Introduction
A large portion of the world’s population experiences chronic pain resulting from a nearly infinite number of causes. Pharmacological treatments such as analgesic drugs (opioids, or pain killers) are commonplace to alleviate chronic pain, however analgesics are drastic both in terms of their mechanisms and their side effects. This is evident from the current opioid epidemic in the United States, which has rendered millions of Americans addicted to opioids, the majority of which became addicted from abusing medical prescriptions. As a result, it is highly advantageous to treat chronic pain with techniques different from pharmacological treatments. The area of research founded in this motivation is that of treating chronic pain with neurostimulation, or electromagnetic brain stimulation. The methodology of electromagnetic brain stimulation is to directly or indirectly apply electric currents to specific regions of the brain. Which regions of the brain are targeted depends on the type of brain stimulation that is used. To detail these methods, this paper will discuss the theory, experimental results, and risks of vagus nerve stimulation, deep brain stimulation, transcranial direct current stimulation, cranial electrotherapy stimulation, and repetitive transcranial magnetic stimulation.
Before discussing the brain stimulation techniques individually, a brief electromagnetic background pertaining to the treatments’ functionalities will be presented. There are two types of treatment techniques within the field of brain stimulation, both of which are based on the concept of stimulating, triggering, or activating a region of the brain by applying electrical currents to it. The difference between the two treatment techniques lies in how the electrical currents are applied. One technique uses electrodes placed on the scalp or implanted in the brain and body to directly apply a current to the region. The other technique uses an external magnetic field to induce eddy currents in the desired region of the brain. Vagus nerve stimulation, deep brain stimulation, transcranial direct current stimulation, and cranial electrotherapy stimulation all fall into the category of direct current application, while repetitive transcranial magnetic stimulation falls into the category of magnetically induced currents.
VNS (Vagus Nerve Stimulation)
Since the mid 1990s, vagus nerve stimulation (VNS) has been used to treat refractory epilepsy, but recently it is has been pursued for treating chronic pain in adults by suppressing nociceptors (pain receptors). VNS operates by attaching two bipolar electrodes to the left vagus nerve in the neck, and generating 30 [Hz] electric pulses between 0.2 - 2 [mA] for 30 seconds, with a five minute break in between [1]. The pulse generator connected to the electrodes is surgically implanted in the left side of the chest, with a battery life of six to nine years. The vagus nerve is used because the nucleus tractussolitarii, nucleus raphemagnus, locus coeruleus, and subcoeruleus are thought to be relay stations for producing analgesic effects in regions of the brain like the periaqueductal gray (PAG), which is the main region for modulating pain.
In a study conducted by Kirchner et al., patients that were going to have vagus nerve stimulators implanted in their body, to treat epilepsy, were subjected to painful stimuli before and after the vagus nerve stimulator was implanted, and the respective Visual Analog Scale scores were compared. According to previous investigations, it is known that VNS-produced analgesia in animals is facilitated by capsaicin-sensitive C fibers in the vagus nerve and other areas in the central nervous system [2]. The key components in the central nervous system that are affected are the nucleus tractussolitarii, the lateral reticular nuclei, and the nucleus raphemagnus. To stimulate these areas, the vagus nerve stimulator is surgically implanted in the upper left side of the chest, which is connected to two bipolar electrodes placed on the vagus nerve in the neck. A pulsed current is then applied between the two electrodes, the amplitude of which plays an important role in the ensuing effects. Animal studies have shown that VNS can both promote and inhibit nociceptive behavior, depending on the amplitude of the current applied between the electrodes. It has been found that stimulation intensities around 30 [μA] impede nociceptor functionality, while lower stimulation intensities encourage nociceptor functionality. Thus, Kirchner et al. sought out to stimulate the vagus nerve with a current amplitude well over 30 [μA].
Kirchner et al.’s study utilized a group of eight men and two women with medically intractable epilepsy, set to receive vagus nerve stimulators, and a control group of eight women and four men without epilepsy nor vagus nerve stimulators. The control and experimental groups were age-matched to avoid any age related health issues. Over the course of the study, three measurement sessions were recorded: a baseline before the stimulator was implanted, two to five days after the implantation, and eight to fourteen weeks after the implantation. For the patients receiving VNS, the stimulation intensity was slowly increased over time, starting from a 30 [Hz], 0.7 [mA] pulsed current, with a pulse duration of 500 [μs], and ending with a 30 [Hz], 1.4 [mA] pulsed current, with a pulse duration of 500 [μs] [2].
Each pain tolerance investigation consisted of a mechanical impact, tonic pressure, and thermal stimulus. The study found that VNS had little to no effect on the pain thresholds of the mechanical impact: 14.3 ± 0.3 [m/s] (velocity of the plastic cylinder) at baseline, 14.5 ± 0.4 [m/s] in the second session, and 14.2 ± 0.5 [m/s] in the third session for the study group, compared to 12.6 ± 0.4 [m/s] at baseline, 13.0 ± 0.3 [m/s] in the second session, and 13.4 ± 0.4 [m/s] in the third session for the control group [2]. Similarly, VNS had no effect on the pain thresholds of the thermal stimulus: 41.8 ± 0.9 [°C] at baseline, 44.0 ± 1.0 [°C] in the second session, and 45.7 ± 0.8 [°C] in the third session for the experimental group, versus 45.5 ± 0.8 [°C] at baseline, 46.7 ± 0.7 [°C] in the second session, and 45.5 ± 0.8 [°C] in the third session for the control group. Unlike the pain tolerance to mechanical impact and thermal stimulus, it resulted that VNS significantly reduced pain from tonic pressure. Over the span of two minutes of a pinching stimulus, the experimental group experienced a baseline pain of 44.5 ± 10.5% VAS (Visual Analog Scale of pain), second session pain of 21.5 ± 7.8% VAS, and third session pain of 25.6 ± 9.1% VAS, compared to the control group’s experience of a baseline pain of 43.2 ± 7.3% VAS, second session pain of 38.9 ± 7.7% VAS, and third session pain of 51.8 ± 6.5% VAS [2]. The results of the pinching stimulus experiments are depicted below in Fig. 1.
Figure 1. Mean curves show pain during pinching in patients (A) and control subjects (B). Baseline is indicated by squares, second session by triangles (VNS, 0.7 [mA]), and third session by diamonds (VNS, 1.4 [mA]). [2]
From the aforementioned results, two observations can be seen. First, increases of the VNS amplitude do not produce increases in pain inhibition, but rather decreases. Second, VNS significantly reduces pain caused by pinching sensations, but does not have an effect on pain caused by mechanical impact or thermal stimulus. Although the results are not striking, the fact that VNS did drastically improve the pain tolerance to pinching, while not affecting the pain tolerance to thermal and impact stimuli, provides cause to consider VNS for future studies on chronic pain reduction. Unfortunately, the authors do not discuss possible hypotheses as to why VNS did not increase pain tolerance to all three stimuli, so it is hard to know for what sources of pain VNS will be effective.
Although it is not a risky operation and treatment, VNS is still not entirely safe or benign. Schlaepfer et al. sought out to examine the risks of VNS and found that after a year long period of VNS treatment, the most prevalent side effects were voice alteration (63%), cough (26%), pain (20%), and dyspnoea (difficulty breathing) (10%) [3]. These side effects were predominantly experienced during or around the time of stimulation, and they decreased in prevalence as time went on. The most serious risk of VNS concerns the surgery to implant the device. Given the invasiveness of implanting a device within the human body, there is the risk of human error during the operation, as well as infection during and/or after the operation. A study has not been done to quantify the infection or surgical error rates, so a figure of merit cannot be given in this paper.
DBS (Deep Brain Stimulation)
A brain stimulation technique even more invasive than VNS is deep brain stimulation (DBS). Deep brain stimulation is most known for treating the motor impairments resulting from Parkinson’s disease, but it has also been examined for reducing chronic pain. As the name indicates, DBS involves surgically implanting electrodes into some part of the brain, which are then connected to an implanted signal generator. The signal generator then produces 5 - 50 [Hz], 0.1 - 3 [V] waveforms with 100 - 450 [µs] pulse widths that are applied between the two electrodes. The areas of the brain that DBS targets are the periventricular/periaqueductal grey matter (PVG/PAG), internal capsule (IC), and sensory thalamus (ST), selected for their roles in modulating and controlling pain [4].
Although there are no proven mechanisms as to why these brain regions help to alleviate pain when stimulated, there are several supported theories. One theory is that stimulation of the PVG/PAG releases endogenous opioids, which are known to have analgesic effects. This theory has come from several studies that found that PVG/PAG stimulation induced analgesia can be counteracted with the administration of naloxone, a drug used to prevent opiate/opioid overdoses [4]. Another theory is that regions of the brain connected to the PVG are also activated, like the medial dorsal nucleus of the thalamus, which in turn may influence how the patient reacts emotionally to their pain, effectively weakening the sensation.
To discuss the efficacy of DBS, a meta-analysis conducted by Bittar et al. will be referenced. This meta-analysis compiled the results from six DBS studies from 1977 - 1997 that stimulated the PVG/PAG, IC, and ST in patients with both nociceptive and deafferentation pain caused by various afflictions. Nociceptive pain is caused by damaged body tissue, and deafferentation pain is caused by afferent nerve connection damage. Each study utilized patients that met the following three criteria: the pain was of known organic origin, all reasonable conventional methods had failed or were poorly tolerated, and the patients did not have neuroses/psychoses or severe depression [4]. The study participants then underwent the operation to surgically implant the deep brain stimulator and electrodes in their brain, and received the DBS treatment after that for an extended period of time.
The six studies consistently followed up on the patients for anywhere between one month and 15 years, recording their pain levels via the VAS or a custom defined pain scale. Due to the lack of consistency with regards to the pain evaluation methods, the results of the meta-analysis are tabulated in a “success” or “failure” manner. Success was defined as complete or partial relief for some studies, or a reduction in the VAS scores for the other studies [4]. It was ultimately found that DBS is more effective at reducing nociceptive pain than it is at reducing deafferentation pain (Table 1), and it is most effective when stimulating the PVG/PAG or when stimulating the PVG/PAG and the ST/IC at the same time (Table 2). The 63 - 87.3% efficacy of DBS for treating nociceptive pain is remarkable. These results clearly demonstrate that when applied to the PAG/PVG, or the PAG/PVG and the ST/IC, DBS significantly reduces the amount of pain felt by the patient.
Table 1. Overall long term rates of success and failure with respect to the two categories of pain; nociceptive and deafferentation[4]
Table 2. Long-term success as a percentage of total number of cases with respect to the site of electrode implantation. [4]
Despite the high efficacy of DBS’s analgesic effects, a limiting factor of it is the invasive surgery required to implant the device in the brain and body. DBS requires holes to be drilled into the skull, through which the electrodes are precisely inserted into the correct region of the brain. To facilitate this operation, a neurosurgeon and neuroimaging tools are required to ensure correct placement of the electrodes. In tandem with the difficulty of the surgery, there are many risks that come with such an invasive procedure and device. From the surgery itself, there is a 1-5% chance of seizure or internal bleeding, and a 2-25% chance of infection [5]. Aside from surgical risks, there are also risks associated with stimulating the brain directly, such as muscle contraction and dysarthria (difficulty talking). Although the risks sound rather intense, DBS has proved to be one of the most effective analgesic brain stimulation techniques.
tDCS (transcranial Direct Current Stimulation)
Similar in theory to DBS, tDCS stimulates certain regions of the brain via two electrodes. Unlike DBS however, tDCS refrains from an invasive medical procedure to apply the electrodes to the brain. In tDCS, one electrode is simply placed on the scalp over the motor cortex, and the other is placed over the supraorbital area, contralateral to the first electrode [6]. Both electrodes are connected to a current source that applies a 2 [mA] direct current between them for around 20 minutes at a time. The placement of the electrodes on the scalp allow for stimulation of the primary motor (M1) and primary sensory (S1) cortices, inducing more activity in them to increase the sensory and pain thresholds of the patient. tDCS has aimed for this mechanism because of findings from studies like one conducted by Schabrun et al., that found that S1 excitability was reduced during and after experiencing pain, and M1 excitability was also reduced after experiencing pain [7]. Thus, if tDCS can increase the excitability of the S1 and M1, it could increase the pain threshold of the patient.
Based on the aforementioned hypothesis, Vaseghi et al. conducted a study concerning the ability of tDCS to modulate excitability in the S1 and M1 and thus modulate sensory and pain thresholds. To do this, the authors gathered four healthy men and eight healthy women, and over the course of fifteen days, administered tDCS to four regions of the brain and administered one sham-treatment, each separated by three days and in random order. For each participant and each treatment type, the participants were subjected to peripheral electrical stimulation and pressure stimulation before, during, and after which the pain and sensory thresholds were measured.
The results of the study can be seen in Table 3, which shows that tDCS produces only a slight increase in pain threshold after 30 minutes. Additionally, the results are varied when comparing the electrical pain and pressure pain threshold effects of tDCS when applied to the M1 versus the S1. For example, tDCS applied to the S1 actually lowers the pressure pain threshold after 30 minutes, which is the opposite of the expected outcome. Additionally, the application of tDCS to the M1 appears more effective than tDCS applied to the S1, however the differences are so small, that it is hard to say for certain.
Table 3. The effects of M1 and S1 tDCS on the level of STh, PTh, PpTh (sensory threshold, electrical pain threshold, and pressure pain threshold) [7]
The results in Table 3 above are unconvincing, due to the fact that at most, there was a 0.18 point increase in electrical pain threshold. The authors suggest that the tDCS treatment may have decreased activity in brain regions like the thalamus and brainstem nuclei, and thus the patients required more stimulus in order to feel the pain. This argument does not sound trustworthy, so unfortunately it must be said that tDCS is one of the least promising brain stimulation treatments.
Despite the lack of positive results, a benefit of tDCS is that its risks are the least severe of all of the treatment techniques described in this paper. The most notable risk of tDCS is skin irritation from the electrode/scalp contact, which can be minimized by using wet sponges. Unlike other stimulation methods, tDCS is impervious to excitotoxic damage from overdriving neurons, thanks to causing neuronal excitation via cation channels rather than the neurons themselves [8]. For the same reason, there is not a risk of accidentally inducing seizures either.
CES (Cranial Electrotherapy Stimulation)
CES is quite similar to tDCS because it also uses two electrodes temporarily attached to the scalp, with a current applied between the two. There are several differences between CES and tDCS that radically change the behavior of the two however: CES applies a current amplitude of 100 - 500 [µA] as opposed to approximately 2 [mA] for tDCS, and CES applies a pulsed current as opposed to a direct current. Additionally, the electrodes used in CES are placed on the scalp in different locations from tDCS, those being on the ear lobes, on the temples, on the maxilla-occipital junction, or on the mastoid processes. A lower current amplitude is utilized by CES because according to a modeling study, the lower current amplitudes can penetrate farther into the brain, incurring insignificant attenuation between the cortical and subcortical regions, allowing for stimulation of deeper areas of the brain [6].
As with all of the brain stimulation techniques discussed here, CES has been investigated for treating various forms of pain, because the preexisting treatments have not been highly effective. One such source of pain is that which results from a spinal cord injury (SCI), often taking the form of chronic musculoskeletal or neuropathic pain. As a result, Tan et al. conducted a pilot study on veterans with SCI to find out if CES is a viable treatment. The study compared the pain relieving effects of one-hour active CES versus sham CES for 21 days, applied via electrodes clamped to the ears [13]. 38 men participated, all of whom had experienced SCI for anywhere between 6 months and 60 years. Half of the participants received sham CES, and the other half received active CES, the two of which were indistinguishable due to the active units producing a current amplitude that resided at a subthreshold level of 100 [µA].