This week, the New England Journal of Medicine published data from the ADAGIO Phase 3 clinical trial of Israeli drugmaker Teva’s Azilect (rasagiline). The trial was designed to test whether Azilect, which is already approved for symptomatic treatment of Parkinson’s, may also have a disease-modifying effect. For early-stage PD patients, a 1-milligram dose of Azilect met all three clinical endpoints for disease modification. However, a 2-milligram dosage met only two of three endpoints. Because of this inconsistency, we cannot definitively conclude that the drug has disease-modifying effects. The Michael J. Fox Foundation spoke with Karl Kieburtz, MD, MPH, about how patients should interpret the news.

NOTE: The medical information contained in this article is for general information purposes only. The Michael J. Fox Foundation has a policy of refraining from advocating, endorsing or promoting any drug therapy, course of treatment, or specific company or institution. It is crucial that care and treatment decisions related to Parkinson’s disease and any other medical condition be made in consultation with a physician or other qualified medical professional.

MJFF: Let’s start with a few basic questions. What kind of drug is rasagiline?

KK: Rasagiline belongs to a class of drugs called monoamine oxidase inhibitors (MAO-inhibitors). To date these drugs have typically been used alone or in combination with levodopa as a symptomatic therapy for both early- and late-stage Parkinson’s patients. Another drug in the same class is called selegiline.

MAO-inhibitors are not without side effects, and there are some warnings associated with prescribing them. Some MAO-inhibitors can have interactions with certain other drugs, including selective serotonin reuptake inhibitors (SSRIs), which are widely prescribed to treat depression. They may interact with a chemical called tyramine found in foods including alcohol and certain cheeses, which can result in blood pressure changes. As a result, people on some of these drugs must adhere to a restricted diet, and this can be difficult. However, studies suggest that rasagiline may be less prone to interact with tyramine in this way, and patients enrolled in the ADAGIO trial did not follow a restricted diet.

MJFF: Before the ADAGIO trial, what evidence suggested that rasagiline might be disease-modifying?

KK: Results from both test-tube and animal experiments suggested rasagiline may have a beneficial effect in illnesses like PD. There were also preliminary results which Teva was following up on with the ADAGIO study. In 2004 the Parkinson Study Group conducted the TEMPO trial on rasagiline. Our results indicated a possible disease-modifying effect, though TEMPO’s overall duration was shorter than that of ADAGIO.

MJFF: How did the researchers design the ADAGIO trial to test for disease modification?

KK: The ADAGIO researchers employed a design known as a “delayed-start trial.” The Food and Drug Administration (FDA) has suggested that, in the absence of a biological marker, a delayed-start trial is the design most likely to satisfactorily demonstrate disease modification.

A delayed start trial works like this: At the outset of the trial, participants are divided into two groups: “early start” and “delayed start.” The early-start group receives the actual treatment, and the delayed-start group receives a placebo. Neither the patients nor the clinicians know which group is which. Results are carefully tracked for a certain period of time.

When that pre-defined time period has elapsed, the delayed-start group is switched to the actual treatment. At this point the clinicians are looking for a very specific result. If the treatment being tested is exerting a disease-modifying effect, then the improvements seen in the group that started on placebo should not be able to “catch up” with the improvements seen in the group that received actual treatment from the beginning. Even though the “delayed start” group may see some improvement once they switch from placebo to the drug, there will always be a gap between the two groups. If you can demonstrate that the gap exists and that it remains over an extended period of time, FDA has suggested that it’s reasonable to claim that your treatment is slowing the progression of the disease.

MJFF: What were the specific endpoints that the drug had to meet in the trial in order to be considered disease-modifying?

KK: There were three endpoints that had to be met over the course of the 72-week trial, which was divided into two periods of 36 weeks (nine months) each. The trial tested two different dosages of rasagiline: one milligram (1 mg) a day and two milligrams (2 mg) a day. Regardless of dosage, the early-start group received active treatment as soon as the trial began and the delayed-start group began receiving active treatment at week 36.

The first endpoint was to definitively show the gap in disease progression, as described above, between the early-start group and the delayed-start group (who were on placebo during this part of the trial) at week 36. The disease had to be less advanced in the early-start group at this halfway point of the trial.

The second endpoint was to conclusively show that there remains a gap between the two groups when analyzing the data from the entire 72 weeks of the trial — after 18 full months, enrollees’ UPDRS scores had to demonstrate that the disease was still less severe in the early-start group.

The third endpoint was to show that the rate of progression between the two groups was not converging during the last six months of the trial. If the lines were slowly converging (and the gap was slowly diminishing), it would mean that the drug might simply be having a slowly building symptomatic effect rather than having actually slowed disease. You really need this third endpoint to help you interpret the second endpoint.

The 1 mg group met all those criteria, which was consistent with a disease-modifying effect, but the 2 mg group did not. Specifically, the 2 mg group did not show an absolute difference between the early and delayed-start groups at the end of the 18 months. In fact, for this group, the placebo looked a little better.

So, even though the 1 mg group met all its endpoints, the fact that the 2 mg group did not presents a conundrum. In their last paragraph, the authors of the paper write: “Given the negative findings for the 2 mg dose, we cannot definitively conclude that rasagiline at a dose of 1 mg per day has disease-modifying effects.”

MJFF: It’s difficult to understand why the higher dose would not achieve the same results as the lower dose.

KK: That’s true. There is some speculation that this had something to do with the symptomatic power of the 2 mg dosage. These trial enrollees were very early-stage patients. Perhaps the symptomatic benefit to those who started on the 2 mg dose resulted in such minimal symptoms that by the end you literally couldn’t see a difference between the treated group and the control group.

To test this hypothesis, the researchers separated the data for those participants on both 1 mg and 2 mg whose disability was more severe — a UPDRS score of 25 points or higher, still relatively mild in relation to PD overall but more severe in relation to the enrollees in this trial — and conducted the analysis on these data separately. In this after-the-fact analysis, both the 1 mg and 2 mg groups hit all three endpoints.

But that is after-the-fact. On the pre-specified analysis, 1 mg seemed to do what was expected, and 2 mg did not. So that leaves us in a bit of an uncertain state.

Certainly the challenge for the 1 mg to hit those three endpoints was high, and it did. But is the 1 mg group a false positive or is the 2 mg group a false negative? It’s hard to know. That’s where we stand. The after-the-fact analysis tends to favor the 2 mg being a false negative, but we need to be cautious in interpreting that analysis.

MJFF: So what does this news mean to a patient — especially a newly diagnosed patient? Is there any downside to starting on a regimen of 1 mg of rasagiline as soon as you learn you have PD?

KK: This quickly becomes a broader question that incorporates some of the other things we’ve discussed — the cost and benefit analysis will look different to every patient. Some of the factors to weigh and discuss with your doctor might include:

--The side effects profile. According to this paper, there isn’t much in the way of side effects. Placebo and active looked quite similar. But every drug can have side effects, and the severity typically varies from person to person.

--For many people, there is a psychological or emotional burden associated with starting a daily medication regimen that will be with you for possibly an extended time.

--Then there’s the question of cost. I believe rasagiline would cost about $10 a day if you were paying out of pocket; of course, this will also vary depending on people’s insurance plans, deductibles and so on.

The alternative is not taking any medicine, and instead choosing to focus on non-pharmacological aspects of disease management: diet, exercise, and so on. The question of what any given individual should do largely depends on who they are.

But in terms of the classic question — How do the safety and tolerability look compared with the efficacy? — rasagiline looks pretty good.

MJFF: Clinical trials always seem to require newly diagnosed patients who have not yet taken any medication for their PD. If everyone diagnosed with Parkinson’s immediately starts on rasagiline, what are the implications for clinical research going forward?

KK: Here we could take a lesson from multiple sclerosis (MS) trials. There are MS drugs that are believed to affect disease progression, and MS researchers hold conferences and engage in international debates on the ethics of conducting placebo-controlled trials in early MS (that is, trials that require some participants not to take drugs that could slow the progression of their disease).

Is there an implication here for future PD trials? I think so. I believe that planning to do trials in early, untreated PD, is probably not a good plan for the future. In fact, my guess is that within a few years people will be saying, “What do you mean, ‘early, untreated PD’? Why would you have ‘untreated PD’?” That concept won’t even exist anymore.

MJFF: Does it stand to reason that in spite of the mixed conclusions from this paper, Teva is likely to seek a disease-modifying indication for Azilect?

KK: Clearly there is a business decision to be made here. But for now, no one outside of Teva knows what Teva is going to do.

From a marketing perspective, the company can’t advertise the drug as disease-modifying unless they get that indication from a regulatory body — the FDA or its European counterpart, EMEA. But in practice, it’s hard to say how much this matters. Whether a physician prescribes the drug to modify the signs and symptoms of the disease, or prescribes it because it is believed to be disease-modifying, the fact is that it’s available and can be prescribed today, and many patients — particularly newly diagnosed patients — are going to want it. Regardless of whether Teva is able to obtain a “disease-modifying” label, the demand for the drug is likely to be pretty high.

MJFF: We’ve focused on newly diagnosed patients a lot. Let’s talk for a minute about what these findings mean for someone with more advanced PD, or much higher UPDRS scores.

KK: This is a very important question. And it’s something that goes back to the study design — in this case, a drug was tested at a very early phase of PD for very short periods of time. The authors speak to this in their last sentence: “It will be important to determine whether benefits seen at 18 months will endure and translate into reduced cumulative disability in clinically meaningful areas such as impairment of gait and balance and cognitive dysfunction.” That’s the big question.

This study was done in people who were not on dopamine agonists, levodopa or any other anti-Parkinson’s medications. Of course, most of the world of people with Parkinson’s disease is on those medications. The timeframe during which you wouldn’t be on them is very brief, a year or two at most, early in your disease course. But people have Parkinson’s for 10, 20 years. What does this study mean for someone who’s a decade into their disease and wondering what they can do? Is this data relevant to them? Not directly.

Rasagiline has been studied in individuals who have motor fluctuations, and the addition of rasagiline to ongoing dopamine replacement therapy — either levodopa or a combination of levodopa and dopamine agonists — does result in less “off” time and fewer motor fluctuations. So the drug is useful as a symptomatic adjunct to levodopa in some who have more advanced PD. But whether it has any impact on disease progression at this stage is unknown; no delayed-start study similar to ADAGIO has been done in more advanced PD.

But there are a lot of people who fall in between these two states — very early and very advanced. People who don’t have three or four hours of off time every day, or really bad dyskinesias, who are going along and doing pretty OK for now but worrying about what’s down the pike. There are a lot of people like that, who are going to have some very reasonable questions: Will this help me? In the long run, will this keep me from getting motor fluctuations? Will this help preserve my cognition? This study does not address those questions. We don’t have data for them.

MJFF: So, at the end of the day, what should be the takeaway, if any? It sounds like each patient simply has to weigh the costs and benefits, depending on the state of his or her own disease and disability, and work closely with his or her doctor. There are no simple answers, as usual.

KK: Is there a clear, urgent message for patients in terms of treating their Parkinson’s disease? No. There’s no quick answer. There is no definitive therapeutic recommendation that arises for people with Parkinson’s disease. My expectation is that many people with early otherwise untreated PD will choose to go on this medication, after talking with their doctors and considering their other treatment options.

This is a complicated study — it’s 10 pages in the New England Journal of Medicine, with another whole supplemental data set online. That’s unusual for this journal. I’ve known about this study for years and it took me a while to get through this article. If the outcome seems complicated to patients, that’s not surprising — it is complicated, even for researchers.

In spite of the lack of a clear therapeutic recommendation, we have still learned some important things. This is the first rigorous delayed-start PD study that has been done in accordance with FDA guidelines, and it establishes a lot of interesting precedents.

First of all, 1,176 people were enrolled in a relatively expeditious fashion. Of those, 85 percent made it into the more complicated second-stage analyses. This has shown that people can be recruited and participate in such a design, in large numbers, across 129 centers in 14 countries. It’s quite feasible to do. That’s important; a lot of people had believed it would be impossible to get people to participate in such a large effort over this period of time.

Secondly, it has given us insights into trial design. People are now looking at the 18-month trial duration and asking whether it has to be done this way. Could we accomplish the same thing in 12 months?

So, first of all, it’s feasible; second of all, it’s helped trigger some new thinking (including the meeting MJFF and the Parkinson Study Group co-hosted with the FDA and the American Association of Pharmacological Scientists last year) about how to carry out a trial with this design.

We’ve learned important things, and it’s likely that we’ll see more studies of this type being done in the future.

Karl Kieburtz is Professor of Neurology and Community and Preventive Medicine at the University of Rochester Medical Center and chair of the Parkinson Study Group. He is a consultant to Teva and he consulted with authors of the NEJM paper as they wrote it.