Last week, Eisai Co. Ltd. and Astex Pharmaceuticals Inc. received a complete response letter for their request to expand the use of Dacogen (decitabine), which already is approved for myelodysplastic syndromes, into acute myeloid leukemia (AML) patients 65 years of age and older.

The negative decision was expected after the FDA's Oncologic Drugs Advisory Committee voted 10 to 3, with one abstention, that data provided for Dacogen did not support a favorable benefit/risk profile. The drug did, after all, fail to significantly improve survival compared to chemotherapy – but that failure illustrates the difficulties drug developers have faced in designing a trial that both addresses the unmet need in elderly AML and satisfies the FDA's preference for comparative studies.

An Oldie But Goodie: Unmet Need

AML is a disease of the elderly, with a median age of onset of 69 years. Younger patients typically are given chemotherapy as a bridge to a bone marrow transplant. Elderly patients are less tolerant of intensive chemotherapy because of concurrent co-morbidities, leading to death in 29 percent of elderly AML patients that take it.

"For elderly, there's really an unmet medical need at this point because they can't take the traditional regimen, and even if they can take it, it doesn't produce dramatic improvement in survival," said Michael Becker, founder and senior partner at MD Becker Partners LLC.

Idamycin (idarubicin hydrochloride, Pharmacia and Upjohn) was the last drug approved as a first-line AML treatment more than two decades ago. Not that firms haven't tried. "There's been a path of destruction behind companies that have tried to develop products in the AML setting," Becker said.

Despite the unmet need, the FDA has made it clear that AML drugs must be compared to another treatment, even in the elderly patients who seemingly lack options. The agency rejected Clolar (clofarabine, Sanofi SA), Zarnestra (tipifarnib, Johnson & Johnson) and Onrigin (laromustine, Vion Pharmaceuticals) for trying to gain FDA approval with single-arm studies. (See BioWorld Insight, Sep. 7, 2009.)

The Dacogen trial had a comparison arm that allowed patients to choose, with physicians' advice, either cytarabine (also known as Ara-C for arabinofuranosyl cytidine) or supportive care. An overwhelming majority (88 percent) chose the chemotherapy, validating the FDA's view that it's the standard of care, despite the toxicity.

Even with a properly designed clinical trial that included a special protocol assessment (SPA), Dacogen failed to gain FDA approval because the drug didn't show a statistically significant increase in overall survival compared to the chemotherapy and supportive care.

Unapproved Comparison

Rather than pit its nucleoside analogue prodrug, sapacitabine, up against cytarabine, Cyclacel Pharmaceuticals Inc. decided to compare the drug candidate combined with Dacogen to Dacogen alone.

The fact that Dacogen isn't approved for AML doesn't bother Spiro Rombotis, Cyclacel's president and CEO, because the FDA signed off on its Phase III SEAMLESS trial with an SPA after it had seen Eisai's Dacogen monotherapy data.

The agency has concluded that low-dose cytarabine, off-label Vidaza (azacitidine, Celgene Corp.), best supportive care and clinical trials are the only options for elderly AML patients who are not fit for standard induction therapy.

"We had the choice of using either all of the above, in a dealer's choice design, where the physician could choose and advise the patients on what of these four options to take, or we could chose one in order to isolate the effect," Rombotis said.

Choosing just Dacogen seemed reasonable because, while Eisai failed to prove that Dacogen was better than the current options, the median survival was 2.2 months longer in the Dacogen-treated group than the group given the choice of cytarabine or supportive care.

"In essence, what that study proved is that Dacogen is not inferior to low dose Ara-C, and had it been a noninferiority study, it would have been approved," Becker said.

Cyclacel decided to make 70 years of age the cutoff for entering the SEAMLESS trial because its drug is so well tolerated. In the Phase II trial, sapacitabine was delivered for 12 cycles, three to six times the number that were tolerated or effective in trials for either Dacogen or Clolar.

"A drug that has even some toxicity, while perhaps effective, will not be tolerated. You will not be able to get enough cycles in to measure an effect on survival," Rombotis said.

When All Else Fails

Despite the FDA's apprehension about single-arm trials for elderly patients, the agency has been more lenient in using accelerated approvals for patients that have relapsed. Wyeth's (now Pfizer Inc.'s) Mylotarg (gemtuzumab ozogamicin) received an accelerated approval in 2000 based on observed response rate in 142 AML patients across three clinical trials

But Mylotarg was pulled off the market in 2010 after a confirmatory trial showed that adding the drug to chemotherapy led to a greater number of deaths compared with chemotherapy alone.

Ambit Biosciences Inc. hopes the FDA's experience with Mylotarg won't retard its chances for gaining accelerated approval for quizartinib in relapsed AML. The company is running a 333-patient single-arm trial measuring the activity of patients 60 years and older who have failed one treatment, as well as those 18 years or older who have failed two therapies.

Based on the data from the first 62 patients, Ambit is feeling pretty confident about quizartinib's chances. The drug induced a complete response in 40 percent of patients across both cohorts. "A 40-ish percent response rate is really unprecedented in this disease with a single agent," said Athena Countouriotis, Ambit's chief medical officer.

In addition to having a larger Phase II data set, Ambit will have something Mylotarg didn't have when gaining FDA approval: overall survival data, albeit without a comparator group.

Ambit's excitement for quizartinib has to do with the drug's target, FLT3. Activating mutations in the gene, which occur in 30 percent of AML patients, are linked to lower overall survival. Overexpression of the protein occurs in an additional 25 percent of AML patients.

Countouriotis likens FLT3 to the Bcr-Abl gene fusion in chronic myeloid leukemia (CML) patients. Targeted treatment by Gleevec (imatinib mesylate, Novartis AG) and Sprycel (dasatinib, Bristol-Myers Squibb Co.) has increased the five-year survival rates of CML patients in the U.S. from 30 percent in 2000 to 89 percent in 2011.

First-generation FLT3 inhibitors had mild effects on AML, but Countouriotis said she believes that the higher activity of quizartinib will lead to increased response rates and overall survival.

Data from the Phase II trial are expected in the third quarter of 2012, and while Countouriotis said she's hopeful that results will be sufficient to convince the FDA to approve the drug, Ambit is prepared to run a Phase III trial to satisfy the agency. And besides, she pointed out, unmet need or not, the FDA is going to want to know that data are coming before issuing an accelerated approval.