Intended for healthcare professionals


Functional neurosurgery for Parkinson's disease

BMJ 1998; 316 doi: (Published 25 April 1998) Cite this as: BMJ 1998;316:1259

Has come a long way, though much remains experimental

  1. Niall Quinn, Professor of clinical neurology,
  2. Kailash Bhatia, Senior lecturer in clinical neurology

    Acquired brain lesions have long been known to modify the symptoms and signs of Parkinson's disease. After many false starts, in which various surgical lesions abolished tremor only at the expense of a hemiplegia, and with the advent of stereotactic techniques, neurosurgeons found that discrete lesions of the globus pallidus or thalamus could improve features of parkinsonism without (usually) causing a hemiplegia. Tremor appeared most responsive to thalamotomy, so this procedure became widely applied in the 1950s and 60s. Despite often permanent relief of tremor and rigidity, thalamotomy had no effect on akinesia, the core disabling feature of Parkinson's disease. Also, although unilateral surgery in this disease, which classically presents unilaterally or asymmetrically, was associated with low morbidity, as the disease progressed a second, contralateral, lesion was often made but with an unacceptably high (25%) incidence of pseudobulbar speech and swallowing difficulties. After the introduction of levodopa in 1967, the first treatment that dramatically alleviated akinesia, surgery took a dive until the mid-1980s, since when surgical approaches to treating Parkinson's disease have experienced a renaissance.

    While using a stimulating electrode to guide lesion placement for Vim thalamotomy, Benabid's group in Grenoble found that high frequency discharges could abolish tremor.1 Why not, therefore, simply insert a chronic stimulating electrode without making a destructive lesion? This technique of deep brain stimulation provided excellent control of tremor in patients with both Parkinson's disease and essential tremor.1 The mechanism underlying this effect is still debated, but since deep brain stimulation mimics the effects of a lesion, the functional result of the high frequencies used is probably inhibition, rather than stimulation, of the neurons surrounding the electrode tip. Importantly, bilateral stimulation could be applied, or stimulation applied on the second side after a previous destructive lesion, without the high morbidity previously associated with bilateral destructive lesions. Akinesia, however, was still not alleviated.

    Current models of striatal output pathways, supported by in vivo recording and lesioning studies in primates with parkinsonism induced by 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP), indicate that both the subthalamic nucleus and the internal pallidum are overactive in parkinsonism.2 In 1982 Laitinen et al reported that selective posteroventral pallidotomy produced major improvements in akinesia, rigidity, tremor, freezing, falls, speech, and also levodopa induced involuntary movements (dyskinesias).3 Many centres have since investigated pallidotomy in Parkinson's disease. Some use only imaging and stimulation with the lesioning electrode to determine lesion location, while others argue that the optimal site can be properly delineated only by additional microelectrode recordings, a debate that remains unresolved.

    Generally unilateral pallidotomy dramatically reduces contralateral, and mildly reduces ipsilateral, levodopa induced dyskinesias (incidentally, the opposite of what existing, and therefore inadequate, models would predict). The degree of benefit to off period parkinsonian features is usually much more modest, about 20-30%.4-6 Some centres have reported negligible morbidity, but others have experienced occasional deaths or stroke or other unwanted sequelae secondary to haemorrhage, infarct, or misplaced lesions. Some have reported low morbidity after bilateral (sometimes simultaneous) lesions, but others have experienced a high rate of neuropsychological (abulia) or pseudobulbar sequelae.

    Might pallidal stimulation produce equivalent benefits with lower morbidity, especially when applied bilaterally, or contralateral to a prior destructive lesion? In this rapidly evolving field, few full peer reviewed papers have yet appeared. Unilateral or bilateral pallidal stimulation also appears to reduce levodopa induced dyskinesias. However, the stimulation site giving maximal suppression of dyskinesias may be associated with worsening of akinesia, whereas electrode settings that improve akinesia are less helpful for dyskinesias.7 Rigidity seems to be improved whatever the setting, but freezing of gait may appear for the first time, or worsen. This focal effect of stimulation within the internal pallidum may limit the efficacy of this technique.

    Since the smaller subthalamic nucleus is also overactive in Parkinson's disease some centres have explored the effects of bilateral stimulation of the subthalamic nucleus (since spontaneous lesions in or near the subthalamic nucleus may cause hemiballism, surgeons have been understandably reluctant to create lesions in this structure). Such stimulation has a much greater effect on underlying parkinsonism,8 including tremor9 and freezing of gait, but usually neither improves nor worsens levodopa induced dyskinesias. Nevertheless, the antiparkinsonian effect is so striking that, in contrast to pallidotomy, levodopa dosage can be dramatically reduced, with consequent reduction of dyskinesias as well.10 Also, at least three centres, one of them in Britain,11 have made subthalamic lesions with major benefit to parkinsonism, usually without increasing dyskinesias. Most of these procedures must still be considered experimental, and many questions remain unanswered. How, therefore, should clinicians looking after patients with Parkinson's disease view their current status?

    No definitive answer is yet available. However, for severe drug resistant unilateral parkinsonian tremor a unilateral thalamotomy might still be appropriate, although several years down the line, when akinesia and dyskinesias are prominent, one may wish one had done a pallidotomy or subthalamic procedure. Prominent tremor on the second side may indicate a thalamic stimulator. Disabling dyskinesias despite optimal adjustment of medication may be helped by a unilateral pallidotomy, although the underlying off period parkinsonism may improve only moderately. For the second side, it is uncertain whether a stimulator or a second lesion is the best course. The most effective intervention seems to be a bilateral subthalamic procedure. Most experience has been gained with stimulation, but bilateral lesions can undoubtedly be effective.

    One problem in comparing these techniques is that some centres only use stimulation and others only make lesions. In all centres there is an inevitable learning curve. Lesions have the advantage that, once produced, the effect is permanent, but the morbidity might be greater than for deep brain stimulation, particularly with bilateral procedures. However, deep brain stimulation may be less effective in the pallidum. Stimulation also usually requires multiple postoperative visits to vary the choice of electrode settings, pulse width, amplitude, and frequency. There is always the (low) risk of infection and ulceration of wires through the skin, and the initial procedure entails an additional operation to implant the stimulator. Moreover, the equipment implanted in unilateral deep brain stimulation costs about £5000, £4000 of it for the pulse generator, which has to be replaced after four or five years; bilateral stimulation doubles the price. Use of stimulators is also to some degree commercially driven, whereas lesioning is not.

    Functional neurosurgery for Parkinson's disease is therefore in its second childhood. We still need to establish what targets and what techniques are indicated for what clinical pictures, but a bilateral subthalamic nucleus procedure appears potentially the most effective. Moreover, the long term efficacy of these lesioning or stimulation procedures needs to be compared with the best results from fetal nigral cell grafts and optimal medical treatment such as continuous parenteral administration of apomorphine.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    View Abstract