Last checked at 
* MEDI4736, tremelimumab combination tested in lung cancer
* Safety data reported for 24 patients, efficacy for 18
* AstraZeneca says tolerability profile encouraging
* Objective response rate in patients 28 percent (Adds comments from researcher, paragraphs 5 and 11)
By Ben Hirschler
MADRID, Sept 27 (Reuters) - Early results for a closelywatched cancer drug combination from AstraZeneca thatboosts the immune system show the cocktail is promising, thoughlimited patient numbers mean the data is far from conclusive.
The British drugmaker, which fended off a $118 billiontakeover bid from Pfizer in May in part by talking upits cancer drug prospects, has high hopes for the combination oftwo experimental drugs known as MEDI4736 and tremelimumab.
The company is still exploring a range of doses, so testingof the drugs in lung cancer is taking time to yield results anddata on only two dozen patients was reported at the EuropeanSociety for Medical Oncology (ESMO) congress on Saturday.
Chief Executive Pascal Soriot had said earlier this monththat the ESMO numbers would be limited.
Still, researcher Scott Antonia of the Moffitt Cancer Centerin Florida said the early signals were encouraging, both forsafety and efficacy. "It looks very, very promising," he said.
AstraZeneca expects to have more definitive results laterthis year and also plans to start a pivotal clinical trial withthe combination either late this year or at the start of 2015.
Immunotherapy treatment is the hottest area of cancerresearch - widely tipped to become a market worth tens ofbillions of dollars in annual sales - and combinations areviewed by many oncologists as the best way to use the new drugs.
Safety, however, is an issue, especially after results fromanother small study with a similar Bristol-Myers Squibb cocktail showed about half of patients experienced serious sideeffects, with three treatment-related deaths.
In the case of AstraZeneca's combination, six out of 24advanced lung cancer patients had adverse events rated asserious, or grade 3/4, and three had events that led todiscontinuation of treatment. There was one treatment-relateddeath.
So far, 18 of the patients have been assessed for efficacyand five of these, or 28 percent, had tumour shrinkage,according to research presented at the meeting in Madrid.
Although direct comparisons are difficult, Antonia said thiswas much better than the efficacy benefit seen with theBristol-Myers combination in lung cancer.
"In terms of what you would hope to see at this point, weare very much on track," Edward Bradley, head of oncology atAstraZeneca's biotech unit MedImmune, told Reuters. "It's earlydays but we're pleased with where we are and I think it's a verymanageable tolerability profile."
$6.5 BILLION FORECAST
AstraZeneca already presented data on a handful of patientsat the American Society of Clinical Oncology earlier this year.The new results build on that by providing more safety data andshowing some evidence of clinical activity in sick patients whohave failed to respond to other drugs.
MEDI4736 is part of a class of drugs known as anti-PD-L1therapies, which work by blocking a tumour's ability to evadethe immune system's defences. Tremelimumab is a so-calledanti-CTLA4 drug that unlocks a different brake on the immunesystem.
The two-pronged approach is designed to expose cancer cellsas fully as possible to the killing power of the body's ownimmune system. But boosting the immune system can cause damagingside effects, including colitis, a serious inflammation of thecolon, as well as liver and thyroid problems.
Immunotherapy drugs are seen as AstraZeneca's most importantpipeline assets and the company has predicted that MEDI4736 could generate annual sales of $6.5 billion, including its usein combinations.
AstraZeneca is vying with rivals Bristol-Myers Squibb, Merck& Co and Roche in the immunotherapy race.
It is viewed by analysts as being behind these leaders butthe company has a long history in cancer treatment and believesit is in a good position to develop a wide range of drugcocktails.
Because such immunotherapy does not work for all patients,some companies have looked to focus on people whose tumours testpositive for a likely response. However, most of the patientsassessed in the AstraZeneca study were actually PD-L1 negative.
"This supports our strategy to explore this combination morebroadly, particularly in the PD-L1 negative population," Bradleysaid.
Currently, immunotherapy is most advanced as a treatment formelanoma but research is advancing rapidly into other tumourtypes, with non-small cell lung cancer - a major killer - thebiggest commercial opportunity.
As a result, doctors and investors alike are following theAstraZeneca drug combination very closely, particularly afterthe earlier disappointment with Bristol-Myers' combination usingnivolumab and its already approved drug Yervoy. (Editing by Pravin Char and David Holmes)
