West Coast Editor

As Genentech Inc. hinted earlier, marketing clearance of Avastin for breast cancer has been pushed back by the FDA's request for an independent review of patient scans from the pivotal trial, and satisfying the agency could take longer than analysts expected.

"They didn't give us any specifics [earlier], but we didn't know it would be a year," said Christopher Raymond, analyst with Robert Baird & Co. in Chicago.

The South San Francisco-based firm's stock (NYSE:DNA) dipped on the news, closing Monday at $78.33, down $3.74, possibly because investors grew jittery about Avastin's chances at the FDA against non-small-cell lung cancer as well.

Genentech expects word in October about Avastin for NSCLC, which could add another $1 billion yearly to the bottom line.

Monday's sell-off aside, trouble from the FDA seems less likely in this indication, since the NSCLC application is based on the more solid endpoint of overall survival rate, rather than progression-free survival (PFS), as in the breast-cancer trial.

"Lung cancer [approval] is a very real possibility," said Raymond, who put Avastin's odds at 80 percent or greater in that indication. Raymond's rating on Genentech is "neutral," with a price target of $94.

Already approved for colorectal tumors, anti-VEGF Avastin (bevacizumab) sold $423 million in the U.S. during the second quarter, a 72 percent jump over the same period last year. (See BioWorld Today, July 13, 2006.)

Wall Street, though, had expected $15 million more from Avastin for the quarter, and Genentech needs the broadened label in order to hit the higher targets, especially with consumer groups pressuring the company about pricing. Avastin sold $1.1 billion last year, and many analysts believe the breast cancer indication will add at least another $1 billion, and possibly much more.

In breast cancer, at issue for the FDA are data from the E2100 trial, done by a network of researchers led by the Eastern Cooperative Oncology Group, which stopped the 685-patient study early because Genentech's antibody worked so well. The PFS in metastatic breast cancer patients given Avastin plus paclitaxel hit 11 months, compared to six months with paclitaxel alone.

But regulators, in their complete response letter to Genentech, want E2100's results audited as they would be if the company had sponsored the study - a demand Genentech didn't expect, although the firm's "language [two months ago] was hinting at a delay" and listeners might more wisely have expected something substantial, Raymond said.

During a July conference call related to second quarter earnings, Susan Desmond-Hellman, Genentech's president of product development, told investors that the company had not "established exactly what the [necessary] documentation will be, but in the middle of those negotiations [with the FDA], we became concerned about the timing."

Now, Genentech estimates the resubmission of its supplemental new drug application will not happen until the middle of next year. Then the six-month approval countdown will begin, which means the label expansion will take until at least the end of 2007.

E2100 tested patients who had either HER2-negative disease, or were HER2-positive and had either been previously treated with Genentech's breast-cancer therapy Herceptin (trastuzumab) or were deemed unsuitable for Herceptin, which sold $747.2 million last year, beating 2004 by 56 percent.

Herceptin is among the drugs that have been approved on the basis of unblinded PFS trials. Another is Taxol (paclitaxel), from Bristol-Myers Squibb Co., of New York. "Breast cancer has the bigger opportunity" of Avastin's two pending cancer indications, Raymond said.

Avastin, first approved in 2004 for colorectal cancer, has gained drug compendia listings for breast and lung cancers (in May 2006 and September 2005, respectively), which means insurance companies and Medicare will reimburse for them. The drug is widely used off label, but Genentech needs formal FDA clearance to rev sales more powerfully in those indications.

Analyst Bret Holley, with CIBC World Markets in New York, predicted Avastin's approval at the end of 2007 or the start of 2008, and deemed Genentech a "sector outperformer," though he reduced his price target to $96 from $98 on news of the delay and estimated U.S. sales of Avastin for breast cancer, if approved, could reach $800 million-plus by 2010.

Avastin is partnered with Basel, Switzerland-based F. Hoffmann-La Roche Ltd., which owns a majority of Genentech.