
By the time McKillop left in 2006 to chair the ill-fatedRoyal Bank of Scotland, its new drug pipeline wasfractured and AstraZeneca had a reputation for dismal researchproductivity that it has only recently started to shake off.
The biggest pipeline wipe-out came in 2004 when Exanta,which could have been the first of a new class of oralanticoagulants to reach the market, was deemed too dangerous byU.S. experts.
Since then AstraZeneca has continued to have more than itsfair share of late-stage drug failures, including products fordiabetes, stroke, depression and rheumatoid arthritis, all ofwhich were once seen as potential multibillion-dollar sellers.
That track record has not been forgotten by investors whoremain sceptical about the claims for a line-up of life-savingnew treatments that have featured prominently in the battle towin hearts and minds in recent weeks.
In its fightback against Pfizer, AstraZeneca's current CEOPascal Soriot made bullish pipeline projections a central plankof his defence, predicting a 75 percent increase in annual salesto $45 billion by 2023.
AstraZeneca Chairman Leif Johansson told Reuters on Mondaythat his firm was now "the most transparent pharmaceuticalcompany in the whole world", following its decision to publishits 10-year planning forecasts on May 6.
The challenge for Soriot and Johansson is to make good onthat promise.
"The problem with putting big sales numbers out there is youhave to live up to them. If anyone believes the industry hassuddenly become less risky, they are misguided," said MarkClark, an analyst at Deutsche Bank.
"It is not just the risk of your own product, there is alsofast-follower risk these days, because rival companies areincreasingly hot on the heels of a new idea."
There is a good reason why, in normal times, drugmakers liketo keep much of their early research under wraps, since they arefearful that promising too much simply sets them up for a fall,as AstraZeneca found out a decade ago.
SPECULATIVE
While investors agree AstraZeneca has some promising newdrugs, especially for cancer, they view the 2023 forecasts ashighly speculative and no sure bet to propel the shares toPfizer's 55 pounds-per-share offer level.
The valuation gap between AstraZeneca's current share priceand Pfizer's offer is currently more than $25 billion.
"I do not know whether AstraZeneca will hit their 2023 salestarget, but I do know that 2023 is a long way off, and suchsuccess is far from certain," said one top-10 shareholder.
Some of AstraZeneca's biggest sales projections are for itsnew immunotherapy drugs for cancer, such as MEDI4736, which itbelieves could achieve peak sales of $6.5 billion a year.
But AstraZeneca is lagging rivals such as Bristol-MyersSquibb, Merck & Co and Roche in thefield of tapping the immune system to fight cancer, and itsattempts to catch up hinge critically on the power of clinicaltrial results.
Some of that data will be presented at an American Societyof Clinical Oncology (ASCO) conference on May 30-June 3.
"They've clearly implied there's going to be very positivedata, and it had better look bloody good," said one top-30investor.
Converting promising early research into winning commercialproducts is a growing challenge for drugmakers, reflecting bothincreasing scientific competition in hot areas like oncology butalso the high hurdle set by governments and insurers in payingfor expensive new drugs.
That is equally true in other areas where AstraZeneca hasflagged promising pipeline assets, such as respiratory drugs,where it is up against powerful players like GlaxoSmithKline that have their own products in development.
Some analysts, like Citi's Andrew Baum, reckon AstraZenecadoes have a truly impressive pipeline, though even his 49pounds-a-share valuation for the shares on a standalone basisremains short of the 55 pounds offered by Pfizer.
Bernstein analyst Tim Anderson is more cautious, writing ina note that AstraZeneca's "track record in R&D is not exactly ashining star, and this may have at least some predictive value".
That view may resonate with the fund manager who commentedback in 2004: "AstraZeneca looked like it had a great pipeline,but suddenly it doesn't look quite as good any more." (Additional reporting by Chris Vellacott; Editing by WillWaterman)
