A Critical Review of Noninvasive Brain Stimulation Technologies in Alzheimer’s Dementia and Primary Progressive Aphasia

TRANSCRANIAL DIRECT CURRENT STIMULATION (tDCS)

The earliest form of noninvasive brain stimulation, tDCS, passes a constant electrical current through a large positive electrode (anode) to the head and underlying brain tissues that exits at a negative electrode (cathode) usually of the same size [28]. The mechanisms are not completely understood but the underlying principle is that the electric current can upregulate neuronal activity under an anode through subthreshold depolarization of the membrane potential (not triggering an action potential), and downregulate neuronal activity under a cathode through hyperpolarization of the membrane potential [29, 30]. However, the effects seem to depend on position and orientation of the cells, neurochemical cascades, and the overall net electric field produced by stimulation parameters. Despite the mechanisms remaining unclear, tDCS has been shown to alter the timing and rate of neuronal firing to produce neuroplasticity, which may strengthen synapses downstream [31]. tDCS has been found to be safe with no serious adverse effects in accordance with recommended procedures for system requirements, setup, and delivery [32–35]. The most common side effects are tingling or burning sensations during stimulation, and skin redness, itching, fatigue, headache, and nausea temporarily after stimulation [36, 37] (Table 1). Although many trials report “targeting” specific brain regions, the electric current courses through numerous brain regions beyond the specified area, and in turn, stimulation is diffuse and likely reaches many cortical and subcortical areas. As such, references to “targeting” brain regions with conventional tDCS are misguided, as possible effects on the area of interest cannot be separated from potential effects on “non-target” areas.

OPEN IN VIEWER

Table 1

Pros and Cons for the Noninvasive Brain Stimulation Methods

Many randomized controlled trials (RCTs) have examined the effects of tDCS on global cognitive functioning in ACS whereas only one has done so for PPA. When tDCS was delivered at 2 mA for twenty sessions over the temporal lobes bilaterally using an accelerated, alternating protocol—that is two consecutive stimulation sessions each day for 20 min with reversal of the side of stimulation (left temporal then right temporal with 5-min break between stimulations)—significant improvements on brief measures of global cognition were found in patients in advanced stages of ACS relative to sham. Importantly, the change of nearly 5 points on the Montreal Cognitive Assessment was clinically meaningful [38]. Several other RCTs [39–43], including two trials involving at-home delivery of tDCS, also found statistically significant enhancement in ACS on similar measures of global cognition after 2 mA of tDCS, though not all [44]. The most common site in delivering stimulation was over the left dorsolateral prefrontal cortex (DLPFC) at durations ranging 20–30 min and between 10 [41] to 180 [39] sessions. However, none of those trials had effects that reached the threshold for clinically meaningful change. Unlike ACS, delivering tDCS with such parameters (ten 30-min sessions at 2 mA) did not enhance measures of global cognition in a mixed group of patients with PPA comprised of all three variants [45].

Regarding episodic memory, tDCS trials have varied greatly in the stimulation parameters as well as the outcome measures used. Not surprisingly, the results have been mixed. A single session of tDCS applied over the bilateral temporoparietal cortices at 1.5 mA for a brief duration (15 min) was found to show improvement on a word recognition task relative to sham in ACS [46]. It is important to note that the enhanced recognition was not clinically significant, but such large improvements would be improbable with only a single session. Whereas no significant changes were seen in total learning, delayed recall, or recognition performance on a verbal episodic memory task among patients with ACS after more sessions of tDCS at a higher intensity (six 30-min sessions of 2 mA) applied over the left temporal cortex in comparison to sham [47], another study with considerably more sessions (eighty-four 30-min sessions at 2 mA) applied over the bilateral frontal lobes did show improvement in total learning on a similar task among patients with mild-to-moderate ACS [48]. However, the change in total learning did not meet the threshold for clinically meaningful improvement. In another trial with ten 25-min sessions at 1.5 mA, a composite index of episodic memory functioning was enhanced after tDCS was delivered over the right parietal cortex as well as the right prefrontal cortex in patients with mild-to-moderate ACS, though there was no sham for comparison and the change also did not appear to reach clinical significance [49]. There have been no tDCS trials that have explored effects on episodic memory in PPA, likely because the difficulties with recall are explained by problems with language processes instead.

tDCS effects on other cognitive abilities, such as language and executive functioning, have been less investigated in ACS and have been more of a focus in PPA. For ACS specifically, there have been mixed results regarding enhancement of language skills, though differences in electrode positioning might underlie this (left DLPFC versus right frontal/parietal cortex) [39, 49], and in the two studies that evaluated executive functions, no significant effects from tDCS were seen [47, 49]. Most tDCS trials in PPA examine effects on language skills when combined with speech therapy, rather than have it as a standalone treatment, and deliver 2 mA over the left inferior frontal cortex or left parieto-temporal cortex for 10–15 sessions at durations ranging 20–30 min. Under these conditions, naming ability was enhanced significantly more when tDCS was added to speech therapy in comparison to sham stimulation alongside speech therapy [50–58]. Although many failed to meet the threshold of clinically meaningful change, two trials did and both applied tDCS to a mixed PPA group involving all three variants over the left inferior frontal cortex at 2 mA for a minimum of 10 sessions for 20-min in duration [53, 54]. It is worth mentioning that disease stage may shape the effects of tDCS, as those with greater levels of impairment have shown larger responses [55], but even so, there may be a point where the underlying neural circuits become too damaged to be modulated by tDCS and other methods.

A newer tDCS technique, high-definition tDCS (HD-tDCS), was developed nearly 15 years ago that involves the use of much smaller electrodes. It allows more flexibility in electrode configurations, as more than two can easily be placed. Often, five electrodes (one anode, four cathodes) are placed in HD-tDCS, typically in a ring pattern with the anode at the center. HD-tDCS has been associated with greater precision in delivery of the current [59, 60], depending on electrode placement, and a head-to-head comparison with conventional tDCS has indicated greater intensity and longer lasting effects for HD-tDCS [61]. Despite it being more than a decade since the method was invented, to date, only one completed trial has examined HD-tDCS on cognition in ACS and one trial for PPA (included in studies showing increased naming above). HD-tDCS was applied in a 4×1 ring configuration in both, but electrode placement was individualized to deliver the current over the left DLPFC using computational modeling from magnetic resonance imaging scans from each patient in the ACS trial. When six 20-min sessions of HD-tDCS were administered at 2 mA in an accelerated protocol (three sessions per day with 15 min breaks in-between), increases on global cognition and episodic memory (delayed recall and recognition combined index) were found in patients with ACS relative to sham [62], albeit data were not reported to determine if the changes reached clinical significance.

A critical question about noninvasive brain stimulation, or any potential intervention, is whether the duration of a treatment effect is long-lasting. Favorable results were reported in a trial that delivered ten 25-min sessions of tDCS at 2 mA over the right prefrontal cortex, with enhancement on two global measures of cognition seen out to eight weeks in patients with ACS [41], with clinically meaningful change evident for one. Moreover, multiple trials in patients having mostly the nonfluent/agrammatic and logopenic PPA variants also found that enhancement of cognitive abilities persist out to eight weeks when tDCS was applied at 2 mA for 20–25 min for a minimum of 10 sessions [50, 57], whereas stimulation for longer periods did not [45]. Persistence of tDCS effects on cognition beyond an eight-week timeframe remains to be understood, as treatment related effects have been reported out to twelve weeks in a trial with nonfluent/agrammatic PPA [50], but effects returned to baseline levels in another trial with similar stimulation parameters in patients with ACS [40].

TRANSCRANIAL ALTERNATING CURRENT STIMULATION (tACS)

tACS applies electric current to the brain through anode and cathode electrodes similar to tDCS. However, tACS delivers a sinusoidal current at specific frequencies, alternating polarity of the electrodes during stimulation (Fig. 1). Although the mechanisms continue to be under investigation, tACS has been shown to synchronize rhythms of neuronal activity, producing cortical excitability that might strengthen synaptic connections [63]. The most commonly reported side effects with tACS include the same physical symptoms as tDCS, but visual phenomena (phosphenes) have also been experienced [64]. In ACS, aberrant brain oscillations or rhythmic patterns of neural activity have been found in association with cognitive symptoms [65–67] and range from slower (delta: 1–4 Hz, theta: 4–7 Hz) to higher frequencies (alpha: 8–12 Hz, beta: 13–30 Hz, gamma: >30 Hz). While these may represent targets for modulation [65, 66], most studies on ACS have focused solely on gamma oscillations (see recent reviews [68, 69]), with some evidence that gamma patterns can increase in ACS following tACS [65, 71].

OPEN IN VIEWER

Fig. 1

Electrical Stimulation Waveforms. A constant level of electrical current is passed through the anode electrode that exits at a cathode electrode for the entire duration of simulation for tDCS. A sinusoidal current is applied in tACS by alternating polarity of the electrodes at a specific frequency during the stimulation period. A random current from a normal distribution is typically applied in tRNS at random frequencies, though the oscillations as well as overall current intensity can be shaped to be towards high/low frequency bands or a net positive/negative current intensity.

Three RCTs to date have examined gamma tACS at 40 Hz on cognition in ACS and there have been no published trials of tACS in PPA [72–75]. When tACS was applied for thirty 20-min sessions (over 6 weeks) over the bitemporal lobes (peak to peak intensity of 2 mA), a measure of global cognition improved and had a clinically meaningful change after stimulation compared to sham [73]. Episodic memory was investigated in another trial that applied more sessions of tACS for a longer duration at a lower intensity (forty 30-min sessions with peak-to-peak intensity of 1.5 mA) over the left DLPFC in an accelerated approach (administered twice daily over 4 weeks). However, treatment was combined with a cognitive training program and patients with mild-to-moderate ACS were mixed with non-dementia cases [72]. Relative to cognitive training alone, the addition of tACS significantly enhanced performance on a comprehensive battery of episodic memory; still, the effect fell short of the threshold for clinically meaningful change. Because other cognitive abilities are recruited to varying degrees when performing the mix of tasks used in the test battery, lumping all performances into a total score may have concealed clinically meaningful changes on more pure measures of episodic memory. Indeed, after only a single session of tACS for an extended duration and at higher intensity (60 min with peak-to-peak intensity of 3 mA) over the precuneus, total learning on a verbal list-learning task was improved compared to sham and near the threshold of clinical meaningful change [75]. Interestingly, this was accompanied by increased short latency afferent inhibition (an indirect measure that reflects improved cholinergic transmission) as well as increased posterior gamma activity. As compared to the precuneus stimulation, a follow-up trial failed to show a change in episodic memory performance following gamma tACS over the right DLPFC.

As opposed to tDCS, the duration of treatment effects on cognition for tACS remains largely unexplored. Only one study has investigated possible persistence beyond the acute period, and it failed to show a treatment effect lasting out to 12 weeks [73]. Because the effects of tACS at gamma bands beyond 40 Hz has not been thoroughly studied, it is unclear if this specific frequency would be optimal in ACS or PPA. Continued work is also needed to examine whether tACS at other frequency bands outside of gamma may affect cognitive problems, and along those lines, theta tACS is thought to have promise for improving episodic memory in ACS [65]. Another important line of future research will be to identify if tACS can have effects in other cognitive abilities, such as language and executive functioning, to potentially lessen the spectrum of impairments seen in ACS and PPA. Moreover, whether tACS can modulate neural circuits far below the brain region it is applied over remains unknown, and a new technique has been developed that may be promising to employ with tACS in ACS and PPA. It is called temporal interference and the general aim involves applying two electric fields at frequencies too high to modulate neural activity on their own, albeit when combined, provides a window for altering neuronal rhythms to deeper structures while having less impact on superficial cortical areas [76] (Fig. 2).

OPEN IN VIEWER

Fig. 2

Temporal Interference. Delivering multiple isolated electrical currents at different frequencies has been proposed as an approach for tACS to modulate regions or circuitry at significant depth in the brain, such as the hippocampus. The technique involves applying the two currents based on a computational model for electrode positioning and at different frequencies that are too high to initiate neural firing on their own. The model suggests an envelope is created where there is a difference in the overlapping frequencies, which is sufficiently low to then modulate neuronal rhythms. From Violante et al. (2023) [76], with permission.

Transcranial Random Noise Stimulation (tRNS)

tRNS delivers alternating current to the scalp at random frequencies. The frequency range can span a wide spectrum (between 0.1–640 Hz) or be restricted to low frequency (<100 Hz) or high frequency (>100 Hz) bands [77, 78]. The random currents are typically drawn from a normal distribution with the average current applied typically being zero (i.e., zero-mean Gaussian white noise) [79], although the distribution can be biased to have a net positive or negative current (i.e., using a DC offset) [80, 81]. The side effects commonly experienced are the same as in tDCS. Evidence has shown that tRNS can modulate cortical excitability similar to tDCS (motor cortex [82]; visual cortex [83]), although mechanisms for tRNS remain to be elucidated and are hypothesized to be related to stochastic resonance and/or temporal summation [78, 84] (Box 1). Some suggest that tRNS effects on higher-level cognitive cortical regions, compared to the primary sensory and motor cortices, may be less consistent and effective [78, 85]. Whether higher-level cognitive functions in ACS or PPA may be impacted by applying tRNS over the brain remains to be understood. To date, no studies have been published investigating tRNS in these conditions. Moving forward, clinical trials exploring if tRNS has meaningful cognitive effects in ACS and PPA, and if so, a duration for any treatment effects are needed. Combining tRNS with a cognitive training program in ACS or PPA in RCTs may be another possible direction to study its potential, given that tRNS has been found to enhance and speed up training effects in healthy individuals as well as those with neurologic disorders [80, 83]. In ACS, theoretically, tRNS could possibly be employed to retune or normalize hyperconnectivity versus hypoconnectivity in the default mode network akin to tACS, among other networks, which are progressively susceptible to the accumulation of pathophysiology [86–88]. Thus, another reasonable course is to research whether tRNS could modulate abnormal cortical excitability in ACS or PPA for improvement of neural networks [89, 90], given some suggestions of tRNS being as effective as tDCS or tACS in changing cortical excitability in other research areas [78, 91].

TRANSCRANIAL MAGNETIC STIMULATION (TMS)

TMS uses an electromagnetic coil to deliver single or a train of alternating magnetic current pulses to the brain [92]. Single pulse TMS has been used primarily as a means to probe neurocircuitry and mechanisms of action whereas repetitive TMS (rTMS) has been used primarily as a treatment of neuropsychiatric diseases [93] through mechanisms similar to those underlying tDCS/tACS. There are multiple TMS parameters that can be configured such as the coil type (e.g., figure-of-eight coil, H-coil, circular coil), frequency, number of pulses, interstimulus interval, and stimulus intensity to probe specific neurocircuits or treat neuropsychiatric disease. For example, the FDA cleared rTMS with figure-of-eight coils and the H-1 and H-7 coils as interventions of treatment-resistant depression [94], the H-4 coil for treatment of smoking cessation [95], and the H-7 coil for the treatment of obsessive compulsive disorder [96].

Consistent evidence has found TMS to be safe in adults across the lifespan with the most common adverse effects reported to be scalp discomfort or pain at site of stimulation and headache [97–100]. The most serious adverse effect found with TMS is seizure, though the incidence of this is significantly small [101]. For instance, a recent meta-analysis found that the risk of seizure with active and sham/placebo TMS was 0.1% and 0.2% respectively [102]. Nonetheless, safety recommendations need to be implemented (see [93, 103]) and safety and tolerability need to always be monitored when providing TMS.

Considerable evidence across different study methodologies (e.g., case series, RCTs, meta-analyses) has suggested that TMS is safe and may be beneficial for older adults with ACS and PPA across the spectrum of severity (e.g., mild, moderate, severe) [104]. Such safety and benefits may rely on multiple factors including patient sociodemographic and clinical characteristics (e.g., gray matter atrophy, dementia severity, comorbidities) as well as TMS parameters and paradigms (e.g., neuroanatomical target, pulse frequency) [105]. The studies that have been conducted have substantially varied in regard to methodologic approach including the patient population (e.g., dementia severity level), cognitive outcome metrics, study design, and rTMS paradigms [104, 106]. While nearly all studies used a variant of the figure-of-eight coil as well as a sham-designed coil as a control condition, they tended to vary considerably in regard to the rTMS parameters of pulse frequency and intensity, number of pulses, and neuroanatomical targeting method (e.g., 5 centimeter rule, 10–20 EEG system, neuronavigation) [104, 106]. The majority of studies utilized high frequency rTMS stimulation that ranged from 10 Hz to 40 Hz while few utilized low frequency (1 Hz) rTMS stimulation. The number of pulses per session significantly varied with a low of 800 [107] to a high of 2400 [108], as did the length of treatment (ranging from 2 to 32 sessions) [109, 110].

To our knowledge, 11 RCTs have examined the cognitive effects of rTMS in older adults with varying severity of ACS and the primary neuroanatomical target for the majority was the DLPFC [40, 107–116]. Nearly all of the studies assessed changes in global cognitive function and found statistically significant increases after rTMS (e.g., 1–3 points on MMSE, 2–3 points on ADAS-Cog), but the effects were small and did not reach the threshold for clinically meaningful change, with only one meeting clinical significance after rTMS was delivered bilaterally to the angular gyrus for 12 days (stimulating 2 s, 30 trains, 2400 pulses, intensity of 90% of resting motor threshold) [40]. One published trial examined the effects of 15 sessions of rTMS on a measure of global cognition in PPA mixed with the nonfluent/agrammatic and semantic variants, in which no enhancement was found, but it is worth noting that placement of the coil was individualized to patients based on screening of language performance [117]. As such, the location of stimulation varied across patients. The overutilization of global cognitive function measures across trials provides useful information on the global cognitive impact of TMS, though it consequently provides limited information with regard to effects on specific cognitive domains compromised by ACS and PPA.

Only a few studies selected cognitive metrics to assess specific cognitive functions such as episodic memory and executive functioning and all were in ACS, with no trials in PPA. When TMS was applied to patients with ACS through an accelerated approach—often called intermittent theta-burst stimulation (iTBS)—consisting of three rounds of stimulation over the left DLPFC for 14 days (3 pulses, 50 Hz bursts every 200 ms at 5 Hz, intensity of 70% of resting motor threshold, rounds separated by 15 min breaks), significantly improved total learning and delayed recall were seen on two different verbal episodic memory measures relative to sham [111]. Only delayed recall on the list learning task showed a clinically meaningful change. In two other trials applying fewer total sessions (10–20 sessions) of rTMS at a higher intensity (90–100% of resting motor threshold) and over different locations (bilateral cerebellum and left parietal lobe), only statistically significant changes in episodic memory were seen [107, 113]. Regarding executive functioning, significant effects of rTMS were seen in some trials with patients having ACS [109, 113] and not others [111], but none reached clinical significance. In order for rTMS to be of significant therapeutic value to older adults with ACS and PPA, its effects specifically on episodic memory or other cognitive skills will need to be examined further and optimized. Otherwise rTMS may never reach full therapeutic potential to improve the cognitive domains specific to and most impacted by ACS and PPA.

rTMS effects on language have been the primary cognitive ability investigated in the limited RCTs published in PPA and interestingly more of a focus in ACS than episodic memory or executive functioning. In ACS, despite naming ability being significantly enhanced with rTMS relative to sham in most of the trials [111, 116], none reached the threshold for clinically meaningful change and there was one that could not be determined due to unreported data [116]. In the single study that did not find effects on naming ability in ACS with rTMS, twenty sessions (2000 total pulses with 20 Hz frequency, intensity at 100% of resting motor threshold) over the DLPFC showed clinically meaningful improvement on an auditory sentence comprehension measure compared to sham instead [115]. Auditory comprehension deficits have been identified at all levels of ACS severity [118]. Albeit compromises in executive functioning have been indicated as a potential driver, no such measures were given to examine if changes in executive functioning may have played a role. Only two RCTs have examined rTMS effects on language skills in PPA, in which one allowed concomitant speech therapy be provided alongside rTMS as part of routine clinical care. Both trials varied placement of the figure-eight coil for rTMS (left DLPFC if handedness was right-side dominant and vice versa; various frontal and temporal areas) and had different parameters for stimulation. When 15 sessions of rTMS were applied to patients having nonfluent/agrammatic and semantic PPA for a total of 1500 pulses each (20 Hz trains with an interval of 20 s, intensity at 100% resting motor threshold), speech word count was enhanced showing a clinically significant change compared to the sham [117]. Furthermore, a composite index of language functioning was significantly better than the sham condition in the second trial that applied more rTMS sessions with fewer total pulses (20 sessions, 1000 pulses each) at a different frequency and intensity (10 Hz trains, 20 pulses per train with 2-s intervals in-between, at 120% resting motor threshold) to a mixed PPA group of all three variants, but both the rTMS and sham groups showed clinically significant changes [119]. It is worth considering that an active rTMS coil was held near the scalp for the sham condition, which still could have allowed magnetic field penetrance into the scalp and underlying brain areas, potentially leading to the observed improvements.

Duration for treatment effects on cognition has been assessed for TMS more than any other noninvasive brain stimulation method. Enhancement of global cognition, episodic memory, and language have been reported to persist out to 8–12 weeks in multiple trials in patients with ACS [111–113], with maintenance of clinically meaningful changes evident for two studies [108, 115]. Disease stage might influence the duration of treatment effects though and may relate to the extent of damaged neural circuits, given some indication that individuals at lower stages of ACS severity (e.g., mild to moderate) had clinically meaningful enhancement out to 12 weeks, while those in the advanced stage did not [114]. Unlike ACS, a duration of treatment effects on cognition for TMS in PPA has largely been unexplored. Only one study has investigated persistence beyond the acute period in PPA, with enhancement of a composite language index out to twelve weeks for a mixed PPA group with all three variants [119]. Despite promising results, optimal stimulation parameters for producing long-lasting effects on cognition remains an important line of future research given considerable variation in rTMS protocols across ACS trials and the limited data for PPA.

NONINVASIVE VAGUS NERVE SIMULATION (nVNS)

nVNS delivers a constant electrical current in cycles (referred to as pulses) to the skin, most commonly the left ear, as the outer ear is connected to the auricular branch of the vagus nerve and the left side is less likely to interact with cardiac function (e.g., heartrate) [120]. This method is commonly referred to as transcutaneous auricular vagus nerve stimulation (taVNS) and is often dosed to be at or below sensory threshold, indicative of whether or not individuals experience physiological symptoms. As such, the most frequent reported side effects with nVNS at threshold include dizziness, tingling or pain at the stimulation site, headache, and fatigue [121–123]. A mechanism for the current modulating the vagus nerve when applied noninvasively is not established. There is some evidence that nVNS can alter sympathetic nerve activity [124], reduce depressive symptoms [125], decrease seizure frequency [126], and improve upper limb movement recovery after stroke [123]. Delivering stimulation to the cymba concha and posterior tragus are the best anatomical targets, instead of other parts of the ear, because functional neuroimaging studies have suggested that stimulation triggers increased activation within important areas of the brainstem [127, 128]. The vagus nerve has projections to the locus coeruleus and dorsal raphe nucleus in the brainstem as well as the nucleus basalis in the basal forebrain [129], structures involved in episodic and semantic memory functions that can be affected in both ACS and PPA. To date, no RCTs have been published investigating nVNS in ACS or PPA, but there have been some open-label studies examining effects of an invasive vagus nerve stimulator on cognition in ACS.

When a stimulator was surgically implanted around the vagus nerve (called cervical vagus nerve stimulation [cNVS]) in an open-label trial, nearly all patients with mild-to-moderate ACS received 0.75 mA of stimulation for 30 s every 5 min based on side-effect tolerability. Two measures of global cognition were administered multiple times over 6 months after implantation, and neither showed meaningful improvement at any time point [130]. A follow-up trial examined long-term treatment for an extended period out to 12 months, but again no meaningful improvement was seen on measures of global cognition [131]. While the lack of a robust effect in ACS from long-term treatment casts some doubt about nVNS having potential to lessen cognitive deficits in ACS or PPA, cVNS has been reported to produce enhancement in non-neurodegenerative conditions, such as verbal memory functioning in patients with epilepsy [132, 133]. There are a wide range of protocols that can be applied with nVNS, involving current intensity, frequency, pulse width, and duration, and thus, different outcomes could occur with an alternative set of parameters.

In light of the limited research, whether cognitive functions in ACS or PPA may be enhanced by nVNS remains to be understood. Clinical trials exploring if nVNS has meaningful cognitive effects in episodic memory, language, or executive functioning are needed to inform if nVNS might hold promise as a treatment in ACS and PPA. Despite some positive effects being seen in healthy individuals as well as those with neurological conditions after nVNS [121, 134–136], the stimulation parameters have varied across studies (e.g., ranging from 5 Hz to 30 Hz) and may not necessarily apply for ACS and PPA. Degeneration of neurons in the locus coeruleus occurs in the initial stages of ACS and progresses over the disease course, with reports of neuronal loss being approximately 30% in the mild cognitive impairment stage and nearly 55% in mild-to-moderate ACS [137]. Perhaps even more so than other technologies, the effects of nVNS may largely depend on the disease stage, at least for ACS. Knowledge gained from stimulation parameters in other conditions might not translate given that protocols may require some modification to the amount of cellular loss in the locus coeruleus in order to potentially be successful. Nonetheless, prior work in healthy individuals has shown that naming and episodic memory performance can be enhanced more after nVNS when paired with a training program [121, 136], and as such, combining nVNS with cognitive training might hold promise as a direction to explore its potential in ACS or PPA in RCTs.

CONCLUSIONS

Only tDCS, tACS, and rTMS have been investigated as treatments for cognitive functioning in ACS and PPA through RCTs (Fig. 3). Caution is advised in interpreting available results as there was significant heterogeneity in the methodologic approach across most trials, including study design, participant characteristics (e.g., dementia severity), stimulation parameters, and cognitive outcome metrics. Delivering different stimulation protocols efficiently informs which treatment approaches show promise for further investigation and which ones do not, but shifting towards the study of standardized protocols will become necessary to fully realize robust techniques with reproducible effects. Knowledge about the therapeutic potential of noninvasive brain stimulation technologies in ACS and PPA is in the very early stages at present.

OPEN IN VIEWER

Fig. 3

Positioning of Noninvasive Brain Stimulation Methods in RCTs in ACS and PPA.

While most trials with each technology have reported positive effects, 85%, 66%, and 62% of the trials for tDCS, tACS, and rTMS did not identify clinically significant changes. In order to advance the field, determining clinically meaningful improvement is vital for identifying stimulation parameters and participant characteristics required for more robust treatment effects. Along these lines, no evidence-supported parameters have emerged as of yet for tACS and rTMS with either ACS or PPA, in part due to the limited trials with tACS and considerable variation in rTMS protocols across trials (Table 2). However, there is early evidence for certain stimulation parameters for tDCS being effective. tDCS over the left inferior frontal region for 20 min at 2 mA for a minimum of 10 sessions has produced consistent and meaningful language improvements in PPA comprised of all three variants, while no clear tDCS parameters for clinically relevant changes have arisen yet for ACS. Given the overlap in language deficits seen in ACS and PPA as well as the neural networks implicated, it is possible that the tDCS paradigm showing promise in PPA may have efficacy for language deficits in ACS. An RCT will be needed to perform a direct comparison to understand if the protocol is applicable in ACS. Much is still to be learned, but evidence has accumulated suggesting treatment effects on cognition for tDCS and rTMS can be long-lasting (when disregarding thresholds for clinical relevance), with many studies pointing to effects persisting out to 8 weeks for tDCS and 8–12 weeks for rTMS.

OPEN IN VIEWER

Table 2

Summary of Current Evidence for Noninvasive Brain Stimulation Methods

NA = not applicable. +/– = statistical but not clinically meaningful changes. – = no statistically significant effects. ++ = statistical and clinically meaningful changes for at least one trial.

Identifying clinically meaningful changes by an improvement of 1 SD is a highly conservative threshold. Although it is a conventional cutoff used within the neuropsychological field, a 1 SD change should not be considered the gold standard for widespread use. Multiple methods could be used to establish a study-specific threshold for clinical significance based on the specific sample, test properties, and test-retest interval involved in the study (see recommendations listed below). Choosing the method for determining clinical significance should be well-justified and made with careful consideration of the study aims, measures, and endpoints. Nonetheless, it is worth highlighting that some cognitive test(s) may have poor properties for detecting clinically relevant changes in a trial. For instance, if a selected measure narrowly captures a range of functioning in ACS and PPA, it can exhibit ceiling effects (scoring at the upper range in mild stages), making it unsuitable as an outcome metric. Combining multiple cognitive tests into a composite or index score can remedy this limitation, but such an approach could lead to a different obstacle for identifying clinically relevant changes in a trial. Because an assortment of cognitive abilities are typically captured in composite scores, any substantial change in specific cognitive functions from noninvasive brain stimulation may be concealed within the overall score. For this reason, more global measures of cognitive function may not be ideal for examining clinically relevant treatment effects. Optimizing noninvasive brain stimulation methods to improve specific cognitive skills affected by ACS and PPA will be vital in achieving therapeutic potential for future clinical applications. Moving forward, future studies should explore clinically meaningful change as an endpoint, treatment effects on specific cognitive domains, and clinical features of participants that may alter treatment effects.

RCTs should consider that some baseline characteristics might affect how robust responses are to noninvasive brain stimulation than others, perhaps due to disease stage or unknown medication interactions altering effectiveness of stimulation parameters. When individuals with clinically meaningful changes are aggregated in the entire treatment sample (including those with no response), statistically significant effects may be identified but the chances of meeting clinical relevance are lessened. As such, important clinical features and stimulation parameters that relate to considerable treatment responses (larger or smaller) in individuals can be overlooked. Parsing out individuals experiencing more robust responses than others may shed light on baseline factors that could affect responsiveness to brain stimulation therapies and provide criteria for selecting candidates who may respond better in larger clinical trials. Future clinical trials on noninvasive brain stimulation in ACS and PPA can be optimized for detecting clinically significant changes by following four recommendations.

Selecting outcome measures that map onto specific brain regions/circuits of interest affected by ACS or PPA.

A prominent approach for noninvasive brain stimulation as a treatment in ACS and PPA has been to examine its efficacy for lessening cognitive deficits. Both conditions progress irreversibly, and thus, enhancing cognitive functioning during the course could translate to better outcomes (e.g., quality of life, level of care partner support) and fill an unmet need, as currently available treatments have not proven to reduce cognitive deficits already manifested. An accelerated protocol (2+ consecutive sessions each day) is a promising stimulation approach that may yield greater success and needs further investigation, as the few tDCS and rTMS trials employing it produced meaningful cognitive improvements. Delivering multiple sessions in a day can have important implications for possible future clinical application since condensing the number of sessions to fewer days would reduce the total in-person visits for treatment, minimizing time and travel related constraints on individuals. Nevertheless, pharmacologic agents aim to slow or delay clinical progression, prolonging the time it takes for significant cognitive decline to occur. A similar approach for clinical efficacy could be taken with noninvasive brain stimulation. Indeed, there have been some trials that provide early support for this notion. With tDCS, patients in advanced stages of ACS showed markedly less decline on global cognitive measures after eighty treatment sessions (20-min in duration at 2 mA) over an 8-month period relative to those receiving sham stimulation [140]. Similar findings have been shown with 32 sessions of high frequency rTMS (1600 pulses a session) spread over a 6-month period for patients with ACS [110]. Devices for several of the technologies are already portable (tDCS, tACS, tRNS, nVNS), allowing for home application, which permits stimulation to be delivered to participants more easily over extended periods. With the recent creation of standardized protocols for home-based noninvasive brain stimulation [141–143], determining if the technologies may delay the cognitive effects of progression in ACS and PPA will likely garner increased attention as an alternative treatment approach in the future.

AUTHOR CONTRIBUTIONS

Christian LoBue (Conceptualization; Visualization; Writing – original draft; Writing – review & editing); Shawn McClintock (Writing – original draft; Writing – review & editing); Hsueh-Sheng Chiang (Writing – original draft; Writing – review & editing); Jessica Helphrey (Writing – original draft); Vishal Thakkar (Writing – original draft; Writing – review & editing); John Hart (Writing – original draft).