One of the controlled clinical trials with contrasting data was conducted at Mt. Sinai Medical Center in New York.19 Two neurologists who participated in the Mt. Sinai study-C.W. Olanow and James H. Godbold-were among the letter writers who presented alternative data to the findings of the British study. They wrote:
"We performed a prospective double blind study in 101 comparable patients with Parkinson's disease. They were randomly assigned to receive selegiline or placebo plus levodopa or bromocriptine. After five years, only eight deaths had occurred (five in patients taking placebo and three in patients taking selegiline)."19
The British Mortality Data
The finding of higher mortality in the deprenyl/L-Dopa group-the first of its kind-must be scrutinized carefully, as it has serious implications for Parkinson's disease patients and others who have relied on the findings of previous studies.
A look at the mortality data in the British study leads to several unanswered questions. First is the question of when the deaths referred to in the BMJ paper occurred. According to the authors:
"No significant difference in mortality was present at the time of the three year analysis. A further interim analysis in December 1994 showed that the mortality in the group treated with levodopa alone was significantly different from the rate in the group given levodopa in combination with selegiline."
Since the BMJ paper, which was published in Dec. 1995, deals with 5-6 years of data, it seems as if the three-year analysis referred to in the paper occurred in 1992 or 1993. Thus, it appears that the 60% greater mortality reported for the deprenyl/L-Dopa group occurred between 1992/1993 and 1994/1995.
However, little information is reported about this critical period. Because of their recommendation that the patients in the group stop taking deprenyl, it is clear that the authors presume that deprenyl was responsible for the extra deaths, with little or no consideration for alternative explanations. Yet, there were things going on during this period that suggest other possible explanations for the extra deaths.
High Drop-Out Rate, Questions About Compliance
First is the fact that many patients left the study during the period in question because they were either lost to follow-up,violated the protocol, deteriorated too much, had adverse reactions, or had their diagnoses revised. In all, 129 (52%) of the L-Dopa only group and 123 (45%) of the deprenyl/L-Dopa group withdrew from the study. This kind of drop out rate raises the possibility that other, unreported drugs could have been taken by the patients during the study, which could have played a role in causing the extra deaths. An additional concern not addressed in the paper is compliance. How many of the patients complied and how was compliance determined?
The authors report that-since it was an open trial-they reassigned patients to other groups when it appeared they were having problems with the treatment regimen in their original group. As they put it:
"Patients who were unable to tolerate the trial drug or gain useful functional improvement (initial improvement of 20% or more in rating scales and continuing improvement above baseline levels of disability) could either be randomized again to a different arm of the trial or withdrawn. Patients have been considered in this report only in relation to their original randomization."
With this kind of change going on at 93 different medical centers, it's entirely possible that there were mistakes made in tabulating and analyzing the mortality data. Yet not a word is said in the paper about what they did to protect against such mistakes.
Differences In Follow-Up
The authors also present data that could explain some of the extra mortality in the deprenyl group. (Table 2) They report 1,372.6 patient years of follow-up for the patients receiving L-Dopa only (treatment arm 1), and 1,500.5 patient years of follow-up for the patients receiving deprenyl and L-Dopa.
Since the deprenyl/L-Dopa group was followed longer than the L-Dopa only group, it's not surprising that more of them died. Nowhere in the paper does it say how many of the extra deaths occurred during the longer follow-up period for the patients receiving deprenyl/L-Dopa.
The Reported Causes Of Death
Another unsettling factor is the authors report that many of the patients in the deprenyl/L-Dopa group died of "Parkinson's Disease". There were 45 reported deaths in the L-Dopa group compared to 76 in the deprenyl/L-Dopa group, but 26 (52%) of those deaths were attributed to Parkinson's disease compared to only 7 (16%) in the L-Dopa group. (Table 3)
The problem is that Parkinson's patients doesn't usually die of Parkinson's disease. In most cases, the disease weakens patients to the extent that they die of other causes such as heart attack, stroke, or cancer. In one of the letters to the editor in the March 16 issue of the British Medical Journal, scientists from the Neurodegenerative Diseases Research Centre at King's College in London wrote:
"Most of the excess deaths in arm 2 was directly attributed to Parkinson's disease itself, this information being obtained from death certificates. We were surprised to find that Parkinson's disease featured as a primary case of death, as most patients with the disease die of its complications."
Instead of trying to explain this anomaly, the authors responded as follows:
"The precise cause of the increased mortality in arm 2 remains to be determined. Selegiline increased the number of early adverse events, and it is conceivable that it may have deleterious effects on the cardiovascular and cerebrovascular system. For example, in the DATATOP study a higher incidence of cardiac rhythm disturbance was reported in patients treated with selegiline."
Instead of answering a reasonable question about data in their own study, the British scientists responded with data from another study (the DATATOP trial). In fact, data from their own study contradicts their suggestion that deprenyl might have deleterious effects on the cardio- and cerebrovascular health. Table 3 (from the British study) shows 20 cardiovascular and cerebrovascular deaths in the L-Dopa only group, representing 44% of the deaths reported for this group, and 21 such deaths in the deprenyl\L-Dopa group, representing only 28% of the deaths reported for this group. Hardly evidence that deprenyl may have been causing heart attacks and strokes!
Recent Papers From The DATATOP Study
Shortly after the results of the British study were published, two papers from the DATATOP study were published in the January 1996 issue of Annals Of Neurology.
DATATOP is an acronym for Deprenyl and Tocopherol Antioxidant Therapy of Parkinsonism. It is the largest clinical trial ever conducted in early-stage, previously untreated Parkinson's patients, including (originally) 800 patients at 25 medical centers in the U.S. and Canada. The DATATOP trial started as a controlled, double-blind study including four groups: deprenyl (10 mg a day), vitamin E (2,000 units a day), both deprenyl and vitamin E at these doses, and placebo.
The clinical endpoint in the DATATOP trial was the length of time it took for patients to develop severe enough symptoms to require L-Dopa therapy. The decision to prescribe L-Dopa was left to the discretion of the treating physicians, who used well-established tests to measure the degree of disability in their patients.
About 18 months after the study started, it was apparent (in spite of the blinding process) that the deprenyl patients were doing much better than the patients in the other groups. As a result, it was decided to break the code, end the study, analyze and publish the data, and enter those patients who wished to continue into an open clinical trial in which every patient would receive deprenyl.
Protocol For The Second Phase Of The DATATOP Trial
Ten months after the original DATATOP trial began, a new protocol was instituted for the patients who already needed L-Dopa. These patients stopped taking anything for a month while they were being evaluated. They were then restarted on their original medications and then, two weeks later, were started on standard L-Dopa therapy (L-Dopa plus Carbidopa). A month later, they came in for a follow-up visit and then again at 3-month intervals. L-Dopa dosage was adjusted during the rest of the trial by the treating physicians for the 371 patients enrolled in this phase of the trial.
The other arm of the second phase of the trial was started after the original trial was ended because of the positive results in the patients receiving deprenyl. At the time, there were 423 patients who did not yet require L-Dopa therapy. Of these, 367 agreed to be withdrawn from their assigned treatments (deprenyl and/or vitamin E) for up to 2 months while they were being evaluated.
During this 2-month withdrawal period, four patients got worse enough to require L-Dopa, 1 withdrew from the trial, and 52 were given deprenyl because of increased disability from Parkinson's disease. The remaining 310 patients agreed to be restarted on their previous regimen, with those who had been taking vitamin E or placebo given an additional 10 mg a day of deprenyl.