The trouble started in mid-October. That's when the price of Amgen Inc.'s shares, which had been at about $67, fell like autumn leaves, dipping just below $62 before an upsurge became evident in the first part of last week.

"It's amazing to me that one of the cleanest stories in biotech is trading [this low]," Eric Schmidt, analyst with SG Cowen, said Wednesday. "We'd be buyers for sure."

After the market closed Tuesday, Amgen offered its second-quarter earnings - which surpassed Wall Street estimates for the sixth quarter in a row. But even this happy news sparked no next-day revival. In fact, the shares lost several dollars and closed near the $60 mark, quite a distance from Amgen's 52-week high of about $72.

What happened? Puzzling out the market's fickle ways is dangerous business, but a pair of developments may have conspired to lower the company's shares and keep them down, at least temporarily.

The main factor may have been publication in The Lancet on Oct. 18, 2003, of findings from a European study showing erythropoietin for anemia in cancer patients failed to help them live longer and might have reduced the efficacy of their treatment.

Results from the study in patients with mouth and throat cancers were picked up by the New York Times and other newspapers. The form of EPO used was NeoRecormon (epoetin beta) from F. Hoffmann-La Roche Ltd., which sponsored the trial. Amgen, of course, sells EPO as Aranesp (darbepoetin alfa) - the second generation of the company's red-blood-cell booster Epogen (epoetin alfa) first approved for chronic renal failure.

The European study in 351 patients at 23 medical centers examined whether treating anemia might boost the effects of radiation, since radiation works less well in anemic patients whose lower-oxygen cells tend to become resistant. One hundred eighty patients were given EPO along with radiation, while 171 got placebo.

EPO did not improve survival or enhance treatment. The drug might even have made the situation worse. Sixty-one percent of the patients on the drug died from cancer or other causes, as compared to 52 percent of those given placebo.

SG Cowen's Schmidt said The Lancet report "should have little impact" on the market. William Tanner, analyst with Leerink Swann & Co., was less certain.

"While results may likely pose little near-term threat, we would be vigilant for the impact of increased scrutiny on EPO usage," Tanner wrote in a research note.

Tanner said five consultants with MEDACorp, a division of Leerink, reviewed the study and found its design "lacked obvious flaws." Fewer smokers were enrolled in the placebo arm (91) as compared to the drug arm (118), and the drug arm had more relapsed patients (18) than the placebo arm (13), but these were considered minor problems.

CERA Threat Not Fully Appreciated,' Report Claims

"Criticism available in the popular press that the patients were not, in fact, very anemic was not highlighted by our consultants," Tanner wrote. Biologically, "the main issues relate to whether the response pertains only to head and neck cancer," he wrote.

Whether it is exclusive to that cancer type or not, other analysts noted, the labels for Aranesp and Epogen indicate the drug is for patients being treated with chemotherapy, not radiation. Furthermore, they are intended to improve the quality of life, not necessarily survival.

Roche figures into another influence that might have taken value from Amgen's stock. The overseas pharmaceutical giant is preparing to enter Phase III trials with continuous erythropoiesis receptor agonist (CERA), a pegylated longer-acting form of EPO that some observers say could make a dent in the market for Aranesp - if CERA gets that far. Amgen already is involved in a war with Transkaryotic Therapies Inc. over its version of EPO, Dynepo (epoetin delta). The same thing could happen with Roche.

Sanford C. Bernstein, an investment research and management unit of Alliance Capital Management LP, said in an Oct. 14 research note that three "key issues" will determine whether CERA can make a difference in the U.S. market: clinical profile; the ability of Roche to break the patent defenses of Amgen and licensee Johnson & Johnson; and CERA's worth against EPO generics, which are on their way.

"Our latest analysis of these three dynamics support our hypothesis that CERA's chances of entering the U.S. market are good - a view not fully appreciated by consensus," said the Bernstein report. Like the European study with NeoRecormon (albeit to a lesser degree), the report was picked up by nonscientific media that included Forbes magazine.

"I don't think there's much new here," Schmidt told BioWorld Financial Watch, noting that EPO has "been around in the medical community for quite a while now. I know Roche is talking about a potential worldwide launch and the product, I'm sure, works. Our issue, if we have one at all, is that we don't think that's enough to get into the U.S. market."

Amgen, Schmidt noted, was "pretty adamant" on that subject in a conference call regarding second-quarter earnings, during which the would-be infringement of CERA on the company's patents was a topic.

"We've got enough experience with the TKT litigation to know their patents are pretty strong, pretty all-encompassing," he added, pointing out that many court decisions on EPO so far have gone Amgen's way. Many of the issues that Amgen has used against TKT (such as pharmaceutical composition of matter claims on all EPO backbones) are applicable to CERA, Schmidt said.

It would take much, in any case, to "reverse the tremendous growth curve this company is on," Schmidt said.

In the latest numbers, combined Epogen and Aranesp sales increased by 58 percent from the third quarter of 2002, jumping from $672 million to $1.1 billion, thanks mainly to Aranesp, approved in the U.S. in mid-2002 for chemotherapy-induced anemia. Epogen sales contributed $626 million, up 12 percent over last year's third quarter.

Enbrel (etanercept), first approved for rheumatoid arthritis, tallied third-quarter sales of $342 million, a 116 percent increase over sales of $158 million in the same period last year, when a supply shortage affected sales (as well as two weeks of missed sales, since the buyout of Immunex Corp. took place July 15).

"I guess in the summertime there was a market slowdown," Schmidt allowed. "But over the last five or six weeks, [prescriptions] have really picked up."

Plenty More In Pipeline For Amgen To Show

Neulasta, the longer-acting version of the white-blood-cell stimulator Neupogen (filgrastim), chalked up $327 million in sales in the third quarter of this year, $23 million of which were international sales. The two drugs brought in $657 million in total sales, which were up 39 percent from $474 million during the same quarter last year.

Cory Kasimov, analyst with Ryan Beck & Co., last week reiterated his firm's "outperform" rating.

"Despite the potential for continued volatility stemming from drug pricing debates in Washington and longer-term competitive threats to Amgen's EPO franchise, we believe the company's valuation is compelling relative to its large, profitable biotechnology peers," Kasimov wrote in a research note. With "longer-term threats," Kasimov was "directly referring to Roche's CERA," he told BioWorld Financial Watch.

"We don't have a lot of information about [CERA] now," he conceded. "It is early, and [EPO] is a huge market. It's going to be one potential competitive threat after another."

Amgen is hosting its first-ever "research and development day" for investors in the first quarter of next year, Kasimov noted, adding that "whether they're doing this because they have exciting opportunities they want to tell investors about or they see threats to their commercial franchise remains to be seen."

Maybe it's a little of both. Phase III data are expected in the fourth quarter of this year on cinacalcet hydrochloride at the American Society of Nephrology meeting, which begins Nov. 12 in San Diego. The drug was licensed from NPS Pharmaceuticals Inc. in 1996 for secondary hyperparathyroidism associated with chronic kidney disease, and Amgen filed a new drug application in September.

Approval is expected in the middle of next year of a supplemental biologics license application for Enbrel in psoriasis. Also next year, Amgen is expected to file a BLA for recombinant keratinocyte growth factor to treat oral mucositis in the hematologic transplant setting.

For now, Amgen is defending EPO against TKT, and "as soon as that one disappears, it now looks like there could be a chance of another" challenge, assuming CERA makes progress, Kasimov said. Like another heavy hitter in biotechnology, Genentech Inc., Amgen is ferocious about guarding its patents. "They have to be," Kasimov said of Amgen officials. "It may make [challengers] think twice about how good is the intellectual property they really have."

In any case, the slide in Amgen's price during recent weeks made it a bargain worth snatching up, he said. CERA "is out there, it's real," Kasimov acknowledged. "But I don't think it's anything too alarming at this point, not enough to derail Amgen to the extent that it has."