By Randall Osborne

West Coast Editor

SAN FRANCISCO - It was a glory day for Genentech Inc. at the opening of the J.P. Morgan H&Q Healthcare Conference here, as the company disclosed two new deals and vowed to forge ahead with a third.

Audience members at the packed keynote luncheon put down their forks and listened attentively while Arthur Levinson, Genentech chairman and CEO, looked to the industry's future.

Some attendees, sent away from the Grand Ballroom for lack of table space, returned with boxed sandwiches and leaned against the wall as Levinson spoke, providing tips for biotechnology firms not so far along as the pioneering Genentech - thereby also providing guidance for investors on what to watch for.

"I've got a pretty good track record," Levinson said, in his characteristic rapid-fire style. When someone makes a drug pitch with "54 flow charts, and 94 references, and you can't make sense out of it, it's not necessarily because you're stupid," he said. "It's typically because not very much is known about the mechanism of action."

And that means the drug is much more likely to fail.

OSI Pharmaceuticals Inc. apparently had the simplicity Genentech favors, and garnered a potential $187 million development deal, joined by Roche Holdings Inc., for OSI's anticancer drug (see related story on page 1).

Alkermes Inc. won a bigger piece of partner Genentech's pocketbook, too, as the latter decided to pay for Phase II/III trials in the first half of this year, testing Nutropin Depot (somatropin) for growth hormone deficiency.

COR Therapeutics Inc., of San Francisco, also had what Genentech likes, and netted a three-way co-promotion agreement that includes potential research funding, along with Madison, N.J.-based Schering-Plough Corp., which is developing Integrilin (eptifibatide) with COR.

Under the terms of that deal, Genentech will co-promote Integrilin, an intravenous glycoprotein IIb-IIIa inhibitor, used to block blood clots in the heart, in the 5,000-plus hospitals at which it currently sells TNKase (tenecteplase) and Activase (alteplase), thrombolytic agents for acute myocardial infarction.

Integrilin is for patients with acute coronary syndrome (unstable angina and non-Q-wave myocardial infarction), including those to be managed medically and those undergoing percutaneous coronary intervention (PCI). The latter may also be treated at the time of the PCI procedure.

COR and Schering-Plough, in turn, will co-promote TNKase and Activase in the more than 2,000 hospitals they current call upon for Integrilin. All parties also have agreed to collaborate on any future large-scale clinical trials combining a fibrinolytic with a glycoprotein IIb-IIIa inhibitor.

Clinical trials in general were a subject of Levinson's keynote speech - in smaller companies especially, "the lack of resolve to kill failures, particularly at the end of Phase II, or even Phase III," he said.

"Beware when you see a negative Phase II or Phase III study, and the company says, 'Well, we're not really that disappointed, because there are some signs of efficacy here. Let's take a look,'" he said. "Then, four months later, they say, 'If we'd only looked at this population instead of this population, the drug would have worked.'"

Levinson compared it to a 200-round roulette game, in which the player loses half his money and discovers later that he won every time he put his money on a multiple of seven. He decides to put all of his remaining money on the number 21, and "it sometimes works, and if it does you're as lucky as anything else," Levinson said. "You have to be very careful. Companies do that all the time, all the time."

Looking ahead, he sees promise in modulators of angiogenesis. "We understand why some of the failures have occurred," he said. He also voted for antigen-targeted anticancer therapies, "based on a Rituxan or Herceptin type of model. We've set a pathway to identifying those particular antigens and the development of antibodies that attack those cancer cells."

Genentech's drugs include Rituxan (rituximab), for relapsed or refractory low-grade or follicular, CD20-positive, B-cell, non-Hodgkin's lymphoma, and Herceptin (trastuzumab), for HER-2 positive metastatic breast cancer.

Signal transduction inhibitors will show "possibly some dramatic successes," too, Levinson said, pointing to the OSI compound that was the subject of the Genentech deal.