By Randall Osborne

West Coast Editor

With its lead product for head and neck cancer in Phase III trials, Onyx Pharmaceuticals Inc. said Friday it filed to sell 3 million shares at $21.53 per share, which would raise $64.5 million.

Adding the underwriters' overallotment option for 450,000 shares, the offering would raise $74.27 million at the sought price. The Richmond, Calif., company's stock (NASDAQ:ONXX) closed Friday at $20.50, down 81.25 cents. The company will have about 17.4 million shares outstanding after the offering.

Analyst David Bouchey, with New York-based C.E. Unterberg, Towbin, said the offering is "pretty substantial. The company actually doesn't need the cash," he added, noting that Onyx has about $44 million, with a burn rate until recently of about $11 million to $12 million per year.

That burn rate "doubles from this point on," Bouchey said, since Onyx's partner, Warner-Lambert Co., of Morris Plains, N.J., a wholly owned subsidiary of New York-based Pfizer Inc., has agreed to pay only the first $40 million in full for developing CI-1042 (formerly ONYX-015), a genetically engineered adenovirus, which also is in Phase I and Phase II trials for other cancer indications.

"After [the $40 million], Onyx starts kicking in," Bouchey said.

U.S. Bancorp Piper Jaffray is lead manager for the offering, which is co-managed by CIBC World Markets. Bouchey is less than optimistic about Onyx's chances with the cancer drug.

"I have a neutral recommendation," he said. "I don't want to say anything bad about the company or the management. It's just the way the science looks. Although the company touts CI-1042 as selective to a defective p53 [tumor suppressor] gene, there's a lot of academic research showing why it doesn't have to be selective."

Also, the trend toward longer survival rates, apparent from the Phase II data, could go away in the larger, Phase III study, Bouchey said.

The Phase II trial enrolled 37 patients, 32 men and five women, ages 22 to 78. Of their number, 32 (86 percent) used tobacco; 34 (92 percent) alcohol. Treatment combined CI-1042 injected directly into their tumors with two systemic, standard chemotherapeutics, cisplatin and 5- flurouracil. Each participant had at least two separate tumors, of which the larger received the combination therapy, while the smaller - control group - got only chemotherapy. Of the 37-patient cohort, seven dropped out for various reasons, leaving 30 for evaluation. (See BioWorld Today, Aug. 2, 2000, p. 1.)

"They're going have to go before the FDA's Oncology Drug Advisory Committee, which may have a cerebral hemorrhage when it looks at this [lack of p53 connection]," Bouchey said.

But the agency may be happy with any efficacy at all, Bouchey said. "There's about a 50-50 chance" the drug will be approved, he added, acknowledging that other analysts differ with his opinion about Onyx. SunTrust Equitable Securities Inc., of Atlanta, for example, last month maintained its strong buy rating, based on trial data.

Still, the competition is lined up against Onyx, Bouchey said. Matrix Pharmaceutical Inc., of Fremont, Calif., in May disclosed Phase III results showing its IntraDose (cisplatin/epinephrine) Injectable Gel resulted in reductions in tumor volume of 50 percent or more in 45 percent of 102 patients treated. Response rates were consistent across all treated tumor types, including breast, melanoma and esophageal. Median duration of response ranged from 36 days for esophageal tumors to 78 days for recurrent breast tumors. The product is a cisplatin and collagen gel designed for intratumor injection.

"That's a single injection, as opposed to Onyx's series of five injections," Bouchey said. "It could be on the market in nine months."

OSI Pharmaceuticals Inc., of Uniondale, N.Y., achieved impressive early results with OSI-774, in a Phase II clinical trial in non-small-cell lung cancer (NSLC). The drug is an orally active inhibitor of the epidermal growth factor receptor (EGFR).

"We believe this forebodes ill for CI-1042, as OSI-774 is also being tested in a 100-patient Phase II trial in head and neck cancer," with data due at a scientific meeting in November, Bouchey wrote in a research note. "We believe this trial could show significant efficacy for OSI-774 in head and neck cancer, since it appears active in NSLC. Both cancers, but especially head and neck, exhibit overexpression of EGFRs."

In any case, the market for CI-1042 will be small, Bouchey said - about 12,000 to 15,000 patients.

"That's a small number, when you consider [marketing partner Warner-Lambert] gets the lion's share of the revenue," Bouchey said. "I don't think it justifies the stock price where Onyx is now."

Bouchey said Onyx probably tried a private placement before the offering.

"I don't know what happened to change their mind," he said. "If I've been able to identify these issues, there are some very smart people out there [who can, too]. People who take a serious look, as you would on a road show, would probably see the same things."