Non-Invasive Neuromodulation of the Central Nervous System: Opportunities and Challenges: Workshop Summary (2015)

4

Therapeutic Uses of Non-Invasive Neuromodulation

Neuromodulatory devices provide tools that can translate insights from cognitive neuroscience into targeted therapies for disorders of the

central nervous system. These tools have been most well developed in the field of psychiatry, but applications for neurorehabilitation and the treatment of neurologic disorders such as migraine, epilepsy, and movement disorders were also discussed by several workshop participants.

According to Alvaro Pascual-Leone, neuromodulation forces clinicians to think from the patient-system point of view, enabling them to personalize interventions by understanding the patient’s particular pathophysiology; in a sense, it is reverse engineering the disorder. Doing that requires the disorder to be deconstructed, moving from concepts such as dementia or depression to specific symptoms, and then characterizing them on the basis of the neural substrate of those symptoms, that is, the specific circuits affected. In other words, he said, these neuromodulatory approaches enable targeting not of the disease itself, but specific symptom complexes that map onto specific neural substrates.

Most of the work on developing therapeutic applications of neurostimulation has focused on transcranial current stimulation (tDCS and tACS) and transcranial magnetic stimulation; however, more recently tRNS has captured the imagination of scientists, clinicians, and even the general public, said Pascual-Leone. At the time of the workshop, two TMS devices targeting treatment-resistant depression are FDA approved and covered by insurance and Medicare in most states: the Neuronetics device with the NeuroStar protocol and the Brainsway device with H-coil, both for the treatment of medication-resistant depression. Since the workshop, a third device has been cleared by the FDA—Magtim Super Rapid.¹ The NeuroStar TMS Therapy alone has treated more than 25,000 patients (NeuroStar TMS Therapy, 2015). With a remission rate, based on both controlled trials and clinical experience, of about 30 percent, Pascual-Leone said this means that perhaps 24 or 25 remitters per day are being helped with TMS.

These results beg the question of why only some people respond to treatment and, indeed, whether neurostimulation is actually responsible for the responses seen. Beatrix Krause and Roi Cohen Kadosh examined individual differences that impact the effectiveness of TES, concluding that patient characteristics that impact the effectiveness of treatment include genetic factors, head or tissue morphology, and state factors (e.g., fatigue, attention, alertness); disease characteristics, including symptoms

__________________

¹See http://www.magstim.com (accessed June 1, 2015).

and co-morbidities; and stimulation characteristics (Krause and Cohen Kadosh, 2014).

Pascual-Leone reviewed the literature on therapeutic applications of neurostimulatory devices, specifically TMS and tDCS, looking at randomized sham-control, parallel-group, multisession studies across multiple indications. For every indication and every device, he found both negative and positive studies. For some indications such as epilepsy and stroke, the specifically targeted area for stimulation was especially important. For example, in studies of generalized epilepsy where the target is unclear, the results were poor; however, when the target was known, and the stimulation was delivered appropriately to the ipsilateral or contralateral cortex, the results were greatly improved.

Most studies reviewed were small and were affected by a high risk of bias for a variety of reasons (e.g., selection bias, performance bias from blinding, etc.). Thus, noted Pascual-Leone, Cochrane reviews² of the quality of evidence across multiple indications concluded that while there may be some indication of efficacy of non-invasive brain stimulation, the evidence is weak. He concluded that larger studies with more diverse groups and settings are needed to avoid bias.

Despite the weak evidence, Pascual-Leone sees great promise in these neurostimulatory approaches if interventions can be individualized. That, he said, requires thinking about the disorders in a different way and demonstrating that the appropriate substrate is being engaged for a specific patient. For example, to address cognitive function, cognitive neuroscience insights might be able to help guide the interventional applications. For this he advocated combining brain stimulation techniques with neuroimaging and neurophysiology, that is, by using magnetic resonance imaging (MRI)-guided approaches for optimal spatial precision and EEG-guided approaches for optimal frequency for a given cortical location or condition. He also stressed the need to combine neurostimulation with behavioral, pharmacologic, and other interventions and the importance of exploring how these devices can be used safely in children, the elderly, and other special populations.

Experimental medicine approaches, where as part of a clinical trial a target is tracked in response to dose of stimulation, could also help understand contributors to efficacy, said Thomas Insel. TMS is well suited

__________________

²See http://www.cochrane.org/CD008208/SYMPT_stimulating-the-brain-without-surgeryin-the-management-of-chronic-pain (accessed July 1, 2015).

for this approach, especially coupled with other electrophysiologic recording techniques, functional imaging, and other outcome measures, said Alexander Rotenberg. However, Roy Hamilton, assistant professor of neurology at the University of Pennsylvania, pointed to the complexity of the problem, which arises from the multidimensionality of the parameter space. For example, with tDCS, factors that must be controlled for include the intensity and location of stimulation, the cognitive construct of interest and how it is measured, the anatomy of the network, and the baseline state of the brain.

PSYCHIATRIC DISORDERS

According to Sarah H. Lisanby, electroconvulsive therapy has “unparalleled efficacy” in terms of benefits to patients with depression, particularly those who have failed to respond to pharmacologic treatments (Greenberg and Kellner, 2005). In one study of depressed patients who reported suicidal thoughts and acts, treatment with ECT resulted in a resolution of suicidal intent in 80.9 percent of those enrolled, some (15.3 percent) after only one ECT session, while most required multiple sessions (Kellner et al., 2005). However, the downside of ECT is memory loss, including loss of memories about the world as well as memories about events in a person’s life (autobiographical memory) (Lisanby et al., 2000).

Innovations, including the use of realistic head modeling to optimize the location of electrodes on the head so that the stimulation is more focal, as well as the use of ultrabrief pulse ECT (Sackeim et al., 2008), have attempted to resolve some of the problems associated with ECT, noted Lisanby. Magnetic seizure therapy (MST), in which magnetic stimulation is used to induce focal seizures, is another experimental technique that provides more focal stimulation in comparison to ECT, and this increase in focality is associated with a reduction in cognitive side effects (Won et al., 2014).

TMS is also used at subconvulsive levels for the treatment of depression. In one study comparing active TMS to sham TMS, TMS was significantly superior to sham after 4 and 6 weeks of five sessions per week (O’Reardon et al., 2007). Similar results were shown in a recent multicenter study with 212 patients diagnosed with major depressive disorder (MDD) with 20 sessions in 4 weeks acutely, and then biweekly for 12 weeks (Levkovitz et al., 2015). Another recent meta-analysis concluded that both high- and low- frequency TMS demonstrated a significant ef-

fect in patients with posttraumatic stress disorder (PTSD) using continuous scale measures of PTSD and depression symptom severity (Karsen et al., 2014). Lisanby noted that the benefits of TMS include a good safety profile, the ability to stimulate focally, and the lack of adverse effects on memory.

Bradley Gaynes, professor of psychiatry at the University of North Carolina School of Medicine, discussed the comparative effectiveness review on non-pharmacologic interventions for treatment-resistant depression in adults, conducted by the Agency for Healthcare Research and Quality (AHRQ) under his direction (Gaynes et al., 2011). AHRQ strives to identify the highest quality evidence available and synthesize that information quantitatively to answer key questions that are of public health or medical importance. Gaynes noted that these efforts are limited by the information available; thus as the evidence evolves, conclusions evolve as well. He further noted that absence of evidence does not equate with absence of effect.

The standard used by AHRQ is to determine whether evidence is of high, moderate, low, or very low quality, reflecting the level of confidence the agency has that evidence supports a claim. In their assessment of non-pharmacologic interventions for treatment-resistant depression, they found that rTMS does not clearly differ from ECT in terms of benefits or harms; however, the strength of evidence was low (Gaynes et al., 2011).

With regard to rTMS compared with sham treatment, they concluded the following:

• rTMS produced a greater decrease in depression severity based on high strength of evidence.
• rTMS was three times as likely to produce a response based on high strength of evidence.
• rTMS was six times as likely to achieve remission based on moderate strength of evidence.
• rTMS produced a greater improvement in health status and daily functioning based on low strength of evidence.
• There was insufficient evidence on the ability of rTMS to maintain response or remission (Gaynes et al., 2011).

In terms of benefits, noted Gaynes, they concluded that rTMS produced better outcomes for depression severity and response rates for young adults and for depression severity in older adults with post-stroke depression, although the strength of evidence was low. In terms of

harms, rTMS produced more scalp pain at the stimulation site than sham treatment, again only with low strength of evidence. The AHRQ report also included other analyses to try to compare effect sizes among trials designed to assess different treatment strategies such as those that used augmentation or switching designs, but there was not enough evidence to come up with any firm conclusions.

Gaynes said the ability to quantitatively synthesize data from TMS studies was hindered by varying definitions of treatment-resistant depression; an unclear number of prior treatment episodes; varying parameters such as coil location, motor threshold, stimulus pulse, and number of pulses; whether TMS was used as an add-on or substitute treatment; and baseline levels of depression. In addition, journal articles from which data were derived often report only group effects, making it difficult to answer simple questions such as how depression severity affects outcomes. He and several other participants also commented that data published in journals may be biased in favor of studies with positive outcomes. He noted other knowledge gaps as well, including information on health-related outcomes such as quality of life, levels of functional impairment, and patient reports as well as efficacy in specific population subgroups.

Moreover, TMS is not without risks. The most common risk is headache, and there is also a risk of hearing loss; however, the most serious risk is seizures (Rossi et al., 2009). Medications can raise or lower the risk of seizures depending on their effects on seizure threshold, said Lisanby, but the safety guidelines that are used to guide TMS dosing were derived in healthy subjects who were not taking medications. Certain populations may be especially vulnerable; for example, children and individuals with comorbidities such as autism, substance abuse, and addiction.

The devices that have been cleared by the FDA for the treatment of depression have the capability of doing both high and low frequency stimulation, but the dosage approved for depression is high frequency. A range of other devices are in development, as noted in Chapter 3, many of which deliver low-frequency stimulation, which carries with it a much lower risk of seizures. A recent meta-analysis that included nearly 250 patients across eight studies found no detectable difference in terms of therapeutic effect between the two frequencies (Chen et al., 2013). tDCS has also been tested as a treatment for depression, and a recent study showed that the combination of sertraline and tDCS was more effective than either in monotherapy for the treatment of depression (Brunoni et

al., 2013). Relative to TMS, tDCS is safe, with no known risk of seizure, said Lisanby. However, similar knowledge gaps remain with regard to mechanism of action and optimization of dose in space and time.

Non-invasive neurostimulation has been used for other psychiatric disorders as well, including PTSD (Karsen et al., 2014), obsessive-compulsive disorder (Berlim et al., 2013), aggression (Dambacher et al., 2015), and addiction (Bellamoli et al., 2014).

NEUROREHABILITATION

The ability of neurostimulation to help patients recovering from stroke also holds great promise, Hamilton said. Because recovery from a focal brain injury depends on the reorganization of networks that serve specific cognitive operations, non-invasive brain stimulation offers both a window into the brain to reveal those networks and a way to leverage change and improve recovery, said Hamilton. Moreover, because stroke is so prevalent—affecting nearly 800,000 Americans each year—and responsible for tremendous morbidity (Go et al., 2014), and given the paucity of effective therapies, there is great unmet need in this area.

According to Hamilton, TMS and tDCS are being applied to three main areas in stroke recovery: hemiparesis, which is motor weakness of one side of the body; neglect, which is the inability to attend to or act upon stimuli on one side of one’s body or space; and aphasia, or problems producing or understanding language. His group starts by understanding how intact cognitive systems work and how injured systems differ from normal systems, using this knowledge to guide the development of brain stimulation protocols. For example, a meta-analysis from his group looked at functional imaging data from aphasic subjects during language tasks compared to normal subjects, and determined that while normal subjects activate left-dominant networks, aphasic subjects also activate additional areas in both the left and right hemispheres (Turkeltaub et al., 2011).

Hamilton and others have pursued the notion that neurostimulation might facilitate reorganization of injured neural networks. This work is predicated on the idea that the two hemispheres are richly interconnected so that, for example, if you inhibit the intact hemisphere or stimulate the damaged area, you restore some degree of balance between the two hemispheres (Hamilton et al., 2011). Indeed, in the Contrastim Stroke Study, 20 subjects received 18 sessions of rTMS stimulation to the non-lesioned

hemisphere prior to task-oriented upper limb rehabilitation. Ten subjects received sham stimulation. More than 88 percent of the subjects receiving rTMS stimulation had a meaningful clinical response compared to only 38 percent who received sham stimulation (Harvey et al., 2014), prompting a much larger Phase III trial that is currently ongoing, said Hamilton.

Studies of tDCS in aphasia have also been promising although more heterogeneous, said Hamilton (Monti et al., 2013). This variability may be due to differences in study design as well as patient factors, including lesion shape and location and the constellation of symptoms. The use of TMS and tDCS for neglect is at an even earlier stage, he said. His lab has shown, for example, that tDCS facilitates visuospatial processing (Medina et al., 2013); and at least one study has shown that a type of TMS called theta-burst stimulation (TBS) may accelerate recovery from spatial neglect (Koch et al., 2012). Michael Fox mentioned another study where TMS was administered to the parietal cortex, resulting in enhanced ability to detect targets on the ipsilateral side (Hilgetag et al., 2001).

OTHER NEUROLOGIC DISORDERS

Neurostimulation has been investigated as a treatment for a variety of other neurologic disorders, including epilepsy, migraine, movement disorders, amyotrophic lateral sclerosis (ALS), tinnitus, and chronic pain. For example, Mark Hallett’s group investigated the use of rTMS delivered to the left and right motor and dorsolateral prefrontal cortex as a treatment for gait abnormalities and bradykinesia in Parkinson’s disease (PD) patients (Lomarev et al., 2006). In comparison to patients receiving placebo, those receiving TMS showed gradual improvement in gait as well as reduced upper limb bradykinesia. The effects lasted for at least 1 month after treatment ended.

The eNeura TMS device is FDA approved for medication-resistant migraine with aura (FDA, 2013). Pascual-Leone said that while this device may have limited clinical impact because most patients with migraine have unpredictable auras and those with aura are often resistant to treatment, the device nevertheless paves the path for patient-applied home use of this technology.

COMBINING NEUROSTIMULATION WITH OTHER THERAPIES

Hallett and several other participants also mentioned the potential of combining multiple neurostimulation technologies or a combination of neurostimulation with behavioral interventions or drugs in order to provide even better results. For example, Yang et al. (2013) combined rTMS with treadmill training in patients with PD. After 12 sessions of rTMS followed by treadmill training over a 4-week period, the combination was shown to improve walking performance and modulate corticomotor inhibition better than either treatment alone in patients with PD (Yang et al., 2013).

Pascual-Leone’s group has also investigated the combination of tDCS with visual illusion for the treatment of neuropathic pain in patients with spinal cord injury. In a sham-controlled study, patients receiving tDCS and visual illusion together reported a reduced intensity of neuropathic pain in comparison with those who received either intervention alone or placebo (Soler et al., 2010).

Many participants commented on the likelihood that effective treatment of many brain disorders may require combinatorial approaches that deliver neurostimulation in combination with pharmaceutical treatments or behavioral-based interventions. Mark Demitrack, vice president and chief medical officer of Neuronetics, for example, said he believes combinatorial work across device platforms as well as combinations of neurostimulation with behavioral interventions are ripe for study now. Ana Maiques, however, said she believes the barriers separating pharmaceutical and device companies are diminishing; and Atul Pande, chief medical officer at Tal Medical, added that most studies are currently conducted against a background of existing pharmacotherapy. Regulatory challenges would likely be complex for a combinatorial strategy; in addition, business models for combination approaches, including potential operational synergies among companies, have yet to be explored, said Jeffrey Nye.

This page intentionally left blank.