Array BioPharma Inc. received word from partner AstraZeneca plc that a Phase II trial and two signal-searching studies with AZD6244 (ARRY-886) failed to meet their primary endpoints.

The news sent Array's shares (NASDAQ:ARRY) tumbling to new 52-week lows on Thursday, eventually closing down $3.06, or 28 percent, at $7.96.

Yet analyst Mike King at Rodman & Renshaw LLC told BioWorld Today he doesn't consider the trial results to be bad. "What I consider to be bad is the trial design," King said.

The randomized Phase II trial compared AZD6244 to chemotherapy drug Temodar (temozolomide, Schering Plough Corp.) in the first-line treatment of advanced melanoma. Initial results indicated no difference in efficacy between the two arms in the primary endpoint of progression-free survival. Array said AstraZeneca subsequently decided not to proceed into a Phase III monotherapy trial in melanoma.

King called AstraZeneca's decision a mistake and said the Phase II trial should not have attempted to prove superiority to Temodar. AZD6244's oral administration and better side effect profile, combined with what appears to be equivalent efficacy, make a sufficient case for the drug. "In today's world, most people don't expect a targeted agent to outperform chemotherapy," he said.

King questioned Array during a conference call about why the trial was designed as a superiority study rather than a non-inferiority study. Although AstraZeneca was responsible for running the study, Array executives offered the possible rationale that non-inferiority trials often require much larger patient populations and thus take longer to recruit.

In a research note, Robert W. Baird & Co. analyst Christopher Raymond also commented that "failure in melanoma is not a rarity." Just last week, Medarex Inc. and Bristol-Myers Squibb Co. announced that three pivotal melanoma trials with the CTLA-4 targeted antibody ipilimumab missed their endpoints. Yet the companies are moving forward with plans to file a biologics license application anyway, pointing to encouraging dose-response and duration of response data. (See BioWorld Today, Dec. 12, 2007.)

Other melanoma programs to stumble in Phase III include the immunotherapy Canvaxin (CancerVax Corp.), the antisense Bcl-2 inhibitor Genasense (oblimersen sodium, Genta Inc.) and the kinase inhibitor Nexavar (sorafenib, Onyx Pharmaceuticals Inc. and Bayer Pharmaceuticals Corp.).

Outside of melanoma, two monotherapy signal searching studies of AZD6244 in colorectal and non-small-cell lung cancer also failed to meet their primary endpoints of outperforming the standard of care in delaying disease progression. In the NSCLC trial, Array said AZD6244 showed "very little difference" in activity compared to chemotherapy drug Alimta (pemetrexed, Eli Lilly & Co.) - a finding that analysts again said could be viewed as positive.

Moving forward, Array said AstraZeneca continues to collect and analyze data on AZD6244 to determine the next steps for the program. The big pharma has partnered with the National Cancer Institute on six additional signal searching studies, and Array Chief Medical Officer John Yates said combination studies are "where this drug seems to be headed."

Raymond pointed out that investigational drugs are often ineffective on a stand-alone basis but prove effective when combined with chemotherapy, citing Avastin (bevacizumab, Genentech Inc.) as a prime example.

Yet King warned that combining AZD6244 with chemotherapy, instead of drilling down on the subgroup of patients most likely to respond to the drug due to specific mutations, would be "the same mistake [AstraZeneca] made with Iressa."

The EGFR kinase inhibitor Iressa (gefitinib) received accelerated approval for lung cancer but later failed a confirmatory trial, prompting AstraZeneca to launch biomarker studies to identify populations likely to respond to the drug. (See BioWorld Today, March 8, 2005.)

AZD6244, a MEK kinase inhibitor, may be more effective in patients with B-Raf mutations, which are common in melanoma and some other cancers and have been linked to increased MEK activity.

Boulder, Colo.-based Array licensed AZD6244 and certain backup compounds to AstraZeneca in 2003 in a deal potentially worth more than $95 million. Array decreased its fiscal year 2008 revenue projections from $34 million to $29 million, removing a $5 million milestone that would have been associated with the initiation of a Phase III melanoma trial. (See BioWorld Today, Dec. 19, 2003.)

As of Sept. 30, Array reported $168.2 million in cash, equivalents and marketable securities. Operating expenses during the quarter were $27.3 million, most of which went to support development of Array's six internal programs in or nearing clinical trials. Raymond said he does not view the AZD6244 data as an indictment of Array's technology platform.