West Coast Editor

SAN FRANCISCO - Amgen Inc.'s reimbursement woes and safety concerns around its erythropoietin stimulating agents (ESAs) made for a tough 2007, but talks continue with the FDA, and work force cutbacks should allow the firm to beat its current full-year earnings guidance, attendees heard during the 26th annual JPMorgan Healthcare Conference here.

And internal efforts such as the biomarkers program could help Amgen backfill its pipeline, Scott Patterson, director of medical sciences, told BioWorld Today. The push already has yielded good results related to Vectibix (panitumumab), cleared in September 2006 for colorectal cancer that has spread despite standard chemotherapy. Vectibix won approval near the end of last year in Europe as well, for patients with epidermal growth factor receptor-expressing metastatic colorectal cancer with non-mutated KRAS.

"It's not that we're assuming our measurements are substitutes for a clinical endpoint," Patterson said. "We just want to know at the biochemical level, is [the drug] doing what it should?"

Vectibix did. Phase III data offered during the European Conference of Clinical Oncology in Barcelona, Spain, in September supported using KRAS as a predictive clinical biomarker for selecting probable responders to Vectibix. The drug's effect on progression-free survival was confined to the 60 percent of patients whose tumors harbor normal, non-mutated KRAS, with no effect in patients who had tumors with mutations in KRAS, regardless of the endpoint studied. Data on progression-free survival and response rate in heavily pre-treated patients were reported earlier.

"We want to know, at the biochemical level, whether [a specific drug] is modulating the pathway we anticipated," Patterson said. "We're adding in, when we can, a pharmacodynamic aspect" to the safety, tolerability and pharmacokinetics ordinarily studied in Phase I trials.

The pharmacodynamic aspect played more commonly into later-stage development in the past, and seldom as extensively as Amgen deploys it at the early stage now, Patterson said, adding that other companies are starting to catch on to the approach as well.

Started four years ago, Amgen's biomarker program also seeks to define strata within patient populations. "We'd like to be able to say how many different types of disease [there are], and whether our responses correlated with one of those strata," he said. "We might begin to explore it in Phase I and take it into Phase II, and expand upon it in Phase III," he said, noting that a postmarketing study goes on with Vectibix.

The biomarker strategy will not work in all cases, but it's already allowed Amgen to rule out drug candidates early, Patterson said. "We had a case where we had a small-molecule, dual-receptor antagonist. It worked at different concentrations. In the single-ascending dose study, everything was fine," but problems with liver counts arose in the multiple-ascending dose attempt. "We found that the level we could employ allowed us to block the pathway through one of the receptors, but not through the other," he said. "We couldn't test the hypothesis we want to in humans. When we take a molecule into man, we want a testable hypothesis."

The brick-wall outcome is "not something we expect to occur regularly," Patterson added. "There's a lot of analytical testing of the assays, so we're sure our measurements are robust enough for us to make internal decisions on the progression of the molecules."

A case also has arisen where dose levels were changed, based on early biomarker findings.

Amgen's trouble with ESAs took center stage when data suggested that the anemia compounds could bring about an increased risk of heart complications and death. The Centers for Medicare & Medicaid Services imposed restrictions on payments for ESAs used in cancer patients. Last month, researchers said ESAs could boost the risk of acute leukemia in patients with myelofibrosis. (See BioWorld Today, Dec. 12, 2007.)

A month before, the FDA told Thousand Oaks, Calif.-based Amgen, which markets Aranesp and Epogen, and Bridgewater, N.J.-based Ortho Biotech Products LP, which sells Procrit under an agreement with Amgen, to strengthen the label warnings to identify tumor types - advanced breast, head and neck, lymphoid and non-small-cell lung cancer - in which ESAs could cause growth. (See BioWorld Today, Nov. 9, 2007.)

Despite all that, Amgen said wagon-circling moves, including layoffs of 12 percent to 14 percent of staffers, should let the company beat its guidance range of $4.13 to $4.23 earnings per share projected in October. Instead, earnings should reach the low end of the $4.30 to $4.50 range, estimated at the start of the year. Thomson Financial analysts predicted about $4.24 per share.

Amgen's stock (NASDAQ:AMGN) closed Tuesday at $46.33, up 94 cents.

Important as ESAs have been to Amgen's success, there's much more to the story - not only the marketed Neupogen and Neulasta, but Enbrel (etanercept) for rheumatoid arthritis, though Enbrel is likely to face pressure from the likes of Abbott's Humira (adalimumab). Neulasta (pegfilgrastim) is used to decrease infection during chemotherapy, and Neupogen is that compound's first and shorter-acting version.

A star in the firm's late-stage pipeline is denosumab. In mid-December, Amgen said the first Phase III pivotal study showed a positive effect on bone density across the skeleton in women with non-metastatic breast cancer given adjuvant aromatase inhibitor (AI) therapy. Injected twice per year subcutaneously, in the HALT Breast Cancer 135 study, denosumab increased bone density worsened by AI therapy, including in highly cortical areas of the skeleton. Increases in cortical bone showed up, too.

Analyst William Tanner, in an October research report, went so far as to call denosumab a potential "white knight." Given what could be an "irreversible contraction" plus Enbrel competition, "successful development of denosumab will likely be paramount to Amgen's long-term viability" as a major player, Tanner wrote. Amgen advanced 13 new molecules into development last year, and plans to put 10 into midstage trials in 2008.

Meanwhile, Patterson hopes to translate the work with biomarkers into successes such as Vectibix, which came out of the long-established understanding of the KRAS gene in EGFR's signaling pathway. Such victories might not be the norm, but they're possible. "We just don't know as much about human physiology as we'd like," he said.

The JPMorgan conference continues through Thursday.