Article Text

Autonomic effects of selegiline: possible cardiovascular toxicity in Parkinson’s disease
  1. University Department of Geriatric Medicine, Glan Clwyd District General Hospital, Rhyl, Denbighshire LL18 5UJ, UK
  1. Dr Esa Heinonen, PO Box 425, 20101 Turku, Finland. Telephone 00358 2 2727 383; fax 00358 2 2727 432.
  1. Orion Pharma, Clinical Research, Turku, Finland
  2. Department of Neurology, University of Oulu, Oulu, Finland
    1. Orion Pharma, Clinical Research, Turku, Finland
    2. Department of Neurology, University of Oulu, Oulu, Finland
      2. ANDREW LEES,

        Statistics from

        (1) We read with interest the article by Churchyard et al 1 showing that selegiline in combination with levodopa was associated with significant impairment of the normal reflex responses controlling blood pressure to head up tilt and standing. It was suggested by the authors that this could be one explanation of the finding of increased mortality in the combined levodopa/selegeline arm of the Parkinson’s Disease Research Group of the United Kingdom trial.2 We are currently conducting a study investigating the cardiovascular reflexes in patients recruited from a community disease register for Parkinson’s disease, diagnosis being based on identical clinical diagnostic criteria to the reported study. Patients were excluded from the study if they were known to have diseases associated with impaired autonomic system function or if they were currently being treated with drugs other than those for Parkinson’s disease. We have identified 27 such patients, five receiving selegiline and levodopa and 22 receiving levodopa. With similar techniques and equipment as the authors we have measured orthostatic blood pressure at 1 minute, Valsalva ratio, and heart rate on standing (30:15 ratio). Age, sex, duration of disease, Webster scale, and daily levodopa dose did not differ significantly between the group of patients on levodopa alone and on levodopa and selegiline combined (table). There was no significant difference in the results of autonomic function testing between the two groups with analysis of varience (ANOVA) andt tests. Both groups of patients showed falls in blood pressure on standing and this did not seem to be more pronounced in the group on levodopa plus selegiline compared with patients on levodopa alone. This finding is surprising given the much lower dose of levodopa prescribed in our study compared with that of Churchyard et al. However, our failure to show significant hypotensive interaction between selegeline and levodopa may have resulted from the lower dose of levodopa prescribed in our study. We did not specifically look at blood pressure changes during 45′ head up tilt. As yet we have not been able to retest patients in whom selegiline has been withdrawn to see if any significant changes become apparent. In view of the small sample sizes reported by Churchyard et al and our own study, we think that both sets of results should be interpreted with caution. In elderly patients with Parkinson’s disease even modest doses of levodopa seem to be associated with orthostatic hypotension. It seems unlikely that the additional hypotensive effect of selegiline could be the sole explanation of the apparent increased mortality with levodopa and selegiline compared with levodopa alone.


        Table 1

        Mean (SEM) median indices for patients taking levodopa only or levodopa and selegiline

        (2) We have read with great interest the article by Churchyard et al 1-1 that appeared in theJournal. There are several issues that we comment on.

        The study design is not scientifically sound: the trial is not randomised, the number of patients in the two groups is unequal, and half of the patients were selected from an earlier study.1-2According to the authors the two groups were matched for age, duration, and severity of the disease, and the dose of levodopa. Due to the study design the patients could not, however, be matched according to the autonomic functions, although there were no significant differences in the occurrence of clinical postural dizziness. The change in the systolic blood pressure after standing was, however, significantly lower in the selegiline group (group II) after the withdrawal than in the group receiving only levodopa (group I). Also one patient in group I had a clear hypertensive reaction (systolic blood pressure>200 mm Hg) while standing. This suggests that the autonomic function in group II even without selegiline was more compromised than in the group I, thus showing a selection bias.

        It is not, however, a new finding that orthostatic hypotension could occur when selegiline and levodopa are given together. Orthostatic hypotension is often found in late stage Parkinson’s disease due to autonomic disturbances.1-3 All dopaminergic drugs such as levodopa or dopamine agonists have been shown to cause orthostatic hypotension.1-4 In a recent double blind study Turkkaet al 1-5 prospectively followed up patients in a double blind study measuring autonomic function with a similar method to Churchyard et al 1-1 for several years. They found that with time there was significantly more orthostatic hypotension assessed with the tilt test in the selegiline plus levodopa group than placebo plus levodopa group. However, only two out of 27 patients in the selegiline group fulfilled the criteria of orthostatic hypotension.

        The mode of action of orthostatic hypotension with dopaminergic drugs is not clear. Selegiline is known to potentiate the therapeutic effects of levodopa but also its adverse effects if the dose of levodopa is not reduced sufficiently.1-6 The authors claim that selegiline would impair the noradrenaline and heart rate responses to tilt although no significant changes were detected between the groups. The authors make hasty conclusions that selegiline would be cardiotoxic with no scientific verification. If a drug causes orthostatic hypotension it does not mean that the drug would be cardiotoxic!

        Furthermore, the authors showed that the clinical condition of the patients deteriorated considerably after withdrawal of selegiline. As the disability was considerably worse after the withdrawal than during selegiline treatment, the effect of adequate dopaminergic medication on cardiovascular responses remains unknown. It is possible that adding a sufficient amount of dopaminergic medication to group II after withdrawal of selegiline would cause similar orthostatic hypotension to that found when selegiline was combined with levodopa. Indeed, Kaufmanet al 1-7 analysed patients with Parkinson’s disease with orthostatic hypotension and found no significant difference between patients treated with levodopa, dopamine agonist, or the combination of selegiline and levodopa.

        Finally, the authors suggest that their findings would be related to the metabolites of selegiline. The authors did not take into account that these amphetamine metabolites are of levoform. Furthermore, d-methamphetamine or d-amphetamine often cause increases of blood pressure, not hypotension.1-8 Similarly, there is a dose proportional increase of blood pressure in rats caused by d-p-hydroxy amphetamine or d-p-hydroxy noephedrine.1-9The authors imply that selegiline would cause withdrawal effects in humans and refer to an experimental study on rats (33cited by Churchyard et al 1-1). The authors also suggest that selegiline loses its selectivity during treatment. A postmortem study showed, however, that there was no increase of monoamine oxidase-A inhibition along with long term selegiline treatment.1-10


        1. 1-1.
        2. 1-2.
        3. 1-3.
        4. 1-4.
        5. 1-5.
        6. 1-6.
        7. 1-7.
        8. 1-8.
        9. 1-9.
        10. 1-10.

        The authors’ reply:

        Heinonen and Myllyla have raised several issues relating to our recent article which appeared in J Neurol Neurosurg Psychiatry.2-1 As they note, the study, which examined autonomic function in those on long term treatment (mean>5 years), was of necessity not randomised. However, selection bias was minimised by recruitment of all outpatients on levodopa and selegiline or levodopa without selegiline within a given time and regardless of the presence or absence of symptoms suggestive of autonomic dysfunction. That there was no bias towards more severely affected cases being placed on selegiline is strongly suggested by the excellent match between the two groups for age, duration and severity of disease, and the dose of levodopa and the greater frequency of symptomatic postural dizziness in those on levodopa alone. Moreover, as noted by Heinonen and Myllyla, the combination of selegiline and levodopa was also associated with a higher incidence of postural hypotension than levodopa and placebo in a prospective, randomised double blind study.2-2 Heinonen and Myllyla note that the percentage change in systolic blood pressure, but not the absolute levels, was lower in those previously on selegiline and levodopa three months after withdrawal of selegiline† han in those originally on levodopa alone. However, both the standing systolic blood pressure and the percentage drop were greater than when patients were taking selegiline and levodopa. That is, the hypotensive effects of standing had improved three months after stopping selegiline even though the average daily dose of levodopa had been increased (not significant) and seven (44%) patients required addition of the potentially hypotensive ergolene pergolide. This suggests that the orthostatic effect was specific for selegiline and levodopa rather than a non-specific consequence of dopamimetic treatment itself. Therefore, we think that our finding that selegiline and levodopa cause orthostatic hypotension is robust and represents an unwanted and potentially dangerous drug interaction.

        Coyle, et al examined standing systolic blood pressure, Valsalva ratio, and 30:15 ratio in 22 patients on levodopa alone and five on selegiline and levodopa. They found that standing precipitated a small and asymptomatic fall in systolic blood pressure that was similar in both groups and by contrast with our finding that levodopa and selegiline caused a greater drop in standing systolic blood pressure than levodopa alone.2-1 However, the number of those on selegiline and levodopa was small and, as Coyleet al note, the daily dose of levodopa was lower. Moreover, although both groups were apparently matched, the mean age of those on levodopa alone was seven years more than those on the two drugs and the SEM of age of those taking selegiline and levodopa was 10%, implying a much greater age range in this group and so raising the possibility that increased orthostatic hypotension in those on levodopa alone as a result of aging2-3 might have exerted a confounding influence on their data.

        Coyle et al reasonably caution that autonomic function has been examined in only a few published cases. Collectively, we2-1 and Turkka et al 2-2 have studied 43 patients on levodopa and selegiline with both groups finding that the combination of drugs was associated with greater orthostatic hypotension than levodopa alone. In a separate study of 20 patients on selegiline (unpublished data) we have again found a similar impairment of cardiovascular reflexes which occurred as a result of selegiline and levodopa treatment. Thus, we think that our findings indicate that the combination of selegiline and levodopa impairs cardiovascular control in at least some and perhaps even a majority of patients with moderately advanced or severe Parkinson’s disease.

        We agree with Heinonen and Myllyla that the causes of orthostatic hypotension induced by selegiline and levodopa are unknown. We considered two potential causes: an amphetamine effect or increasing inhibition of monoamine oxidase-type A which might be expected to increase plasma concentrations of the false transmitter tyramine, which also causes orthostatic hypotension.2-4 Selegiline may lose some of its selectivity for monoamine oxidase type B at high doses or with chronic treatment at<10 mg/day.2-5-2-7 Riedereret al found that >65% of monoamine oxidase type A activity was inhibited by treatment with selegiline at 5–10 mg/day in the brains of seven patients with Parkinson’s disease and noted that the degree of inhibition of monoamine oxidase type A seemed to rise with increasing duration of treatment.2-8 Selegiline comprises the l-isomer of deprenyl and is thought to be largely metabolised to l-amphetamine and l-metamphetamine2-9 which have less potent cardiovascular toxicity than the d-isomers. Although amphetamines cause supine hypertension, they have also caused orthostatic hypotension by impairing sympathetic function in humans.2-10 To our knowledge cerebral and plasma concentrations of the l-isomers and d-isomers of amphetamine have not been measured in those on long term selegiline treatment. Thus the possibility that high cerebral or plasma concentrations of l-amphetamine and l-metamphetamine sufficient to cause toxic cardiovascular effects as a result of long term treatment with selegiline and levodopa cannot be excluded. In the submitted study of 20 patients on selegiline already referred to we found that the impaired cardiovascular reflexes due to selegiline and levodopa treatment normalised within four to seven days of stopping selegiline. This time course was consistent with either an amphetamine effect or loss of selectivity for monoamine oxidase type B resulting in excess plasma tyramine. However, in this study selegiline seemed to cause supine hypertension as well as orthostatic hypotension. This combination would be consistent with either an amphetamine or a tyramine effect.

        The failure of patients with severe orthostatic hypotension induced by selegiline and levodopa to develop a compensatory tachycardia indicates impairment of cardiovascular reflexes2-1 and as such is consistent with significant, albeit reversible, drug toxicity by whatever mechanism.


        1. 2-1.
        2. 2-2.
        3. 2-3.
        4. 2-4.
        5. 2-5.
        6. 2-6.
        7. 2-7.
        8. 2-8.
        9. 2-9.
        10. 2-10.
        View Abstract

        Request permissions

        If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.