Share this article on:

NEWS FROM THE AAN ANNUAL MEETING: Why Did Preladenant for Parkinson's Fail in the Phase 3 Trial?

Rukovets, Olga

doi: 10.1097/01.NT.0000453584.95134.47
Back to Top | Article Outline


In a failed phase 3 trial of early Parkinson's disease patients, neither preladenant (an adenosine 2a receptor antagonist) nor rasagiline (the active control) showed significant benefit over placebo.

PHILADELPHIA—One year ago, pharmaceutical company Merck announced that it would be halting trials of preladenant, an adenosine 2a receptor (A2A) antagonist, for early Parkinson's disease (PD) after inconclusive findings on a phase 3 multicenter study. The announcement was especially disappointing after the promising results from both animal models and a phase 2 trial of the drug as an add-on therapy to levodopa (

Here, at the most recent AAN Annual Meeting in Philadelphia, Robert A. Hauser, MD, FAAN, professor of neurology, molecular pharmacology and physiology, and director of the University of South Florida Health Byrd Parkinson's Disease and Movement Disorders Center of Excellence, and colleagues presented the conflicting trial data. The phase 3 results were difficult to interpret, said Dr. Hauser, because the active control in the study, rasagiline, also did not show benefit over placebo.

“This trial demonstrates the value of an active control. In this case, the failure of rasagiline tells us that there is a problem with the execution of the trial. Rather than concluding that preladenant does not work, we conclude that the trial failed and we can't really draw conclusions about preladenant in early disease,” Dr. Hauser told Neurology Today.



So far, “in early Parkinson's disease, there was one other trial of an A2A antagonist istradefylline and that was a negative trial, so we don't know if A2A antagonists work in early PD,” he said. In patients with motor fluctuations on levodopa — moderate to advanced patients, Dr. Hauser continued, “all of the A2A antagonists have worked in phase 2, so I'm quite convinced that they work in that situation, but there's been difficulty in demonstrating that efficacy in phase 3.”

According to the data, he noted, in the early PD trial, there were striking differences in placebo effect in various regions of the world, “and this raises the issue of the ability of sites to select patients and do an accurate job in rating patients.” The placebo responses in this study were highest in Turkey, India, and Latin America — countries which also enrolled the lowest numbers of patients.

“When you go from a small number of sites to a large number of sites, a lot of noise is introduced into the system because you are using a lot more investigators,” said Dr. Hauser. “So, it's important to have rigorous entry criteria. It's important to have experienced raters, but it's also important not to put excessive pressure on sites to enroll in that it might cause them to stretch the bounds of inclusion/exclusion criteria and introduce unnecessary noise into a study.”



Back to Top | Article Outline


Janis Miyasaki, MD, FAAN, an associate clinical director of the Movement Disorders Centre at Toronto Western Hospital in Canada, told Neurology Today that these study results should be taken with “a huge rock of salt” since the patients did not have a significant response to rasagiline.

One point that was not clear from the study, Dr. Miyasaki said, was whether they were using disease duration or time since diagnosis as a measure — “diagnosis is a clearer input versus retrospectively looking at patients' recollections of when their symptoms started. That's important because, for non-dopaminergic therapy, there is potentially a time window where you are going to see response.”

According to the presentation, she added, median disease duration was .4 years, so this could also mean that “a significant number of patients will end up being SWEDD patients — Signs of Parkinson's Without Evidence of Dopamine Deficiency on a scan. They may have either dystonia or essential tremor, which would also potentially speak to why the patients aren't getting a response to the active comparator.”

Another concern Dr. Miyasaki had was that the baseline score “included both ADL and motor function. That's an unusual score to look at. Usually, we would separate out those two scores so you know exactly what the motor impairment is.”

When recruiting centers are not specialized or experienced in conducting PD trials, “the type of patient who is recruited is not as reliable,” especially since there is often pressure from drug companies to act quickly, she said.

“This sort of international, multicenter trial is hugely expensive regardless of who is doing it. My concern is that we will discourage drugs and drug classes that could be promising and could have real efficacy for our patients based on how the study is conducted,” she said. “Pharmaceutical companies are profit-driven and the speed to get drugs to market is a significant issue for them. Every week that this is delayed is a significant loss for them. I'm really concerned about the future of drug development when companies are so interested in the speed of a study that they really discount the value of expertise in conducting a trial.”

Another part of the problem, Dr. Miyasaki noted, is that academic institutions “need to recognize they are not performing well in terms of ethics reviews and contract review to allow highly specialized investigators to be competitive. This is a huge issue in academic medicine.”

Based on these results, she added, it's not clear “whether you should throw out preladenant. It's clearly proven to be effective as add-on therapy, and there's no real logical reason for why it shouldn't work as monotherapy. So given that disconnect between what we logically know and what we see in this study, you have to wonder what's going on in the study.”

Generally, clinical trials for neurologic disease can be very tough to recruit accurately because many neurologic illnesses don't have a specific test to make a diagnosis, she pointed out. “We have to find some way to address this.”

Back to Top | Article Outline


The trial included 1,007 patients who had been diagnosed with Parkinson's disease (PD) within 5 years, were not using levodopa or other agonists, and had a score of at least 10 on the Unified PD Rating Scale (UPDRS) Part III, a 14-item evaluation of motor function. These patients were randomized to preladenant 2 mg, 5 mg, or 10 mg twice-daily; rasagiline 1 mg (active control) once-daily; or placebo. The patients who were treated with placebo in the first part of the study were switched to preladenant, (5 mg, twice-daily) in the second part of the study.

The primary endpoint for this study was the change from baseline on the UPDRS parts II and III after 26 weeks. [Part II consists of 13 items, which describe activities of daily living (ADL) as well as ratings of any difficulty walking, tremor, and sensory symptoms.] Patients were 63 years old on average, and the mean baseline score on the UPDRS parts 2 and 3 was 28.5.

After 26 weeks, however, neither preladenant nor rasagiline was found to be superior to placebo.

The investigators reported a dose-ordered response. The patients in the preladenant 2 mg twice-daily group actually worsened, showing a significant change from baseline of 0.30 and a 2.60 difference when compared with placebo (p = .0033). The preladenant 5 mg twice-daily group improved by 1 point from baseline and a 1.30 difference from placebo (p = .1382). The preladenant 10 mg twice-daily arm also improved by 1.8 points with a difference of 0.40 from placebo (p = .6378). In the 1 mg once-daily rasagiline group, improvement from baseline was not significant. The placebo group had the biggest overall change from baseline of 2.2 points.

Study author Robert A. Hauser, MD, FAAN, a professor of neurology, molecular pharmacology and physiology, and director of the University of South Florida Health Byrd Parkinson's Disease and Movement Disorders Center of Excellence, said: “The main result was that the trial did not find efficacy for preladenant as monotherapy in PD, but the trial also was unable to identify benefit for rasagiline, the active control. This is therefore considered a failed trial and it is difficult to interpret these results and what they mean for preladenant in early Parkinson's disease.”

—Olga Rukovets

Back to Top | Article Outline


•. AAN Annual Meeting abstracts from the failed phase 3 trial of preladenant.
    •. Hauser RA, Cantillon M, Pourcher E, et al. Preladenant in patients with Parkinson's disease and motor fluctuations: a phase 2, double-blind, randomised trial. Lancet Neurol 2011; 10(3):221–9.
      •. More research on Parkinson's disease in Neurology Today.
        •. Neurology studies of A2A receptors for the treatment of Parkinson's disease.
          © 2014 American Academy of Neurology