Neurostimulation in Parkinson’s Disease (PD)
Parkinson’s disease (PD) is a neurological disorder that has a wide spectrum of symptoms. The most notorious are the dysfunctions in movement coordination, but also including cognitive abnormalities such as dementia.
Tremor, bradykinesia and rigidity are the three characteristics commonly associated with the motor disorder in PD. Tremor, in particular, has been studied in great detail from experimental and theoretical perspectives. It is an involuntary, almost periodic oscillation of a limb (or any other part of the body) that becomes a disabling feature in PD.
It should be noted that physiologic tremor is present in all PD patients, but it does not always make it a disturbing factor in movement coordination.
The molecular basis of PD is relatively well known. It has its origins in the degeneration of the dopaminergic neurons of the substantia nigra (Figure 1).
Figure 1. Left: Normal. Right: Parkinson’s. Degeneration of dopamine neurons in substantia nigra is hallmark sign of Parkinson’s disease.
This causes a dysfunction in the areas where these cells project. The net, final effect of the dopaminergic loss is an enhanced synchronization in certain brain regions.
The consequences of the loss of dopaminergic transmission in the basal ganglia (Figure 2), were found to be as follows:
Figure 2. The basal ganglia are a group of structures found deep within the cerebral hemispheres.
Firstly, the response of basal ganglia cells to cortical inputs was found to be enhanced after the dopaminergic lesion.
Thirdly, evidence obtained in rodent models of Parkinson’s disease (PD) showed the consequent enhancement of rhythmic synchronization between the basal ganglia and cortical networks
Hence, after the loss of dopaminergic transmission in the basal ganglia, the result is an increase in inhibitory (GABAergic) transmission from the internal globus pallidus to the thalamus (Figure 2).
The enhanced inhibition to the thalamic neurons will have distinct results. This depends upon the conditions in which the thalamic cells are in the moment they receive the synaptic input. In some cases it will excite thalamic neurons (this apparent paradoxical effect was explained in the blog Epilepsy III).
What normally occurs is, that a rhythmic excitation from the thalamus to the brain cortex emerges. This will force a particular oscillation in the cortex that may disturb normal motor control. The rhythm is normally about 3 to 6 Hz resting, which is the tremor frequency in PD patients.
Thus, we have a relatively simple trigger, the loss of the neurotransmitter dopamine. It ends up enhancing excitability and promoting more synchrony than normal in particular brain areas.
This is the typical scenario underlying many neuropathological syndromes: increased excitability that promotes hypersynchrony.
If they occur in the hippocampus and temporal lobe areas, epileptic seizures may materialise. If they appear in the basal ganglia and thalamus, Parkinson’s disease (PD) tremor will emerge (Figure 3).
Figure 3. Depending on where hypersynchrony occurs, the subject will experience either epileptic seizures (hippocampus) or Parkinson’s disease (basal ganglia, thalamus).
Once again, as mentioned in the epilepsy blogs I, II and III, different molecular factors cause same result: the increased synchronization above normal levels. This scenario leading to an abnormal brain was expounded in great detail in “The Brain-Behaviour Continuum − The Subtle Transition Between Sanity and Insanity” (World Scientific).
The idea that brain oscillations dictate the tremor in patients, originally derives from intracerebral recordings from PD patients performed in the early sixties. Those recordings showed that thalamic cells discharge spikes rhythmically at the frequencies of the resting tremor.
Currently, a host of more detailed observations has been gathered.
Compared with the complex and not-so-rhythmic appearance of some epileptic seizures, Parkinson’s disease is characterized by more robust periodic oscillatory neuronal activity. Considering this, it is conceivable to expect more success in halting the PD pathological brain rhythms by direct electrical stimulation to the brain.
For instance, the phenomenon of phase-resetting (Figure 4) can be used as a DBS protocol, because this method needs a very periodic oscillation.
Figure 4. Phase-resetting.
There is a wide array of theoretical studies, using computer simulations, that demonstrate the effectiveness of feedback signals in the suppression of collective synchronization (phase-resetting).
The advantage of these methods is that no precise knowledge of the system’s constituents is required. Only the macroscopic properties are of relevance in these approaches. (Recall similar comments in the epilepsy blogs at this level of description we don’t need molecular/cellular details.)
In current times, the successful application of these methods to arrest PD symptoms is well accepted in the community of investigators and clinicians. Deep brain stimulation in Parkinsonian patients stops the periodic activity for a while – with a return to the pathological rhythm a few seconds after turning off the stimulator.
The following figures, taken from Titcombe et al., 2001, demonstrate the success of DBS arresting the pathological rhythm.
Figure 5A. Simulator on.
Figure 5B. Simulator off.
On the Figure 5A, we can see that switching on deep brain stimulation arrests the periodic limb movement.
On the Figure 5B, we can see there is a fast transition to the tremor after turning off the stimuli.
These two recordings represent the Parkinsonian rest tremor finger velocity. Figure 5A panels show the halting of the tremor, in four patients, upon turning the stimulation on at the dashed vertical line. The traces on the right represent the transition to the pathological tremor oscillation – after switching off the DBS at the dashed vertical line. We can see that Subject A needed a longer time to halt the tremor.
The location of the deep brain stimulation was, for each subject, in the subthalamic nucleus (STN), ventrointermediate thalamic nucleus (Vim), and internal globus pallidus (GPi). The frequency of the stimulation is in brackets. (List of thalamic nuclei.)
Figure 6. STN = subthalamic nucleus etc.
Although this is an obvious success, it does not really help in understanding the cellular mechanism of action. This is because the deep brain electrodes excite not only cell bodies in the neighborhood but also passing axons from remote areas.
Nevertheless, the result of the perturbations induced by DBS can be thought of as the destabilization of fixed points. In Parkinson’s disease (PD) and related pathologies these are high-amplitude periodic neural rhythms a.k.a. large-scale oscillation (read more).
Neural oscillations, or brainwaves, are rhythmic or repetitive patterns of neural activity in the central nervous system.
GIF LINK: Analysis of Oscillatory Neural Activity in Series Network Models of Parkinson’s Disease During Deep Brain Stimulation.
The standard electrical stimuli currently used in deep brain stimulation protocols for PD are based on high frequency pulses. Note in the figure that all frequencies are over 100Hz, in similar fashion as the previously discussed vagal nerve stimulation in epilepsy.
However, high frequencies may have non-desirable secondary effects. Therefore, other stimulation paradigms (models) can be equally or more effective, and more gentle than the almost continuous brain perturbations imposed by the current devices.
There is theoretical evidence that methods based on nonlinear delay feedback and phase-resetting are efficient at desynchronising populations of cells. As opposed to the “linear” brute-force approach of the on-off high-frequency stimulation, these “nonlinear” stimulation protocols selectively counteract the synchronization mechanisms.
Accordingly, a recent study presented a deep brain stimulation recording apparatus that sends appropriately timed stimuli to selected brain areas in order to cancel out pathological synchronous activity leading to PD tremor.
Figure 7. EEG vs. EMG.
The device relies on recording EEG and EMG activity in the individual and sending adequate pulses to deep brain sites. These devices, provided they work on patients, could represent major advances in the field of deep brain stimulation in PD and related disorders.
In the final analysis, then the neurostimulation desynchronises neural networks. It has been shown, that sub-thalamic nucleus (STN) stimulation (developed in the early 1990s) desynchronizes basal ganglia networks. This results in the motor units firing more independently and thus reducing the higher synchrony characteristic of the tremor.
Interestingly, this also occurs by treatment of the patients with dopamine. Thus, two very different therapeutic methods, pharmacological and biophysical, result in a similar amelioration of the symptoms.
The inspection of the brain regions that change activity as a result of either subthalamic stimulation or levodopa therapy (that restores dopamine) revealed great similarity between the two protocols.
The side effects of STN (subthalamic nucleus) stimulation are not of great concern in current times. It does not seem to lead to significant cognitive dysfunction, although some deficiencies in specific cognitive tasks have been reported in patients with deep brain implants.
The approaches aforementioned relied on the destabilization of synchrony. Conversely, other motor abnormalities in PD may still necessitate a relatively stable periodic dynamics. Gait dynamics is an example of this.
There are considerable gait disturbances in PD patients, such as freezing (episodic gait impairment) and abnormal variability in stride length and interval (continuous gait impairment).
In general, walking is not just a strictly periodic phenomenon, but shows small stride-to-stride fluctuations. Hence, in these cases some periodic forcing could be applied to maintain a less variable and almost-periodic gait activity. This has been done using the socalled rhythmic auditory stimulation (RAS) which improves gait in PD.
Auditory pacing (Figure 8) serves as an external clock to set the pace. Treadmill walking similarly is useful in PD patients as it provides external cues to restore or stabilize the impaired rhythm.
Figure 8. Auditory pacing (credits.)
The commonly used RAS (rhythmic auditory stimulation) is strictly periodic forcing. On the other hand, normal gait is not strictly periodic but displays fluctuations. Therefore, modifying rhythmic auditory stimulation into nonrhythmic by adding some variability to it improves gait dynamics in patients with moderate PD.
Let us close with a reflection on comparing deep brain stimulation protocols in PD (Parkinson’s disease) with those used in epilepsy.
It is perhaps unfair to compare the success at controlling tremor in PD with the less apparent control of seizures in epilepsy.
First, the tremor is present almost continuously in patients, whereas seizures are not, nor are they easily anticipated.
Second, the circuitry in PD is relatively well known, offering clear targets to desynchronise the circuitry, while each epilepsy has its own anatomy.
Third, the motor dis-coordination, especially tremors, is a quite periodic rhythm so that attempts at phase-resetting can be assayed. But seizures display a wide array of waveform morphology and nonperiodic activity.
For these and other reasons, treatment of motor dysfunctions by direct brain perturbations is more probable to succeed than the control of seizures in clinical settings.
Nevertheless, it is hard not to note similarities in the collective network activity in PD and in epileptiform phenomena. Two basic principles include:
(I) States of more extensive excitability, which promotes higher rates of neuronal spike firing.
(II) The tendency towards synchronous activity, consequent to the enhanced excitability.
Since all people’s physiology and biochemistry are different, so is their reaction to drugs (Pharmaceuticals) and electrical therapies (Electroceuticals). Therefore, the concept of personalized medicine and individualized therapies (Dynamiceuticals) has emerged. This new field involves identifying and understanding each system’s (brain’s) own dynamics.
Figure 9. Dynamiceutical approach.
After all, it is the basic physiology that determines the network dynamics and the possible success of DBS protocols. Hence the Dynamiceutical approach, grounded in the detailed knowledge of the physiological dynamics of the systems to be treated, is important