Neurostimulation for Parkinson Disease

The treatment of advanced stages of Parkinson disease (PD) is a major challenge for modern medicine. This devastating movement disorder is characterized by severe fluctuations of the patient's mobility between a state of sometimes complete akinesia and dyskinetic hypermobility often being equally bothersome for the patient. These symptoms frequently are accompanied by wide fluctuating nonmotor symptoms such as anxiety, depression, hallucinations, and autonomic disturbances. While the early stages of PD usually can be sufficiently treated, this condition presents an ongoing therapeutic problem. After the revolutionary discovery of levodopa as the most efficient drug therapy¹ during the early stages of the disease, it has been found that long-term complications limit the use of this drug.² The great hope for transplantation of embryonic cells to restore endogenous L-dopa production was met with initial success³ but has turned into a disappointment, with 2 double-blind studies demonstrating lack of efficacy when rigorously tested against sham surgery.^{4 - 5} The proof-of-principle studies of gene therapy for PD likewise created much hope⁶ but also failed to show benefit in a sham surgery–controlled trial, based on release of preliminary data.⁷ A preliminary conclusion for many experts is that placebo effects are overwhelming for the invasive therapies of PD.

Deep brain stimulation seems to be the exception to this rule. The basic principle of this treatment is the modulation of the so-called basal ganglia loop, a circuit connecting several large cell groups in the depth of the brain with the cerebral cortex. This loop serves as a substrate for the selection and preparation of executive behavioral programs assisting many motor, cognitive, and emotional tasks of the cortex.⁸ The way different nerve cell groups in this loop collaborate is not yet fully clear, but this complex regulatory network consists of cell groups that activate or inhibit the subsequent nerve cell group within the pathway. The substantia nigra activates this circuit.

The major pathology of advanced PD is the degeneration of the substantia nigra, which leads to a lack of net activation of this loop. The complex architecture of this loop with mostly inhibitory connections between the different nuclei leads to underactive or hyperactive cell groups. Two of them, the subthalamic nucleus and the pallidum internum, are predominantly hyperactive.⁹ Deep brain stimulation with continuous stimulation between 110 and 180 Hz appears to inhibit such overactive cell groups, and this seems to be the major therapeutic effect of deep brain stimulation. These 2 latter nuclei are therefore the targets for this therapy, with most procedures performed in the subthalamic nucleus.⁹ Moreover, treatment with deep brain stimulation is associated with fewer adverse effects than classic lesional surgery,¹⁰ and the stimulation parameters can be adjusted with an external programming device to achieve a better treatment effect.

In this issue of JAMA, Weaver and colleagues¹¹ report the results of a controlled trial of deep brain stimulation for advanced PD, in which they tested deep brain stimulation of the subthalamic nucleus or the globus pallidus internus against best medical treatment. The novel aspects of the study are the following: (1) this is the largest study of deep brain stimulation conducted to date, including 255 patients with a rigorous randomization; (2) this is the first study that has randomized patients to best medical treatment or deep brain stimulation of 1 of the 2 known targets (the subthalamic nucleus or the globus pallidus internus) (2:1:1), although the 2 groups of surgical patients are combined in this report; (3) the primary outcome variable was the time patients spent with unimpeded motor function (“on” time) documented with diaries; (4) the quality of the clinical improvement was measured by blinded neurologists; and (5) important secondary outcomes include quality of life, neuropsychological and psychiatric outcomes, and adverse effects.

Thus far, only one other large-scale randomized trial has tested deep brain stimulation of the subthalamic nucleus against best medical treatment, for a 6-month period in 156 patients.¹² The 2 studies report almost the same disease severity at baseline: the Hoehn and Yahr stages (which grades PD between 1-5) were 3.3 for the present study and 3.7 for the former study, the mean duration of the disease was 12 and 13 years, the motor examination in the immobile condition showed 43/108 and 47/108 points on the appropriate scale, and the levodopa equivalence dosage was 1295 and 1175 mg, respectively. Both studies also demonstrated similar outcomes for the stimulated groups. The “on” time (periods of good symptom control or unimpeded motor function) was increased by 4.5 and 4.4 hours, respectively, the daily time with at least acceptable mobility increased to 10 and 12 hours, and the quality of the worst state improved to similar values of 30/108 and 28/108. The quality of life improved by 17% and 23%. These changes represent all large effects. In summary, the 2 trials show similar results with respect to efficacy.

However, the study by Weaver et al¹¹ also documents important adverse effects of this procedure. At least 1 serious adverse event occurred in 49 patients who received deep brain stimulation therapy compared with 15 patients who received best medical therapy. The cumulative risk for device-related problems was 10% at 6 months, and stimulation-related problems, which mostly can be corrected, were even more frequent. There was one surgery-related death in this trial population,¹¹ but the risk for surgical-related death or permanent disability due to the procedure can only be reliably estimated in large (>1000) patient cohorts, in which these adverse events were found to be 0.4% and 1%, respectively.¹³ Similar to other prospective studies, Weaver et al found that psychiatric adverse effects thought to be related to deep brain stimulation were not more frequent than in the best medical treatment group. Nevertheless, cumulative data from 55 centers have shown that the suicide rate following deep brain stimulation was 13 times higher in the first postoperative year (24/5311 patients) and doubled after 4 years.¹⁴ Risk factors have been identified but prevention algorithms still need to be developed.

The findings of neuropsychological adverse effects reported by Weaver et al¹¹ are similar to the findings of earlier studies.¹⁵ Verbal fluency, working memory, and processing speed seem to be affected by deep brain stimulation. However, the treatment does not induce a general cognitive decline,¹⁵ which has been suspected particularly for elderly patients in small open studies.

Although deep brain stimulation is the most important innovation for treatment of advanced PD since the discovery of levodopa, many questions are still unanswered. For instance, the optimal timing for the implantation is unknown. The majority of patients undergo deep brain stimulation surgery more than 10 years after disease onset when the patients are already incapable of working and when the disease-related psychosocial decline has already begun. As quality of life is improved with this treatment it may improve psychosocial functioning in general for these advanced stages. With the aging of the general population, PD will become even more common and patients with PD will get older. Therefore, the present results showing similar efficacy and tolerability of deep brain stimulation in younger and older patients must be replicated because it is at variance with some other reports demonstrating lower rates of operative and postoperative complications in younger patients.¹³ Postoperative long-term outcomes have been published up to 5 years postoperatively but longer-lasting observations are certainly needed. Moreover, the relative advantages of the 2 deep brain stimulation targets remain undefined. In small uncontrolled studies stimulation of the globus pallidus internus had a diminishing effect over years but had less neuropsychiatric adverse effects than stimulation of the subthalamic nucleus.¹⁶ The 2-year follow-up of the study by Weaver et al¹¹ will answer some of these questions.

Overall the results of this important study by Weaver et al¹¹ have convincingly confirmed the 6-month efficacy of deep brain stimulation for advanced PD in the largest patient group studied thus far. However, this study, along with previous research on this therapy, shows that such progress cannot be made without costs in terms of adverse effects.