Emerging and Experimental Neurosurgical Treatments for Parkinson’s Disease

Scientific Background

A schematic model has emerged over the past three decades to describe the essential neurophysiology of Parkinson’s disease. This model describes two parallel and antagonistic basal ganglia circuits, the direct and indirect pathways, which are respectively responsible for initiation and inhibition of movement. Activity in both of these pathways is modulated by dopaminergic output from the substantia nigra pars compacta (SNc), so in general terms, degeneration of dopaminergic neurons in this nucleus disrupts the balance between these parallel systems for initiation and inhibition of movement, leading to the motor disturbances observed in Parkinson’s disease. ^, A more detailed understanding of these circuits is essential to understand the major trends among emerging and experimental treatments for Parkinson’s disease. (For further information on circuitry relevant for Parkinson’s disease, see Chapters 102 and 103 ; for other aspects relevant to Parkinson’s disease, see Chapters 110 , 111 , and 113 .)

The direct and indirect pathways share a common outflow circuit involving the complexed internal globus pallidus (GPi) and the reticular nucleus of the substantia nigra (SNr), which together inhibit thalamic (excitatory glutamatergic) stimulation of the motor cortex. ^,

In the direct pathway, dopamine produced in the SNc acts at D ₁ receptors in the striatum to activate GABAergic neurons, which project to and suppress the inhibitory output from the GPi and SNr, thereby (through inhibition of inhibition) disinhibiting excitatory thalamic output to the motor cortex. The net effect of dopamine on the direct pathway is hence to excite the motor cortex. ^,

In the indirect pathway, dopamine produced in the SNc acts at striatal D ₂ receptors to suppress inhibitory GABAergic projections to the external globus pallidus (GPe). GABAergic projections from the GPe inhibit the excitatory glutamatergic projections from the subthalamic nucleus (STN) to the GPi, whose inhibitory GABAergic projections in turn inhibit thalamic excitatory output to the motor cortex. In the absence of dopamine, the default logic of the indirect pathway therefore amounts to a triple-negative (dis-disinhibition, through three sets of inhibitory GABAergic projections: striatum to GPe, GPe to STN, and STN, after an intervening excitatory glutamatergic projection to GPi, to thalamus), with resultant inhibition of the motor cortex. The presence of dopamine adds a fourth negation by acting at striatal D ₂ receptors to inhibit GABAergic projections from the striatum to the GPe, thereby preventing complete inhibition of movement by the indirect pathway. ^,

Thus, according to this model, under default conditions, the direct pathway is responsible for initiation of movement, while the indirect pathway suppresses initiation of movement. The presence of dopamine favors movement by activating the direct pathway and suppressing the default inhibition of the indirect pathway. In the absence of dopamine, on the other hand, the direct pathway is less active (resulting in less initiation of movement), and the indirect pathway performs its default function, inhibiting initiation of movement. ^,

Although this model simplifies the connections among the involved nuclei and neglects the connections with other structures, it nevertheless explains some of the cardinal features of Parkinson’s disease and provides insight into the selection of anatomic targets for surgical approaches to treating the disease.

Advances made in the 1990s and 2000s shape our current understanding of the genetic and molecular pathogenesis of Parkinson’s disease. A review of the relevant mechanisms and pathways is beyond the scope of this chapter, but detailed understanding of these pathways has proven highly practical and essential to effective lesioning and deep brain stimulation (DBS) as well as to advances in gene therapies, embryonic cell therapies, and stem cell therapies for Parkinson’s disease that have entered or approached human clinical trials in recent years.

Refinements in the Understanding of Mechanisms and Anatomic Targets for Electrical Stimulation

Stereotactic approaches to the treatment of Parkinson’s disease have been used since the 1950s, and during that decade Lars Leksell and others demonstrated the efficacy of stereotactic pallidotomy and thalamotomy. Thalamotomy, in particular, was found to be effective at suppressing the parkinsonian tremor, but it had little effect on rigidity and could aggravate bradykinesia. The first landmark clinical trial demonstrating the efficacy of levodopa was published in 1961, and for the following two decades pharmacotherapy was considered the treatment modality of choice for Parkinson’s disease. Between 1985 and 1992, however, Lauri Laitinen demonstrated that pallidotomy could be used as an effective adjunct to antiparkinsonian medications in patients whose tremor, rigidity, and bradykinesias were incompletely controlled by pharmacotherapy as well as in patients with drug-induced dyskinesias. The modern era of DBS is marked by the landmark 1993 study by Alim-Louis Benabid and colleagues describing the efficacy of bilateral stimulation of the STN in three parkinsonian patients. The well-known case series of Benabid and colleagues, published in 2003, remains the pivotal long-term follow-up study in the literature on DBS, confirming the durable effectiveness of STN stimulation and characterizing some of its side effects.

A further landmark trial published in 2006 demonstrated that DBS of the STN in addition to pharmacotherapy is more effective than medical management alone in patients older than 75 years of age with severe motor complications of Parkinson’s disease. This work firmly established stereotactic implantation of electrodes for stimulation of deep brain structures into the medical and neurosurgical mainstream, and DBS is presently considered the first-line surgical procedure in the treatment of patients with advanced Parkinson’s disease.

In subsequent years, clinical work on DBS focused in part on establishing the most appropriate targets for stimulation. Mainstream use of DBS in the context of movement disorders has come to focus on three principal targets: the GPi, STN, and ventralis intermedius nucleus of the thalamus (Vim). In 2010, a major clinical trial compared outcomes from bilateral stimulation of the GPi with bilateral stimulation of the STN with respect to overall motor function, as quantified by established clinical paradigms. The study concluded that stimulation in these regions generated similar outcomes with respect to overall motor function ; unilateral pallidal stimulation has also been compared with unilateral STN stimulation, with similar results. Stimulation of the STN is, however, associated with greater reduction in the use of antiparkinsonian medications and with lower stimulation amplitudes.

At present, both the dorsolateral STN and posteroventral GPi are effectively targeted for stimulation in treating the motor symptoms of Parkinson’s disease, with no definitive evidence that one or the other is more effective. The foregoing and other studies have also examined the impact of STN and GPi stimulation on neurocognitive (nonmotor) aspects of Parkinson’s disease; a consistent finding has been an adverse effect on mood when either target is stimulated ventrally, though several authors have noted a trend toward more severe cognitive and mood disturbances with STN stimulation. Some differences have also begun to emerge with respect to motor effects, with STN being a potentially more effective target for bradykinesia and GPi a more effective target for “on-medication” dyskinesia. STN stimulation appears to reduce overall dyskinesia, primarily through reducing the overall use of antiparkinsonian medication, so some reviewers have suggested that a patient with a low threshold for dyskinesia and low preoperative probability of medication reduction might be an appropriate candidate for GPi rather than STN stimulation.

The mechanism of action of DBS remains an active area of research. It has been empirically established that high-frequency electrical stimulation, as applied in conventional DBS, has a net “lesioning” effect, inhibiting the target nucleus. Thus the finding that “functional lesioning” has equivalent efficacy in the GPi and STN is consistent with the physiologic circuit model described earlier in this chapter. ^,

In recent years, a more nuanced understanding of the mechanisms of DBS has emerged, with a recognition that DBS is both an electrical and a biochemical therapy. The fundamental effect of electrical neurostimulation occurs at the cell membrane, where an applied electric field alters transmembrane voltages so as to open voltage-gated sodium channels on the axon. The effects of electrical signal propagation, however, are almost instantaneously biochemical; when an action potential reaches the axon terminal, it induces neurotransmitter release, with effects that may be inhibitory (via GABA release) or excitatory (via glutamate release). A single stimulatory pulse at typical settings can directly trigger tens of millions of synaptic events. ^, It is now recognized that important physiologic effects occur in response to DBS at several levels, from molecules to cells to networks, and in electrical as well as biochemical domains. ^, These effects have been relevant to emerging and experimental neurosurgical treatments of Parkinson’s disease, so we address some of them here.

There are several immediate electrophysiologic effects of neurostimulation, not all of which act in concert. The cathodic phase of a stimulation pulse causes membrane depolarization at the nodes of Ranvier in the region of the stimulating electrode. The degree of depolarization depends on the shape of the axon and its orientation with respect to the electrode. If the resulting depolarization exceeds the axonal activation threshold, an action potential is triggered that propagates in both directions: orthodromically (toward the axon terminal) as well as antidromically (toward the soma). Trains of stimulation pulses, as are used in DBS, therefore not only entrain physiologic, orthodromic axonal action potentials; such pulse trains also induce antidromic action potentials that interfere with intrinsic action potentials propagating in the physiologic direction. Through these mechanisms, DBS can therefore be used to assume control over pathologic neurocircuitry; when the frequency of action potential firing induced by the external field is approximately twice the underlying average firing frequency of the neuronal population, synaptic communication of the stimulated neurons is effectively controlled by the applied stimulation. This principle crucially informs the selection of stimulation parameters in DBS. ^,

Although DBS differs from destructive lesioning, both techniques disrupt information flow between the basal ganglia and thalamocortical nuclei. This disruption in signaling corresponds to the mechanism underlying both therapeutic approaches. Net signaling from the basal ganglia to the thalamus and motor cortex is reduced in Parkinson’s disease, and this circuit persists in a state of pathologically high, hypersynchronized oscillation. In part due to the properties of the efferent synapses of the STN, low-frequency DBS (20 Hz) increases the degree of synchrony, whereas high-frequency stimulation (>70 Hz) suppresses pathologic synchrony in the basal ganglia. Notably, administration of levodopa has similar effects; this observation links the key pharmacologic, lesion-based, and stimulation-based therapeutic strategies for Parkinson’s disease at the electrophysiologic level.

Of note, both the GPi and STN are effectively targeted in the treatment of dystonia. The efficacy of DBS for dystonia was established in a study conducted from 2002 to 2004 by the Deep Brain Stimulation for Dystonia Study Group in Germany, Austria, and Norway, which evaluated 40 patients with primary generalized or segmental dystonia. All patients underwent bilateral stimulation of the ventral GPi, and stimulation was compared with sham stimulation in a double-blind fashion. The results of the study, published in 2006, demonstrate an improvement in motor function and a reduction in disability scores as quantified using standardized rating scales, and they provide class I evidence supporting the efficacy of DBS in the treatment of dystonia.

In the treatment of both parkinsonian tremor and essential tremor, the Vim has been described as the target of choice. The nucleus is organized somatotopically along its medial-to-lateral axis, with the tongue and face represented medially and the lower extremities represented laterally, so precise placement of stimulation electrodes within the nucleus can facilitate optimal tremor control. Parkinsonian patients do not typically perceive an overall benefit from stimulation of the Vim, however, because stimulation of this nucleus has little or no effect on rigidity, bradykinesia, gait, or postural dysfunction; for this reason, Vim stimulation is typically reserved for patients with essential tremor.

Advances in Neurostimulation Technology for Deep Brain Stimulation

The technology behind the implanted systems used for DBS evolved slowly in the first two decades after these systems were introduced. In recent years, however, technological progress has accelerated in response to demands within the field that the principal tool of clinical neuromodulation deliver enhanced spatial and temporal precision. In this section, we discuss technological advances in several areas: “adaptive” or “closed-loop” DBS systems (capable of electrophysiologic recording, stimulation, and using neural recordings to modify stimulation parameters), directional electrodes, improved analysis software and automation of device programming, wireless upgrades, wireless recharging, patient-facing interfaces, and MRI compatibility.