Chronic neurological disorders are the leading cause of disability worldwide and their burden of death and disability is increasing [1,2]. Of all chronic neurological disorders, stroke, multiple sclerosis (MS) and traumatic brain injury (TBI) are the three largest contributors to disability and mortality [1,3]. Despite the incidence of chronic neurological conditions declining from 1990 to 2015, time spent living with a disability, and mortality rates from chronic neurological conditions has been increasing [1]. These findings suggest that people with chronic neurological conditions are living longer with persistent disability [1].

Recent literature has explored the burden of disability for persons with neurological conditions [4-6]. People with chronic neurological conditions have a wide range of gait dysfunctions including, but not limited to, deficits in gait initiation, movement fluidity, and gait speed [4]. In addition, balance impairments demonstrating increased postural sway and instability are a significant problem for the neurological population [4]. As a result, older adults living with neurological disorders are a significant higher falls risk compared to older neurotypical adults [5]. For example, Coote et al. reported that 71% of older adults with MS sustained a fall compared to 41% of age-matched controls over a six-month period [6]. Furthermore, falls in persons with MS were more likely to cause serious injury [6]. Similarly, a systematic review by Lai et al. reported approximately 73% of individuals who have sustained a stroke will experience a fall in the first six months following hospital discharge [5]. The development of new effective rehabilitation interventions to ameliorate persistent neurological-related movement disabilities, such as balance and gait, are required to improve the lives of persons with neurological conditions and for improved sustainability of healthcare systems [1].

Cranial-nerve non-invasive neuromodulation (CN-NINM) is a new and emerging technology that aims to facilitate neuroplasticity in individuals living with chronic neurological conditions [7-12]. CN-NINM technology utilizes translingual neurostimulation (TLNS) via a portable neuromodulation stimulation (PoNS^TM) device [13-16]. When combined with targeted physical rehabilitation, the PoNS^TM device has been reported to improve long-standing balance and gait deficits in persons with chronic neurological conditions [13,14,16].

The PoNS^TM device is a portable device with an electrode array placed on the anterior dorsal surface of the tongue [10,13,17], providing a natural form of stimulation to cranial nerves V (trigeminal nerve) and VII (facial nerve) [16,18-20]. Additional cranial nerves (CN) that have been reported to be stimulated include the glossopharyngeal (CN-IX), vagal (CN-X), and hypoglossal (CN-XII) nerves [11,18]. Stimulation of the CNs subsequently stimulate targeted areas in the brainstem and cerebellum through the lingual branch of CN-V and chorda tympani branch of CN-VII [18]. It is hypothesized that stimulation of these structures modulates activity of corresponding nuclei of the brainstem in the sensory and spinal nuclei of the trigeminal nuclei complex and the caudal portion of the solitary nucleus tract [16]. In turn, connected regions are also modulated, stimulating corresponding neural networks and subsequent neurotransmitter release [16,20,21]. When combined with physiotherapy, this potentiates long-term neuroplastic change [10].

Randomized and non-randomized control trials, as well as small sample research designs have revealed promising results for CN-NINM using TLNS in chronic neurological conditions [18]. For example, improvements in chronic balance deficits, falls, and gait have been demonstrated in individuals living with moderate TBI [7,8,14,15], stroke [15,22-25], MS [26] and Parkinson’s disease (PD) [26], as well as in persons with refractory balance deficits [9].

The purpose of this scoping review is to explicate the evidence available for the use of TLNS in neurological populations. A preliminary search of PROPSERO, MEDLINE, the Cochrane Database of Systematic Reviews and the JBI Evidence Synthesis was conducted on February 14, 2021, at which time no current or underway scoping reviews were identified. This will be the first scoping review addressing TLNS literature to inform future research and healthcare policies. The protocol for this scoping review was specified in advance and published in JMIR Research Protocols [27-30].

The following objectives were addressed by this review:

1. What is CN-NINM, and how is CN-NINM being delivered?
2. What are typical parameters used in TLNS?
3. Which conditions or clinical symptoms have evidence supporting the use of TLNS?
4. How is TLNS being incorporated into neurological rehabilitation?
5. What has previous research shown on the effects of TLNS in neurorehabilitation?
6. What are the current gaps in the evidence base and the implications for rehabilitation science?

This scoping review followed the Joanna Briggs Institute (JBI) methodology to map the available literature on CN-NINM [31]. The study was designed and conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Review (PRIMSA-ScR) [32]. The protocol was registered on Open Sciences Framework and published in JMIR research protocols [30,33].

Inclusion/Exclusion criteria

Eligibility criteria were established a priori utilizing the population, concept, context framework [31]. Studies were selected in accordance with the following criteria.

Population: Studies were included that involved adults with neurological conditions (≥ 18 years of age). Conditions of interest included, but were not limited to, stroke, mild TBI including concussion, moderate TBI, severe TBI, and MS. Given that neuromodulation interventions for depression are typically invasive in nature and not via TLNS, depression was excluded from the current study. Trigeminal neuralgia was also excluded, as our search retrieved an overwhelming number of articles pertaining to the condition and TLNS is rarely used in this population.

Concept: We explored the concepts of TLNS, the PoNS^TM device and CN-NINM. It excluded other forms of invasive and non-invasive brain stimulation such as, transcranial magnetic stimulation, functional electrical stimulation, transcranial electrical stimulation techniques and deep brain stimulation. Studies pertaining to Vagus nerve stimulation were excluded given that this type of nerve stimulation primarily involves an implantable device.

Context: We considered studies conducted in any context and geographical location to widely explore the application of CN-NINM in neurological rehabilitation.

Types of sources: Limited restrictions were placed on the types of sources included in this review. This scoping review considered studies with qualitative, quantitative, and mixed methods design. Furthermore, systematic reviews and gray literature including non-research articles, opinion papers, texts, as well as documents from relevant websites were considered. Posters and unpublished theses were considered gray literature. Studies in French and English published since 2000 were included. This time frame was deemed appropriate as TLNS is a new technology that emerged from the sensory substitution research performed in the early 1990’s at the Tactile Communication and Neurorehabilitation Laboratory at the University of Wisconsin-Madison [16,34]. Abstracts were excluded due to resource constraints and the limited information offered given the brevity of abstracts.

Search strategy

An initial limited search of MEDLINE and CINAHL was undertaken to identify articles on the topic. Text words contained in the titles and abstracts of relevant articles found in the initial limited search, along with key concepts and Medical Subject Heading suggested by an academic librarian at McMaster University were used to develop the search strategy. Comprehensive search strategies were developed in collaboration with the librarian. These search strategies were developed for five databases: MEDLINE, AMED, CINAHL, EMBASE and Web of Science. The detailed search strategy is provided in Appendix 1.

Sources of unpublished studies and gray literature included ProQuest, DuckDuckGo and Google. Consistent terms were used for each browser and the date and time as well as the results of the search were documented (Appendix 2). For the gray literature search, the first five pages of results from each respective database were screened. If a document met the inclusion criteria a backward and forward reference search was performed to identify other relevant documents. The research team also contacted and received gray literature from author Dr. Yuri Danilov and Helius Medical Technologies.

Hand-searching of all reference lists of included studies was undertaken to identify additional studies of relevance.

Selection process

All identified records were collated and imported to Zotero reference management software (Center for History and New Media, George Mason University) and duplicates were removed. All research team members screened the first ten citations of the initial MEDLINE search to ensure consistency in application of the a priori inclusion and exclusion criteria and researcher agreement. Any discrepancies were discussed and resolved collectively by the research team. The titles and abstracts of the remaining articles were screened by groups of two researchers and any disagreements were resolved between the two researchers, or in the case that an agreement was not reached, senior researchers were consulted. Full text of potentially relevant papers was retrieved and imported into Covidence screening and data extraction tool (Veritas Health Innovation, Melbourne, Australia). Full text reviews were performed by groups of two researchers. The results of the search and reasons for exclusion were reported in full and presented in a PRISMA-ScR flow diagram (Figure 1).

Data extraction

The data extraction table proposed in the protocol paper was trialed with two resources to ensure all relevant data were extracted [30]. Minor revisions were made such as extracting funding information as well as description of mechanism of action, type of device used and indications for use (Appendix 3).

The data extraction process was completed by groups of two researchers and to ensure accurate data collection, the extracted data was compared and discrepancies were resolved by consensus.

A total of 11,700 records were retrieved from the five databases selected for this study. Additionally, 99 articles were retrieved from our gray literature search. Of these, 40 studies met all inclusion criteria and were included in the scoping review. Details of the study selection process are outlined in Figure 1. The 40 studies consisted of one systematic review [11], five RCTs [21,35-38], five non-RCTs [9,12,17,39,40], two pilot studies [13,41], six case reports [8,19,20,22,25,42] and nine articles on conceptual literature [10,16,23,43-48]. Furthermore, 12 gray literature sources were obtained which included posters, poster presentations and facts sheets [14,15,24,26,49-56]. Of the included articles, seven were published in Canada [9,14,17,22,35,42,46], three in Australia [13,19,20], 29 from United States [8,10,11,13,15,16,21,23-26,36-41,43-45,47,49-56] and one from Ireland [48].

The systematic review and randomized controlled trials are described in Table 1. The systematic review comprised five studies. Three of the studies were included in our scoping review [38-40], the remaining two studies were excluded as they did not meet our inclusion criteria [57,58]. Two of the five studies [38,58] had usable outcomes measures that could be combined into a meta-analysis (MA). The MA reported improvements in balance and gait following CN-NINM in a variety of neurological conditions, however, because one of the studies [58] was not included in our study (due to use of an alternative device (BrainPort balance device)), we cannot consider the results in our scoping review.

Five RCTs reported on the effects of the PoNS^TM device combined with PT, or an individualized exercise and balance training program, on gait and balance in individuals with MS [37], TBI [21,35,36] and chronic balance dysfunction [38]. In each study, the intervention group was compared to a control group who were administered a sham treatment using a low frequency stimulation (below stimulation threshold) PoNS^TM device [21,35-38]. Two of the studies reported improvements in gait [37,38] and one reported improvement in balance [38]. Three studies reported no significant between group differences in balance and gait [21,35,36]. Additionally, one study found that there was no between group differences for high or low frequency groups in grey matter volume across 11 different brain regions [21].

Five of the 40 articles were non-randomized controlled trials [9,12,17,39,40] and two were pilot studies [13,41] (Table 2) all of which used the PoNS^TM device. These studies investigated stroke, MS, TBI, PD, chronic balance dysfunction, and neurotypical adults. Two studies used fMRI to map brain activity during stimulation [12,40]. Two studies [12,40] reported on the effects of CN-NINM on optic flow, with only one study reporting an upregulation [40]. One study used EEG to measure brain activity and found an increase in alpha and theta activity [8]. Balance was reported to improve in three studies [13,39,41]. Additionally, two studies investigated cognition, both of which reported improvements [17,41]. Finally, Danilov (2013) reported improvements in word processing [41].

Six of the 40 articles used a small sample research design of which one was a case report [22], and five were N-of-1 studies [8,19,20,25,42] (Table 3). The populations studied included SCI, TBI, spinocerebellar ataxia, posterior fossa craniectomy and stroke. Four studies found improvements in balance [19,20,22,42] and one study reported a reduction in postural sway [20]. Two studies reported improvements in gait [20,22]. Quality of life measures were investigated in three studies [8,22,25]. One study reported improvements in depression, memory, emotions, and hand dexterity [25], while another reported a reduction in post-traumatic stress disorder and terrors [8]. Furthermore, one assessed QoL via a variety of outcomes measures, however, did not report or comment on overall improvements of QoL, or lack thereof [22].

Nine articles were conceptual literature comprising four literature reviews [10,43,47,48], one narrative review [46], and four book chapters (Table 4) [16,23,44,45]. The populations included in these publications were TBI, MS, SCI, Cerebral Palsy, refractory balance and gait disorders, PD, stroke, cerebellar degeneration, and posterior fossa syndrome. All 9 articles described PoNS^TM as the device used for TLNS, to facilitate neuroplastic change. Six of the 9 articles referred directly to stimulation of the CN-V and CN-VII via TLNS [16,23,43,44,46,48] whilst three articles identified use of TLNS but did specifically indicate which nerves were being accessed [10,45,47]. Seven articles suggested that TLNS would be an appropriate adjunct therapy to augment and or enhance physical rehabilitation [16,23,43-46,48]. Eight articles reported improvements in balance and gait dysfunction [10,16,24,43,45-48] and 8 articles indicated that TLNS can be applied to a variety of neurologic conditions [16,23,43-48]. Furthermore, these articles noted that TLNS was safe with no reports of serious adverse events [10,16,23,43-47]. Three articles indicated that TLNS would be suitable for use at home [16,43,48].

Twelve articles were identified from gray literature, 7 of which were research based [14,15,24,26,49-51]. Six articles were research posters [14,15,26,49-51] with one thesis publication (Table 5) [24]. These studies included PD, MS, TBI, and stroke populations. Quality of life was assessed in three studies, one reporting a reduction in depression symptoms [26], one stating a general improvement on the Multiple Sclerosis Impact Scale (MSIS-29) [14] and one reporting improvements in emotion and speech [51]. Three studies found an increase in gait speed [15,26,49] and one study found improved sequencing of muscle activation [49]. Lastly, two studies reported an improvement in static and dynamic eye control [15,50].

Twelve articles were identified from gray literature, 7 of which were research based [14,15,24,26,49-51]. Six articles were research posters [14,15,26,49-51] with one thesis publication (Table 5) [24]. These studies included PD, MS, TBI, and stroke populations. Quality of life was assessed in three studies, one reporting a reduction in depression symptoms [26], one stating a general improvement on the Multiple Sclerosis Impact Scale (MSIS-29) [14] and one reporting improvements in emotion and speech [51]. Three studies found an increase in gait speed [15,26,49] and one study found improved sequencing of muscle activation [49]. Lastly, two studies reported an improvement in static and dynamic eye control [15,50].

Five of the 12 gray literature articles were conceptual literature articles (Table 6), of which 2 were fact sheets [52,43] and three white papers [54-56]. All five articles reported that the PoNS^TM device was used to create neuroplastic change in the brainstem and cerebellum via stimulation of CN V and CN VII, augmenting the effects of rehabilitation to improve balance and gait [52-56]. Two publications indicated TLNS was able to produce long lasting effects [55,56]. The populations discussed included TBI, MS, Cerebral Palsy, stroke, SCI and refractory balance and gait disorders indicating wide applicability for neurological populations. CN-NINM was reported to be safe with no adverse event in two of the publications [55,56] and one reported its applicability for home use [53].

The findings from this scoping review indicate that TLNS combined with physical therapy may enhance neuroplasticity, thereby facilitating neurological recovery in persons with neurological conditions. The majority of papers investigated TLNS as a conjunct modality to be used alongside physiotherapy intervention [8,11,13,15,17,19-22,24-26,35-37,39,41,42,49-51] with relatively few studies investigating the utility of TLNS as an individual treatment modality [9,39]. TLNS via CN-V and CN-VII stimulates multiple subcortical regions leading to modulatory effects impacting clinical outcomes including gait and balance [7,11,16,21,23,36,37,43-45,48]. When paired with rehabilitation interventions, TLNS has been found to improve functional ability [45]. Although these results are encouraging, there is some evidence demonstrating that there is little to no difference in treatment effects between control/sham groups and TLNS groups [21,35,36].

All studies used TLNS via the PoNS^TM device, developed at the Tactile Communication Neurorehabilitation Laboratory at the University of Wisconsin-Madison, and licensed by Helius Medical Technologies. The parameters described in the reviewed literature vary widely and are individualized dependent upon the clinical presentation. CN-NINM via TLNS, using devices such as the PoNS^TM, are a safe, portable, and feasible modality to include in home-based exercise programs.

Application to clinical populations

CN-NINM via TLNS was investigated in multiple neurological populations including TBI [8,21,24,35,36,41,42,50], MS [14,17,26,37,45], PD [26,45], sub-acute stroke [13-15,51] and spinal cord injury [22]. Other conditions included spinocerebellar ataxia, posterior fossa craniotomy, or were not specified [12,19,20,38-40]. There are limited systematic reviews and randomized controlled studies for all these populations, possibly due to the emergence of TLNS as a relatively new technology. Most studies were non-randomized controlled trials, pilot studies, or small sample research designs. The results of the studies suggest that TLNS can be paired with exercise to improve functional outcomes such as gait and balance across multiple neurological populations.

In Canada, the PoNS^TM device is approved for use in persons with mild to moderate TBI and MS and recently achieved FDA approval in the US for MS. The expected number of people living with the effects of CVA, TBI, MS are increasing and are associated with chronic physical and mental disabilities [59]. CN-NINM via TLNS has the potential to improve balance and gait, which are common problems in neurological populations, as well as the potential to influence non-motor impairments such as depression and visual processing thus impacting upon quality of life [14,22,26].

Understanding mechanisms of recovery

Neuroplasticity is a physiological process through which areas of the brain can change or reorganize their functions through intrinsic or extrinsic stimuli [60]. Clinically, this understanding of neuroplasticity forms a theoretical rationale for utilizing modalities such as TLNS [7-12]. CN-NINM via TLNS uses a natural form of stimulation by directly targeting sub-cortical structures via the cranial nerves [18,43]. The anatomical and physiological properties of the tongue allow for localized stimulation through neural impulses or “spikes” stimulating the brainstem and cerebellum [18,43]. Neural networks within the brainstem and cerebellum are key regions involved in postural control, balance and locomotion [61-63]. Recovery of the cortico-reticulospinal system, arising in the brainstem, has been associated with improved gait and lower extremity function [64-66] and is a strong predictor of postural control in adults [67]. The cerebellum, directly connected with the brainstem, coordinates movement, balance, equilibrium, and muscle tone [62] and is associated with motor learning [68].

MRI studies included in this review identified increased activity in the aforementioned areas following application of TLNS [32-35]. TLNS has also shown to increase the volume of grey matter in areas of sub-cortical processing such as the cerebellum and decreases in areas of conscious processing such as the frontal lobe [17]. This indicates a potential recovery mechanism that leads to more typical recovery of functions such as gait and balance that are primarily subserved by sub-cortical neural networks [17]. This neuromodulation is transmitted through the central nervous systems via functionally connected neural systems [48]. CN-NINM via TLNS is therefore quite different from other forms of stimulation, such as transmagnetic stimulation (TMS) and transcranial direct current stimulation (tCDS) which utilize external stimulation through focalized magnetic or electrical fields to induce cortical activity.

A primary goal of neurorehabilitation, and a determinant of quality of life, is achieving independence in walking at home and the community, requiring the integration of postural control and balance [69-71]. Therefore, targeting these neural structures using CN-NINM via TLNS combined with PT has the potential to ameliorate balance and gait impairments [13,35,72].

 Dose and safety

A wide range of parameters for pulse-width, frequency, and intensity of the PoNS^TM device were used in the literature, however, the optimal parameters remain to be established [7]. For the PoNS^TM device, voltage and timing are pre-programmed, however the intensity can be modified through the pulse width and the literature reported ranges from 0.3 to 0.6 µs being used [37,49]. Variable intensities were used in the studies as the device allows for subjects or therapists to alter it to a “comfortable level” [37,49]. Currently, a standardized program has not been established in the literature, however Helius recommends a daily training schedule of two 60-minute sessions and one 20-minute session for six days/week for 14-weeks [54-56]. The application of TLNS was always paired with gait training, balance training, breathing awareness exercises or movement control exercises, however the number and length of sessions per day as well as duration of the intervention period varied [17,19,20,51]. These findings present challenges in determining the optimal parameters for clinical application and a standardized program across studies is necessary to compare intervention effects. No serious adverse events were reported however some side effects included vertigo, pain, headache, vomiting, nausea, and fainting [7,54-56].

 Implications for neurorehabilitation

The findings of this scoping review suggest that TLNS primes areas of the CNS for neuroplastic change when combined with PT. The evidence suggests that incorporating motor learning principles such as task specificity and intensity of practice can improve functional outcomes in persons with neurological conditions [73,74]. To maximize neuroplastic change, the activity must be challenging with sufficient intensity to result in an increased production of brain derived neurotrophic factor [75-78]. Therefore, combining CN-NINM via TLNS with neurorehabilitation has the potential to further boost neuroplastic change facilitating functional recovery in activities such as balance and gait [19,20,22,23,48,49] which are common movement problems in people with neurological conditions [4-6]. Furthermore, the wide variety of patient populations studied demonstrates that TLNS has wide applicability to improve balance and gait in persons with neurological conditions. However, it is difficult to ascertain which population, including corresponding clinical indicators, would present as the most amenable candidates to receive this intervention based on the limited sample sizes and relative heterogeneity of populations and participants within current studies.

An additional benefit of TLNS is its safety and portability making it a feasible modality to include in home-based exercise programs. Current neurological rehabilitation models in Canada may not provide sufficient intensity of therapy to maximize functional recovery [79]. The inclusion of TLNS in home-based exercise programs can reduce the economic costs and time-burden of in-person rehabilitation whilst simultaneously increasing accessibility, allowing individuals to further engage in their rehabilitation. Therefore, TLNS has the potential to reduce physical impairments, increasing activity and participation, thereby improving quality of life whilst minimizing health care system costs.

This is the first scoping review investigating TLNS allowing for exploration of both peer review and gray literature. To date the majority of TLNS literature comprises small sample or non-randomized controlled trials in a wide variety of neurological conditions. Several conceptual papers provide the theoretical underpinning of TLNS and its potential contribution to neuroplastic change in the CNS to optimize recovery of balance and gait in persons with neurological conditions.

Several gaps in the literature on TLNS are identified: (i) determining for whom TLNS is most appropriate including but not limited to the medical diagnosis and clinical impairments most amenable to TLNS; (ii) determining the optimal parameters for use including dose, frequency, intensity and duration of interventions; (iii) determining the specific brain regions in which neuroplastic changes occur; and, (iv) understanding end-user perspectives on TLNS, an area of rehabilitation where research is lacking. Therefore, consideration of mixed-methods studies would provide useful clinical information.

To date, the majority of studies have been undertaken with persons with mild to moderate TBI or MS. There has been limited investigation with persons with stroke despite stroke representing the largest global burden of disability [1,3]. Further research is needed to examine the role of TLNS in this population.

A limitation of a scoping review design is that it does not allow for the assessment of risk of bias in the identified studies [29]. In this review, several studies were completed by researchers with financial interests in the PoNS^TM device. Therefore, independent investigation of the effectiveness of TLNS is required. Scoping reviews do not allow for the determination of effectiveness as a statistical analysis cannot be completed [29]. However more robust studies with larger sample sizes are needed before a future systematic review is conducted to determine effect size more accurately. Likewise, the generalizability of the results cannot be determined [29].

With the emergence of new therapeutic technologies, such as the PoNS^TM device, for people living with neurological conditions, it is imperative to explore the evidence base for appropriate clinical application and future research directions. The results of this study suggest that TLNS has the potential to produce lasting changes on gait and balance if paired with an individualized physical rehabilitation program. It is a feasible form of treatment that can be incorporated into home exercise programs across various neurological conditions including TBI, MS, CVA, PD and SCI. The existing body of evidence for its application is limited and further research is needed to define parameters for future use.

The authors report no conflict of interest.

International Registered Report Identifier (IRRID): PRR1-10.2196/2996

Financial support: None to declare.

Authorship Statement: Senior researchers JVG and DB conceived the idea of the scoping review and developed the research questions; both contributed meaningfully to drafting and editing the manuscript along with the rest of the research team. TN and SD wrote the introduction. KB developed the study methods. KB, TN, SD, KL, TH, and AB contributed equally to the study selection process and data extraction. KL and SD contributed to the results section. The discussion was written by TH and AB. All authors contributed to and revised the manuscript and approved the final manuscript.