Results of a phase 2 randomized trial suggest that tarenflurbil, a selective amyloid-beta–lowering agent, was well tolerated and showed some indication of a dose-related effect on measures of daily activities and global function in mild Alzheimer's disease (AD), although without a significant effect on cognition.

In moderate AD, however, there were no positive effects and some evidence of a negative effect on global function.

"These findings justify phase 3 studies of tarenflurbil at the 800-mg twice-daily dose in patients with mild AD," the researchers, led by Gordon K. Wilcock, DM, from Radcliffe Hospital, University of Oxford, in the United Kingdom, conclude.

This separate assessment of the mild and moderate groups was prespecified and assessed after it was found that there was a significant interaction between baseline Mini-Mental State Examination (MMSE) scores and treatment group for 2 of the 3 primary outcomes. The researchers are confident that this is the most appropriate way to interpret their results, they write; "however, the significant treatment interaction in the primary models and reliance on a subgroup analysis mean that this phase 2 study is unlikely to be elevated to pivotal status in support of an efficacy claim."

The trial showed that the drug was well tolerated and provided a "clear indication" of the dose that should be used in the phase 3 trials, Dr. Wilcock told Medscape Neurology & Neurosurgery. "As far as phase 2 studies being a foundation for phase 3 studies, it did really well; that's my personal opinion."

Results of the study, funded by Myriad Pharmaceuticals, are published online April 30 in Lancet Neurology.

Antiamyloid Strategy

Deposition of amyloid-beta in the brain is a neuropathological hallmark of Alzheimer's disease and a potential cause of neuronal damage, the authors write. The process is thought be initiated by the 42-amino-acid-peptide Aβ₄₂ that is found in both soluble oligomeric forms and in neuritic and diffuse plaques, they explain. Tarenflurbil, formerly known as the R-isomer of flurbiprofen, is a nonsteroidal anti-inflammatory agent that also acts to selectively lower Aβ₄₂.

In this placebo-controlled phase 2 multicenter trial carried out in Canada and the United Kingdom, 210 patients with mild to moderate AD living in the community were randomized to receive tarenflurbil 800 mg or 400 mg twice daily or placebo for 12 months. Baseline MMSE scores ranged from 15 to 19 (moderate AD) and 20 to 26 (mild AD). Primary efficacy outcomes were results on the AD assessment scale cognitive subscale (ADAS-cog); the Alzheimer's Disease Cooperative Study activities of daily living scale (ADCS-ADL) and the clinical dementia rating sum of boxes (CDR-sb).

In a further 12-month extended-treatment phase, those who had been on tarenflurbil continued on their randomized dose, but placebo patients were randomized to 1 of the 2 doses of the drug.

Prespecified Interaction Analysis

The authors note that prespecified interaction analysis revealed that patients with mild AD and moderate AD responded differently to treatment on 2 of the 3 primary analyses, the ADAS-cog and the ADCS-ADL (P < .10), so these groups were then analyzed separately.

In patients with mild AD, treatment with 800 mg of tarenflurbil twice daily was associated with significant benefit that increased over time, on global function CDR-sb, a 35.7% effect size, and activities of daily living on the ADCS-ADL, with a 46.4% effect size, even though most of these patients were already receiving stable doses of acetylcholinesterase inhibitors. There was no significant effect on cognition, although the slope over time in cognitive decline was in the right direction, Dr. Wilcock noted.

In those with moderate AD, however, there were no significant effects with the same dose of tarenflurbil on either the ADCS-ADL or the ADAS-cog and a negative effect on the CDR-sb, a 52% decline (P = .003).

The drug was safe and generally well tolerated up to 24 months of treatment, they report. The most common adverse effects of treatment were diarrhea, nausea, and dizziness.

Patients with mild AD who were treated the whole 24 months with the 800-mg tarenflurbil dose had lower rates of decline on all 3 outcomes than those treated first with placebo for 12 months and then switched to either dose of tarenflurbil.

"This wasn't meant to be an efficacy study," Dr. Wilcock noted, "but any efficacy that came out of it was obviously very welcome to us and showed that it was helpful in slowing down some aspects of the disease progression."

As for the phase 3 program, 2 studies are already under way with the aim of treating more than 2000 patients in total, he said. Dr. Wilcock is the principal investigator for 1 of these studies. Based on the phase 2 results, he said, "I'm hopeful, and I hope in the next 12 months we will know from the phase 3 results whether this approach is useful."

Reason for Optimism

In a Reflection and Reaction article accompanying the paper, Paul S. Aisen, MD, from the University of California, San Diego, calls it "important in its assessment of the drug, but it also shows the methodological hurdles in the path toward disease-modifying therapies for AD."

Among these hurdles, illustrated by the current study, he notes, is that AD-modifying activity might not be evident in a phase 2 trial. Disease-modifying interventions are expected to slow progressive decline on cognitive and other measures without any short-term effect, he writes. "Proof of efficacy, or even proof of mechanism, might be impossible in a phase 2 study."

The primary analysis of this study failed, he writes, but planned analyses pointed toward a separate analysis on the basis of baseline cognitive function. Although the data do not prove efficacy for tarenflurbil in mild AD, it does justify continuation of a large phase 3 development program.

"With the need so enormous and the potential benefit suggested (although not proven) by these phase 2 results, the effort is indeed justified, despite the substantial uncertainty," Dr. Aisen writes. "In a few months, we will learn whether tarenflurbil will be the first antiamyloid intervention to be efficacious in a pivotal trial.

"The best news is that several other promising antiamyloid programs are moving forward, despite the methodological difficulties of seeing clear efficacy in phase 2 studies," Dr. Aisen concludes. Biomarker techniques are also improving, allowing more rational selection of drug doses to target secretases, advances that should increase the chances for success. "The realistic possibility that pharmacological control over the amyloid cascade will be achieved, and the enormous effect this would have on world health, should continue to drive these efforts."

Methodological issues also cropped up during a phase 3 trial of another antiamyloid compound called tramiprosate (Alzhemed, Neurochem), in which Dr. Aisen participated. The trial was ultimately negative, but not before investigators ran into difficulty with data analysis due to what they reported was unexpected variation between participating sites.