West Coast Editor

The same cancer drug from Kosan Biosciences Inc. that met with combination-toxicity problems in a study against colorectal tumors has failed to meet its primary objective in a Phase II study against non-small-cell lung cancer, and the company is stopping that trial, as well.

"If we knew why certain drugs in certain tumors fail, we'd all be billionaires," said Robert Johnson, chief medical officer of Hayward, Calif.-based Kosan, adding that the firm is "pretty upbeat" about ongoing efforts to develop the compound known as KOS-862.

Kosan's stock (NASDAQ:KOSN) closed Tuesday at $6.40, down 15 cents.

The NSCLC trial was terminated because it yielded no tumor response in the first step of the two-stage Simon trial design (named after Richard Simon of the National Cancer Institute). But Kosan hardly has given up on the drug, partnered with Hoffmann La-Roche Inc., of Nutley, N.J., in a potential $220 million deal two years ago. (See BioWorld Today, Sept. 24, 2002.)

A Phase II breast cancer trial begun in February continues to enroll patients, with an interim peek likely in the first half of next year, and a Phase II in prostate cancer is expected to start this quarter.

Epothilones work similarly to Taxol (paclitaxel), the chemotherapeutic agent introduced in 1993 by Bristol-Myers Squibb Co., of New York, but are more soluble in water, with higher activity.

"The excitement of the class is that it works in tumors that appear to be taxane resistant," Johnson said. "If you take cells from patients who first have a response to a taxane and then failed, the cells in general have an up-regulation of the multidrug-resistant pump. For some reason, the epothilones as a class aren't very good substrates for this MDR pump."

That, Johnson told BioWorld Today, is "the cake' story. The icing' story is that if you go head to head with a taxane, you could become superior and treat patients for many more cycles," he said, noting that taxanes are "being used more and more as first line in breast cancer."

Kosan also might take encouragement from results gained by BMS with another drug in the epothilone class: BMS's ixabepilone, also known as BMS 247550, which has reached Phase III and reported good Phase II results in breast cancer at this year's meeting of the American Society of Clinical Oncology. BMS also is studying the drug in prostate cancer.

"We have Roche as a partner, so obviously they think there's room for more than one," Johnson said, noting the BMS compound brings with it "a lot of toxicity."

Kosan is no stranger to such problems. During the summer, cumulative drug toxicities in KOS-862 patients who previously had been treated with Eloxatin (oxaliplatin), from Paris-based Sanofi-Synthelabo SA (now Sanofi-Aventis Group), forced Kosan and Roche to stop the Phase II colorectal trial and turn its eyes to other tumor types. (See BioWorld Today, June 24, 2004.)

"Because it was so new to the U.S. market, a lot of our patients had recently come off oxaliplatin," Johnson said. For first-line treatment of colorectal cancer in combination with standard treatment 5-fluorouracil and leucovorin, Eloxatin won U.S. approval at the start of this year. It had been approved as a second-line treatment in August 2002.

Also during the summer, Kosan amended the Roche deal, providing for cost reimbursement at a higher level. In the nine months ending Sept. 30, the company got $17.1 million from Roche, according to Kosan's quarterly report.

Daniel Santi, founder, chairman and CEO of Kosan, told BioWorld Today the company has its lead geldanamycin analogue, 17-AAG, in Phase II and Phase Ib trials, and DMAG, its second-generation geldanamycin analogue, in Phase I. Both are being developed in collaboration with the NCI, and both are polyketide inhibitors of the "molecular chaperone" Hsp90, interrupting processes implicated in cancer-cell growth and survival, he said.

About 15 Phase Ib combination trials and Phase II studies with 17-AAG eventually will be under way.

The company's own formulation of 17-AAG, tagged KOS-953, also is moving along.

"That's a Phase I trial that will convert into Phase II for multiple myeloma," Santi said, adding that another Phase Ib trial has just begun to test KOS-953 against multiple myeloma in combination with Velcade (bortezomib), the proteasome inhibitor from Cambridge, Mass.-based Millennium Pharmaceuticals Inc.

Kosan also will try KOS-953 against melanoma and against chronic myelogenous leukemia in patients who have failed in treatment with Gleevec, from Novartis AG, of Basel, Switzerland.