23 November 2005. When it comes to human tests of prospective Alzheimer disease treatments, researchers have learned to be humble. They are familiar with having their expectations raised by solid epidemiology, intriguing biological rationales, and animal model data, only to see them dashed in subsequent clinical trials. This problem has bedeviled trials of nonsteroidal anti-inflammatory drugs (NSAIDs), which either failed to show an effect (e.g., rofecoxib and naproxen [see ARF related news story], and nimesulide [see ARF related news story]), were too toxic (diclofenac [Scharf et al., 1999] and indomethacine [Rogers et al., 1993]), or cut short over questions of cardiovascular side effects (see ARF related news story). Before this backdrop, then, the latest NSAID trial seems worth writing home about, as its results were modestly encouraging.
At the 35th Annual Meeting of the Society for Neuroscience, held last week in Washington, D.C., Sandra Black of the University of Toronto presented phase 2 results of a 31-center trial of R-flurbiprofen, conducted in Canada under Black’s supervision and in the UK under Gordon Wilcock at the University of Bristol. This single enantiomer of flurbiprofen emerged as a candidate from research in the laboratories of Greg Cole and Todd Golde, who tried to find compounds that minimize gastrointestinal toxicity while taking advantage of possible γ-secretase modulating and anti-inflammatory effects (see Morihara et al, 2002; Eriksen et al., 2003; Gasparini et al., 2005). Flurbiprofen is the only approved NSAID still in the running for an AD therapy, though the one discussed here is a version trademarked by Myriad Pharmaceuticals.
At the conference, Black reported that the trial had enrolled 207 people with mild and moderate AD and randomized them into a 400 mg twice daily, 800 mg twice daily regimen, or placebo for one year. After that, 86 Canadian participants enrolled in a blinded follow-on study, and Black reported 6-month data from that, too. Most of the participants were also taking cholinesterase inhibitors. Primary endpoints of the trial included performance on the ADAS-cog battery of neuropsychology tests, the activities of daily living scale used by the ADCS, and the clinical dementia rating (CDR) scale.
Patients who had moderate AD at study outset showed no improvement. People with mild AD showed small improvements, but they were not statistically significant overall. On the cognitive tests, people with mild AD on placebo declined, those on 400 BID R-flurbiprofen declined a bit more slowly, and people on the high dose declined more slowly still. Some of the patients in the follow-on group appeared to regain some cognitive function previously lost to disease, but this effect was small and preliminary. On the assessments of daily activities and global function, the improved performance following high-dose treatment did reach statistical significance. This applied particularly to those patients who had the highest levels of drug in their bloodstream. This phase 2 trial did not collect CSF samples to track distribution of the drug or its effect on Aβ levels, but the next trial will, Black said.
The drug was safe and effective enough for the trial sponsor, Myriad Pharmaceuticals, to begin a phase 3 trial of the compound, trademarked as Flurizan, at 130 centers in the U.S. This trial will compare the 800 mg BID dose to placebo for a year and is currently enrolling patients (see Clinical Trials.gov for more information).
This is not to say that there were no side effects in the phase 2 study. There were, and the list includes transient eosinophilia, anemia, blood pressure elevation, and mild rash. The 1,600 per diem dose necessary to see a significant effect is pushing the limit of the tolerable dose established previously, so a positive outcome of this phase 3 trial is not guaranteed. While R-flurbiprofen shows its colors in people, researchers in academia and industry are continuing to search for new analogs that modulate γ-secretase more strongly without affecting any of its other important substrates (see Peretto et al., 2005).—Gabrielle Strobel.