Editor

The near-term meaning of this month's blowup of AstraZeneca plc's Iressa, the epidermal growth factor inhibitor against non-small-cell lung cancer, was clear - a strong boost for Genentech Inc. and OSI Pharmaceuticals Inc., whose EGFR drug, Tarceva, was approved for the same indication in November.

OSI's stock jumped more than 45 percent. Genentech's shares benefited, too. AstraZeneca, holding a once-ballyhooed "wonder drug" that apparently doesn't work as well as they hoped, was trying to figure out what to do next.

In the FDA-required trial that enrolled 1,692 subjects, Iressa as monotherapy failed to show a survival benefit vs. placebo in the overall NSCLC patient population or in those with adenocarcinoma (a tumor that originates in glandular tissue).

Iressa (gefitinib) gained marketing clearance in the spring of 2003 on the basis of tumor response rate, with regulators calling for a trial demonstrating clinical benefit later. Tarceva (erlotinib), on the other hand, won the FDA's nod on the basis of a 42.5 percent survival advantage over placebo.

Until the bad news about Iressa, many observers assumed the AstraZeneca compound and Tarceva "were essentially the same drug," and matching the former's success with the latter's was practically a formality, said Winton Gibbons, analyst with William Blair & Co.

That notion turned out to be about as wrong as it could be. Iressa's failure instilled even more confidence regarding Tarceva's future, with the likes of Jennifer Chao, an analyst with Deutsche Bank Securities Inc., reiterating projections of revenues as high as $400 million by 2006.

Does the Iressa imbroglio hold any longer-term lessons? Michael Zasloff, analyst with Ferris Baker Watts, thinks so. One lesson is that accelerated approval (granted to Iressa on the basis of Phase II data) doesn't necessarily bode well. Another is that not all of the "targeted" therapeutics generating so much excitement will prove worthwhile. A third, Zasloff said, is that even major pharmaceutical firms can be less than fully diligent in their development efforts.

"I was surprised in the sloppiness of some of what I read" about Iressa research, he said.

A Phase II study with Iressa in 216 patients with refractory disease yielded data that AstraZeneca (and, it seems, the FDA) found promising, even though no placebo control was included. Half the patients got 250 mg of Iressa, the other half 500 mg. As had been seen in earlier work, about 10 percent of the patients saw their tumors shrink, in some cases dramatically. About 40 percent of trial participants said their disease-related symptoms backed off. Two more Phase II studies, similarly designed, also yielded 10 percent to 20 percent tumor shrinkage.

Two Phase III trials have evaluated Iressa when added to first-line therapy, but the drug showed no advantage there. Still, AstraZeneca pointed to the tumor shrinkage in earlier trials and the reported relief of symptoms, and asked the FDA to grant accelerated approval based on those.

In September 2002, the Oncology Drugs Advisory Committee gave its blessing to Iressa in third-line treatment of NSCLC where no viable treatment options exist. Members liked the tumor shrinkage and the evident improvement of symptoms, even though patients had not been forbidden other drugs, such as narcotics, during the trial.

Iressa's ultimate fizzle might have had investors in OSI and Genentech gloating, but Tarceva is not out of the woods entirely with the FDA. Approved as second-line therapy for NSCLC, Tarceva must undergo more studies of efficacy and these are expected to finish by 2008 (although only unexpected adverse reactions could take it off the market).

Safety and efficacy of Tarceva were proved in one Phase III trial in 731 patients, with a primary endpoint of survival. As with Iressa, about 10 percent of patients who got Tarceva saw a reduction in tumor size, compared with about 1 percent of those given no drug. Median time to progression of tumor growth was 33 weeks in the Tarceva group, compared with 17 weeks for the placebo. The median overall survival was 6.7 months in the Tarceva group, compared with 4.7 months in the placebo group.

Survival at one year was 31 percent for Tarceva, 22 percent for the placebo - statistically significant, although some said that benefit might have dwindled in a larger trial. Furthermore, Tarceva (like Iressa) added no survival benefit in Phase II trials when added to first-line chemotherapy.

Zasloff sees a future for both drugs, allowing that the path for Tarceva is definitively clearer. Whether AstraZeneca carries the Iressa ball or some other company takes on the challenge, the compound might benefit from sophisticated diagnostic tests. Recent studies have shown EGFRs expressed on the cancer cells of those responding to Tarceva or Iressa bore specific mutations that made those cancers particularly susceptible to the drugs, he noted, citing work published in May in the New England Journal of Medicine. A similar paper appeared in Science in April.

The FDA might ask AstraZeneca to test Iressa against placebo in patients with the specific mutation, or test the drug in subjects who express the "correct" receptors for drug activity against those who lack the mutations. Given the costs of Iressa and Tarceva, HMOs and the Centers for Medicare and Medicaid services are likely to limit reimbursement based on similar considerations, Zasloff said.

"If I were in charge of allocating countries' resources I would say, Look, why give folks a particular drug when 90 percent of them are not going to respond?'" he said, predicting a "heavily segmented" market for targeted therapies and noting the diagnostic test used with Herceptin (trastuzumab), the breast-cancer drug from Genentech.

"The real challenge is, what happens to OSI if the FDA or CMS says they no longer have such and such a market [for Tarceva] - they have one-tenth of that?" Zasloff added. "The commercial potential of that drug is profoundly reduced."

But, he told BioWorld Financial Watch, the change is inevitable.

"They're going to force pharmacogenomics into drug development," he said. "It's going to be that way from now on. It has to be."