For a condition as complicated as Parkinson's disease, it is not an easy task to show that a drug slows disease progression instead of simply ameliorates symptoms. Warren Olanow et al.1 designed a trial to test this question for the drug rasagiline. The researchers used a 'delayed start' design, in which one group of subjects received placebo and another group received rasagiline for 36 weeks. After that, both groups were switched to rasagiline for another 36 weeks. The researchers looked for differences in disease severity between the two groups to assess whether the drug was able to slow progression. The findings are not clear cut1. Rasagiline slowed progression of Parkinson's disease at one dose but not at another. What does the data mean?

Serge Przedborski:

Current medications for Parkinson's disease ameliorate its motor manifestations but do not halt or retard the neurodegenerative process. Preclinical studies have led Parkinson's disease researchers to believe that inhibition of monoamine oxidase (MAO), an enzyme involved in dopamine catabolism, may mitigate the demise of dopaminergic neurons. In a new clinical trial, an international group of experts tested this idea using rasagiline, a MAO type B inhibitor, in untreated subjects with Parkinson's disease over 72 weeks using a double-blind delay-start design. The authors were cautiously optimistic in concluding that 1 mg, but not 2 mg, rasagiline per day may be disease modifying1.

Parkinson's disease is a chronic disorder; hence, even if rasagiline blocks neurodegeneration totally, the difference in the clinical scores between rasagiline and placebo may still be small after 72 weeks. Thus, given the variability in the data shown in this study, a question may be whether the 1-mg outcome is a false positive (type 2 error) or whether the 2-mg outcome is a false negative (type 1 error). Also, with such a small effect of rasagiline, are the researchers certain that the difference in the number of subjects who dropped out of the trial early or the potential differences in the profiles of the subjects transferred early from phase 1 to phase 2 did not contribute to the distinct outcomes of the 1-mg and 2-mg rasagiline doses?

Another difficulty in interpreting the results may stem from scoring the subjects with the Unified Parkinson's Disease Rating Scale (UPDRS); is it known whether a two-point change in this nonlinear tool in a subject with a low score is equivalent to that in a subject with a high score?

The authors should be praised for rightfully stressing that “the study results must be interpreted with caution.” “Caution” is indeed the word.

Professor of Neurology and Pathology, Columbia University, New York, New York, USA.

Patrik Brundin:

This study raises both hope and questions. The claim that a 1-mg (but not 2-mg) daily dose of rasagiline has disease-modifying effects in people with Parkinson's disease is exciting, whereas the underlying molecular mechanisms and the dose specificity both still need clarification.

The difference of 1.7 points in the Unified Parkinson's Disease Rating Scale (UPDRS) between the early-start and 36-week delayed-start is statistically significant, but critics have said it is too small to be clinically meaningful. That critique, however, does not take into account a key implication of the study—that several years, as opposed to months, of rasagiline treatment might retard disease progression at a clinically meaningful level compared to placebo. But a trial with a delayed start of, for instance, 10–15 years would be practically, financially and ethically difficult. For the time being, with a 38% reduction in the rate of clinical decline, the study hints that long-term therapy with rasagiline could be clinically valuable.

As for future studies, more specific and objective outcome measures than UPDRS, and powerful brain imaging tools, are needed to assess changes in disease progression.

Professor of Neuroscience, Lund University, Sweden.

Ted M Dawson:

Many clinical studies in Parkinson's disease have failed to provide definitive evidence for neuroprotection in Parkinson's disease. Unfortunately, the current study is no exception. Based on the putative neuroprotective mechanism of rasagiline, the finding that two doses of rasagiline yield varying results does not fit with a strong neuroprotective effect. Even if the low dose of rasagiline has a modest disease-modifying effect, it is relatively small. Thus, the Parkinson's disease community still waits for a robust disease-modifying agent.

This study illustrates a fundamental problem in the field, in that all Parkinson's disease neuroprotective trials to date have been based on acute-intoxication models of Parkinson's disease; in these models, mice, rats or nonhuman primates are exposed to 6-hydroxydopamine or MPTP (1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine), which selectively kill dopamine neurons. Much like prior neuroprotection trials in Parkinson's disease, rasagiline and related compounds were protective in these models.

It may be time to rethink our strategy and investments in Parkinson's disease neuroprotection trials and move beyond trials based on such limited animal models.

Professor of Neurodegenerative Diseases, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA.