* Swiss-led R&D unit's drugs among top pipeline hopefuls
* California's Genentech has long dominated Roche R&D
* CEO sees "exciting opportunities" after dry spell
ZURICH, Dec 19 (Reuters) - Roche is touting prospective new drugs from its long-underperforming Swiss-led research unit after years of leaning on its California-based Genentech arm to restock its medicine cabinet.
This year, the Swiss drugmaker - the top global spender in pharmaceutical research and development (R&D) - has been talking up several medicines being trialed by Roche Pharma Research & Early Development, or "pRED," as potential stars.
Its pRED unit operates independently from Genentech "gRED" research and Roche's Chugai R&D arm in Japan.
A resurgence would be timely. Chief Executive Severin Schwan needs his R&D operations firing on all cylinders to fulfill promises of growth as patents expire on Rituxan, Avastin and Herceptin. These Genentech-developed drugs, which have combined sales of $20 billion a year, are either already exposed to rivals' cheaper copies in leading markets or soon will be.
"In pRED, some exciting opportunities are now coming through after a time when many things did not work," Schwan told Reuters in a recent interview. "It goes in waves. You can't program to have a certain number of molecules coming through the pipeline every year in each unit."
Topping pRED's list of pipeline hopefuls is CEA-TCB, a so-called bispecific antibody drug that brings a patient's cancer-fighting T-cells closer to tumor cells in order to kill them. Other promising products include a replacement for eye drug Lucentis, idasunutlin for diseases including acute myeloid leukemia and an autism drug in mid-stage Phase II trials.
Though approvals remain some way off - idasunutlin's planned filing date is 2019, for instance - Roche insiders in Basel are relieved that pRED appears to have regained its footing following years in Genentech's shadow.
"Thank God," one Roche executive said, requesting anonymity. "It took a while, but pRED is finally starting to deliver."
From the 1990s, Roche has thrived largely on Genentech's hit-making machine. After Rituxan, Herceptin and Avastin came Lucentis and cancer drugs Perjeta and Kadcyla.
Since 2016, Genentech's labs added cancer immunotherapy Tecentriq and multiple sclerosis drug Ocrevus, seen by analysts as hitting $1 billion sales this year.
Additionally, Roche's recently-approved hemophilia drug Hemlibra, predicted by some as a $5 billion-per-year medicine, emerged from Chugai's labs. So did Alecensa for lung cancer, another drug with blockbuster aspirations.
So pRED still has much catching up to do. "Roche's three big drug hopefuls - Ocrevus, Tecentriq and Hemlibra - aren't from pRED, they're from elsewhere in the company," said Michael Nawrath, a Zuercher Kantonalbank analyst.
"Without the Americans, Roche would be just a specialized diagnostics company."
FUMES OF YESTERYEAR
By contrast, the reputation of Roche's own research organization, after inventing Valium in the 1960s and antibiotic Rocephin in the 1980s, has run on the fumes of yesteryear.
Even pRED's much-heralded Rituxan follow-up drug Gazyva has so far produced just modest results with $200 million in 2016 sales, though recent expanded approvals could accelerate revenue.
In 2010, Roche closed a U.S. branch of pRED's labs in Nutley, New Jersey, and slashed 1,000 jobs. Its third research chief since 2012, John Reed, joined four years ago from California's Sanford-Burnham Medical Research Institute.
His mission was to help get pRED's 2,200 scientists in Switzerland, Britain's Welwyn Garden City, Germany and Shanghai back on track by focusing less on "blue sky" projects and more on medicines underpinned by a solid hypothesis.
A fund manager who owns Roche shares sees Reed's arrival in Basel as a watershed. "The effectiveness of Roche R&D in Switzerland has been transformed by John Reed," the investor said. "I am quite optimistic about Swiss R&D starting to produce new drugs."
Some analysts also see pRED closing the gap to Genentech, perhaps partly by emulating its focus on therapeutic antibodies that the San Francisco-based company pioneered.
"Quite a few competitive products in development at pRED are based on antibodies," said Baader Helvea's Bruno Bulic. "It might be that at some point, the student might surpass the master."
In 2009, when Roche bought the 44 percent of Genentech it did not already own for $47 billion, Schwan insisted on keeping the pRED, gRED and Chugai research organizations separate on the grounds that a combined R&D monolith would "kill innovation."
With pRED potentially re-emerging from the drug-development wilderness, Schwan remains adamant it was the right decision.
"I can understand that if somebody looks at it from outside, they might ask why the hell we have three research units," he said. "But I believe it doesn't make sense to put it all together and make one big sauce out of it. You should still see the components." ($1 = 0.9899 Swiss francs) (Additional reporting by Ben Hirschler and Simon Jessop in London; editing by David Stamp)