Nearly 10 million people worldwide suffer from Parkinson’s Disease (PD). Common symptoms include limb tremors, poor movement coordination and initiation, and muscle rigidity. The gradual decline in mobility is crippling and devastating. The first recorded case of PD was reported just over two hundred years ago, but our current best treatments still fall short of anything we might call a cure.
Over the years, many potential solutions for treating PD have been posited, tested, and ultimately discarded; these solutions have included drugs, lifestyle changes, and surgeries, just to name a few. In recent years, however, a new type of therapeutic known as deep brain stimulation has become more mainstream for treating PD. Deep brain stimulation (DBS) for PD works by implanting a long, thin electrode into focal regions of the brain and delivering fast electrical pulses to these regions. These rapid pulses induce changes in brain activity that can alleviate Parkinson’s most aggressive symptoms in just a few seconds (see for yourself in the video below).
In many patients, DBS has met with enormous success. But strangely enough, basic scientists and neurologists today have only a very rudimentary understanding on why DBS works as well as it does for treating Parkinsonian symptoms. As it is, the insertion of even small devices into mammalian brain tissue is usually met with risks of mass inflammation and rejection of the device. Despite this tricky technical roadblock, doctors have discovered that successful implantation and subsequent slamming the brain with high-frequency electrical activity through DBS is apparently sufficient to produce therapeutic results. Somehow we’ve stumbled upon DBS as a meaningful, effective fix for Parkinson’s Disease, a disorder we do not fully understand. To begin to understand the mystery behind DBS’s efficacy, let’s unpack what we do know about Parkinson’s Disease and why other prominent therapies to date have failed.
A few decades of neurology and neuroscience research have revealed a defining cause underlying PD: the progressive death of a small, but vital, group of dopaminergic cells in the substantia nigra pars compacta (SNc) deep within the basal ganglia, an evolutionarily ancient set of nuclei that control many elements of motion, mood and behavior. Dopamine from the SNc has, in recent rodent studies, been shown to play pivotal roles in movement initiation—an elegant corroboration between basic science and clinical practice.
So if dopamine loss is the problem, then can we just fix PD by adding more dopamine or stimulating dopamine neurons? Could we just do DBS in the SNc and compensate for the loss of dopamine?
The answer is no, and if you want to know why, you need only to reach into your pantry and grab that box of Oreos(™). Broadband dopamine release in the brain is frequently the result of receiving a reward or positive outcome; we crave it, we need it to learn and to survive. Dopamine instructs us regarding how our actions influence or enhance our own survival, and it fundamentally makes us feel good. Unfortunately, this also means that attempting to address the dopamine deficit of PD by delivering excess dopamine to compensate can be intensely addicting, much like the four Oreos I’ve consumed since starting writing this paragraph.
If you need more convincing that being hooked on dopamine is addicting, consider this: pharmacological interventions and other experimental manipulations causing the upregulation of dopamine make rats and mice execute reinforced, compulsive behaviors in which they perform repetitive actions for minutes on end, just to receive more dopamine delivery in the brain. This goes to show that rodents are highly susceptible to the addictive effects of enhanced dopamine in the brain.
And in humans? Doctors have tried the strategy of targeting dopamine in order to alleviate symptoms and the SNc deficits of Parkinson’s, but they observed the impulsive, addictive tendencies as well. The advent of the drug levodopa (also known as L-dopa) for PD treatment in the mid-20th century was soon followed by the demise of several patients in its clinical trials; although many users found temporary relief for their aberrant motor symptoms while taking the drug early on, over time these same patients became highly susceptible to gambling, obsessive-compulsive disorder (OCD), and other impulsive behaviors. L-dopa prescriptions were soon discontinued, and doctors in the 1980’s continued on the hunt for an alternative treatment for Parkinson’s.
Alongside the rise of L-dopa on the pharmaceutical market came the proliferation of stereotactic neurosurgery. This is a practice that allows for neurosurgeons to plan their interventions in the brain in such a way that allowed them to target afflicted tissue while avoiding integral intact structures , all while causing minimal damage to the skull. Ultimately, stereotactic surgery enabled neurosurgeons, for the first time, to probe the effects of focal stimulation in areas deep within the brain that were otherwise difficult to reach. In early stereotactic surgeries of the 1960’s, it was discovered that high-frequency electrical stimulation of a deep brain region known as the subthalamic nucleus (STN)—also part of the basal ganglia gang—could alleviate symptoms of patients with various motor disorders (while low-frequency stimulation of the same region would exacerbate symptoms). And thus the seeds for the eventual proliferation for deep brain stimulation in STN were sown.
The STN is currently the most common target for DBS treatment for patients suffering from Parkinson’s, but its involvement was essentially discovered by accident. That high frequency of 100-120Hz that alleviated tremor symptoms? Doctors chose that frequency because in the 1970’s and 1980’s that was the fastest that any state-of-the-art portable device could deliver intracranial pulses. Even though 100-120Hz is roughly 30% faster than STN neurons naturally fire in vivo. Seems legit, I guess?
Despite the remarkable success of DBS for Parkinson’s treatment, scientists and doctors alike remain baffled as to why DBS works as well as it does. For example, why does stimulation of the STN in Parkinsonian patients have opposite effects with low-frequency versus high-frequency stimulation? Also, what is the STN even responsible for naturally?
Basic neuroscientists are far from agreeing on the STN’s normal functions in the mammalian brain. Some neuroscientists posit that the STN is responsible for ‘action selection’—the moment-to-moment execution of behavior—since it is a major output structure for the basal ganglia at large. However, natural dysfunctions or lesions of the STN have been associated with aberrant and obsessive habit formation. The confusion surrounding the success of DBS for Parkinson’s is underscored by a fundamentally incomplete understanding of how microcircuits in the STN and broader basal ganglia coordinate movement. So here are some ideas from a budding basal ganglia enthusiast:
DBS in STN may primarily target inhibitory axons that enter the STN from the pallidum (another primary candidate region for controlling action selection). In first targeting inhibitory axons entering the STN, DBS may be forcefully shutting down the excitatory neurons emanating from STN that may normally carry motor command information outward to other brain areas; in prohibiting outward signals from the STN, DBS may be able to dampen outgoing noisy signals that contribute to tremor-like symptoms.
There is another mechanism by which DBS could actually be changing normal STN function. Given that the DBS electrodes are not perfectly inhibitory-axon-targeted, and are likely also over-stimulating the excitatory neurons in STN, it is also possible that DBS drives depolarization block (a phenomenon by which over-excitation causes forward transmission failure in neurons) in STN and thereby changes downstream firing patterns in ways that manage to block Parkinsonian tremors.
Alternatively, DBS could actually work in reverse; since the electrical stimulation frequency is so high and so exogenous, it could be that these electrical pulses can travel backwards up the axons that transmit information to the STN, rather than affecting the information transmitted from the STN (like above). This phenomenon is commonly referred to as “antidromic stimulation”. Antidromic stimulation could change firing in cortical projections to STN, these changed inputs could refine motor actions to eliminate PD symptoms. Given that DBS essentially slams the STN with a very high frequency of stimulation (higher than they naturally produce), this is actually plausible, and may even be occurring in addition to the other two hypotheses posed here.
Testing these possible theories however, at least in humans, remains an enormous challenge. Solving whether or not DBS works by forward inhibition or reverse/antidromic modifications in human patients would require inserting numerous extra electrodes into a live brain, and monitoring of firing rates in upstream or downstream regions of the STN as DBS is applied within the STN. This type of experiment is not only technically challenging due to the inflammation and damage that extra electrodes cause to brain tissue, but is the kind of experiment mired in ethical and legal questions too. Similar experiments for testing mechanisms of DBS for treating PD in mice would also be difficult to conduct, since there are hardly any widely-accepted rodent models for PD in which to study the problem.
Forwards, backwards, upside down---though the neuroscientific correlates of DBS’s widespread success remain unknown, the results are stunning. DBS is a phenomenal reminder to basic scientists of just how much there is still left to learn about motor output by the brain. And we are slowly making progress in deepening our understanding of how the STN and its partner brain structures across the basal ganglia are responsible for producing skilled, controlled motor actions. Even if we got here by a happy accident.