- Email: [email protected]
Systematic Review and Meta-Analysis of Noninvasive Cranial Nerve Neuromodulation for Nervous System Disorders
Accepted Manuscript Systematic Review and Meta-Analysis of Noninvasive Cranial Nerve Neuromodulation for Nervous System Disorders Linda Papa, MDCM, MSc Alexander LaMee, Ciara N. Tan, Crystal Hill-Pryor, PhD PII:
S0003-9993(14)00335-9
DOI:
10.1016/j.apmr.2014.04.018
Reference:
YAPMR 55823
To appear in:
ARCHIVES OF PHYSICAL MEDICINE AND REHABILITATION
Received Date: 6 January 2014 Revised Date:
21 March 2014
Accepted Date: 14 April 2014
Please cite this article as: Papa L, LaMee A, Tan CN, Hill-Pryor C, Systematic Review and MetaAnalysis of Noninvasive Cranial Nerve Neuromodulation for Nervous System Disorders, ARCHIVES OF PHYSICAL MEDICINE AND REHABILITATION (2014), doi: 10.1016/j.apmr.2014.04.018. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
Systematic Review and Meta-Analysis of Noninvasive Cranial Nerve Neuromodulation for Nervous System Disorders
RI PT
Linda Papa, MDCM, MSc1; Alexander LaMee 2; Ciara N. Tan1; Crystal Hill-Pryor, PhD3 1
Orlando Regional Medical Center, Orlando, Florida 2 University of Central Florida, Orlando, Florida 3 U.S. Army Medical Research and Materiel Command, Fort Detrick, Maryland
M AN U
Word Count: 3,189 (excluding abstract and references)
SC
March 19, 2014
Tables & Figures: 6 Number of Authors: 4 Number of Institutions: 3
Linda Papa, MDCM, MSc
TE D
Director of Academic Clinical Research and Attending Emergency Physician Department of Emergency Medicine Orlando Regional Medical Center 86 W. Underwood (S-200)
Tel.: 407-237-6329 Fax: 407-649-3083
AC C
[email protected]
EP
Orlando, Florida, 32806
Alexander LaMee
University of Central Florida Department of Biomedical Sciences [email protected]
Ciara N. Tan, BS Department of Emergency Medicine
1
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
Orlando Regional Medical Center [email protected]
Crystal Hill-Pryor, PhD
RI PT
Neurotrauma Research Deputy
Combat Casualty Care Research Program U.S. Army Medical Research and Materiel Command
SC
[email protected]
CORRESPONDING AUTHOR:
M AN U
Linda Papa, MDCM, MSc
Director of Academic Clinical Research and Attending Emergency Physician Department of Emergency Medicine Orlando Regional Medical Center 86 W. Underwood (S-200) Orlando, Florida, 32806 Tel.: 407-237-6329
TE D
Fax: 407-649-3083 [email protected]
EP
Key Words: Cranial nerve modulator, rehabilitation, central nervous system disorders,
PoNS™, neuroplasticity, balance disorders, neurological disorders, neurodegenerative
fMRI
AC C
disorders, traumatic brain injury, non-invasive neuromodulation, cranial nerve stimulator,
Grant Support: None
Author Disclosure Statement: None
2
ACCEPTED MANUSCRIPT Systematic Review CN-NINM 1
Systematic Review and Meta-Analysis of Noninvasive Cranial Nerve Neuromodulation for
2
Nervous System Disorders
3
ABSTRACT
5
Objective: To systematically review the medical literature and comprehensively summarize
6
clinical research done on rehabilitation with a novel portable and non-invasive electrical
7
stimulation device, called the cranial nerve noninvasive neuromodulator (CN-NINM) in patients
8
suffering from nervous system disorders. The CN-NINM induces processes of neuroplasticity by
9
noninvasive stimulation of four cranial nerves and targets the subcortical area of the brain. Data
10
Sources: PubMed®, MEDLINE® and the Cochrane Database from 1966 to March 2013. Study
11
Selection: Studies were included if they recruited adult patients with peripheral and central
12
nervous system disorders and were treated with CN-NINM and were assessed with objective
13
measures of function. Data Extraction: After title and abstract screening of potential articles,
14
full texts were independently reviewed to identify articles that met inclusion criteria. Data
15
Synthesis: The search identified twelve publications, five were critically reviewed and two
16
combined in a meta-analysis. There were no randomized controlled studies identified and the
17
meta-analysis was based on pre-post studies. Most of the patients were individuals with a chronic
18
balance dysfunction. The pooled results demonstrated significant improvements in 1) Dynamic
19
Gait Index (DGI) post-intervention with a mean difference 3.45 (1.75, 5.15)(p<0.001); 2) the
20
ABC scale with a mean difference of 16.65 (7.65, 25.47) (p<0.001); and 3) the DHI with
21
improvements of -26.07 (-35.78, -16.35)(p<0.001). Included studies suffered from small sample
22
sizes, lack of randomization, absence of blinding, use of referral populations, and variability in
23
treatment schedules and follow-up rates. Conclusions: Given these limitations, the results of the
AC C
EP
TE D
M AN U
SC
RI PT
4
1
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
meta-analysis must be interpreted cautiously. Further investigation using rigorous randomized
25
clinical controlled trials are needed to evaluate this promising rehabilitation tool for nervous
26
system disorders.
RI PT
24
27
Key Words:
29
Cranial nerve modulator; rehabilitation; nervous system disorders; neuroplasticity; non-invasive
30
neuromodulation
M AN U
SC
28
31
Abbreviations:
33
ABC - Activities-specific Balance Confidence Scale
34
CDP - Computerized Dynamic Posturography
35
CN-NINM – Cranial Nerve Non-Invasive NeuroModulation
36
CNS – Central Nervous System
37
DGI – Dynamic Gait Index
38
DHI - Dizziness Handicap Inventory
39
fMRI – Functional Magnetic Resonance Imaging
40
PNS – Peripheral Nervous System
41
SOT - Sensory Organization Test
42
TBI – Traumatic Brain Injury
AC C
EP
TE D
32
43
2
ACCEPTED MANUSCRIPT Systematic Review CN-NINM 44
INTRODUCTION
45
The goal of rehabilitation programs for patients suffering from nervous system disorders is to
47
achieve the highest level of functioning possible. Part of this rehabilitation involves adaptive
48
changes and dynamic remodeling of the structure and function of cells and tissues in response to
49
both endogenous and exogenous environmental stimuli.1,2 This adaptive ability of the nervous
50
system is referred to as “neuroplasticity.”1,3 Neuroplasticity allows the injured brain to remodel
51
or repair itself by recruiting brain regions that are not impaired, thus increasing neural
52
connections to these areas. This increased activation makes it possible for uninjured brain to
53
compensate for the functions of the injured parts.4 Neuroplasticity can also occur through
54
neuromodulation via neurotransmitters such as serotonin, dopamine, acetylcholine, and
55
histamine. These neurotransmitters or “neuromodulators” are secreted by a small group of
56
neurons and diffuse through large areas of the nervous system to regulate nervous system
57
activity.5-7,8 Neuromodulation occurs naturally in the brain but may be altered by extraneous
58
factors such as stress, medications or electrical stimulation.9
TE D
M AN U
SC
RI PT
46
EP
59
The use of external stimulation to change and regulate the internal electrochemical environment
61
of the brain is a rapidly growing area of research.10-13 Most electrical stimulation methods
62
currently being applied are invasive and include (1) deep brain stimulation (DBS) via implanted
63
microelectrodes in the basal ganglia, thalamus, and in the midbrain for treatment of Parkinson’s
64
disease and chronic pain14-16; (2) vagus nerve stimulation (VNS) via implanted electrodes on the
65
vagus nerve for suppression of epileptic seizures and depression;17-22 (3) subdural implantable
66
stimulators for stroke recovery;23,24 and (4) transcranial magnetic stimulation (TMS) via an
AC C
60
3
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
external magnet applied over the skull to stimulate the cerebral cortex for treatment of
68
depression.12,25 Moreover, peripheral nerve stimulation has also been shown to potentially
69
reduce pain due to chronic migraine headaches, but a common adverse effect is persistent pain at
70
the implant site.26 Similarly, hypoglossal neurostimulation has been used to improve symptoms
71
in patients with severe obstructive sleep apnea, but the side effects of surgical implantation limit
72
its use.27 Vagus nerve stimulation has seen positive results in the treatment of epileptic
73
conditions such as refractory epilepsy.28 Though generally regarded as a safe alternative to
74
intracranial surgery, one study reported that 51% of patients experienced side effects requiring
75
the VNS device be switched off or explanted.29 Non-invasive stimulation techniques are more
76
appealing yet not fully validated. Although advances have been made in treating chronic pain
77
through repetitive transcranial magnetic stimulation and transcranial direct current stimulation,
78
further study is still required.30
M AN U
SC
RI PT
67
TE D
79 80
A novel non-invasive and portable electrical stimulation device, called the cranial nerve
82
noninvasive neuromodulator (CN-NINM or Portable Neuromodulation Stimulator - PoNS™),
83
induces processes of neuroplasticity by noninvasive stimulation of four cranial nerves:
84
trigeminal, CN-V, facial, CN-VII, glossopharyngeal, CN-IX, and hypoglossal, CN-XII, that
85
innervate the human tongue. CN-NINM uses comfortable, superficial, electro-cutaneous
86
stimulation of afferent cranial nerve branches to the tongue via small surface electrodes. The
87
tongue is a natural candidate for electrical stimulation due, in part, to a high density of sensory
88
receptors and a high concentration of electrolytes in saliva. The CN-NINM targets the
89
subcortical part of the brain, including the brainstem and cerebellum, an area largely unreachable
AC C
EP
81
4
ACCEPTED MANUSCRIPT Systematic Review CN-NINM 90
by other existing neurostimulation systems.31 Using “white noise” patterns of superficial
91
electrical stimulation on the dorsal surface of the tongue, the impulses stimulate all receptors to a
92
depth of 200-400 microns in the tongue epithelia.32
RI PT
93
Through a systematic review of the medical literature, this article will comprehensively
95
summarize the clinical research done using the cranial nerve noninvasive neuromodulator (CN-
96
NINM or PoNS™) in the treatment of central and peripheral nervous system disorders. The
97
primary objective is to assess the effect of cranial nerve modulation on objective measures of
98
function and evaluate its potential as a rehabilitation tool for these disorders.
M AN U
SC
94
99 100
METHODS
101
A literature search of PubMed®, MEDLINE® and the Cochrane Database from 1966 to March
103
2013 was conducted using the MESH search terms cranial nerve modulator, non-invasive
104
neuromodulation, cranial nerve stimulator, tongue stimulator and tongue modulator and Portable
105
Neuromodulation Stimulator (PoNS™). Other terms also searched included central nervous
106
system disorders, neurological disorders, neurodegenerative disorders, brain injury and
107
expansions of these terms to match synonyms, subterms or derivatives (Table 1). These terms
108
were searched in all fields of publication (e.g. title, abstract, and key word). The search was
109
limited to the English language articles and to “human” studies. Studies were included if they
110
recruited adult patients with peripheral and central nervous system disorders and were treated
111
with non-invasive cranial nerve modulation and were assessed with objective measures of
112
function. Articles that did not use noninvasive cranial nerve modulation or did not treat nervous
AC C
EP
TE D
102
5
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
system disorders primarily, were excluded. In addition, the bibliographies and reference lists of
114
all articles and all review articles were evaluated for other potentially relevant articles.
115
Acceptable study designs included experimental studies, observational studies, and case control
116
studies with clearly defined outcomes measured. Review articles, opinion papers and editorials
117
were excluded. The abstracts of the publications were screened for relevance and in case of
118
uncertainty regarding the inclusion, the entire text of the article was read. Studies were defined
119
as prospective or retrospective according to whether the method of data collection and the end
120
points were defined before patient enrollment began. The full texts of the articles were then
121
pooled and reviewed by two different authors to identify articles that met inclusion criteria. Once
122
the relevant articles were selected they were reviewed using a standard review form. The
123
elements extracted from each study that were included in the standard review form were: 1)
124
study type; 2) study objective; 3) study setting; 4) inclusion/exclusion criteria; 5) outcome
125
measures; 6) sample size; 7) results; 8) conclusion; and 9) limitations. The review forms allowed
126
the reviewers to objectively assess the content of each article in a consistent fashion. A
127
composite evidentiary table was then constructed.
SC
M AN U
TE D
EP
128
RI PT
113
Studies were also assessed for quality. Two authors applied the Newcastle-Ottawa Quality
130
Assessment Scale for Nonrandomized Studies in Meta-Analysis to evaluate the overall quality of
131
evidence for the studies identified.33,34 This Scale was developed to assess the quality of
132
nonrandomized studies with its design, content and ease of use directed to the task of
133
incorporating the quality assessments in the interpretation of meta-analytic results. It uses a 'star
134
system' in three broad categories: 1) the selection of the study groups; 2) the comparability of the
135
groups; and 3) the ascertainment of either the exposure or outcome. A study can be awarded a
AC C
129
6
ACCEPTED MANUSCRIPT Systematic Review CN-NINM 136
maximum of 4 stars in the Selection category; 2 stars in the Comparability category; and a
137
maximum of 3 stars in the Exposure/Outcome category.
RI PT
138
All data were entered into RevMan software (version 5.2 for Windows; Copenhagen, Denmark).
140
The results of studies were pooled using fixed effects models after consideration of heterogeneity
141
among trials. For continuous outcomes we calculated the mean differences reported from
142
individual studies, and we pooled statistics as weighted mean differences with associated 95%
143
confidence intervals. The mean, SD, and sample size for each outcome (DGI, ABC, DHI) pre
144
and post intervention were used to calculate the weighted mean differences. We used χ2 and I2
145
statistics to test for heterogeneity (25%, 50%, and 75% representing low, moderate, and high
146
heterogeneity) as well as funnel plots. Funnel plots are a visual tool for assessing publication bias
147
(the association of publication probability with the statistical significance of study results) in
148
meta-analysis. They are essentially scatterplots of the treatment effects of each study (horizontal
149
axis) versus the study size (vertical axis). Publication bias is a concern if the funnel plots are
150
asymmetrical.
153
RESULTS
AC C
152
EP
151
TE D
M AN U
SC
139
154
The initial search identified fifty potential publications, twelve of which involved the use of
155
cranial modulation systems and had their full texts reviewed. Of these, five studies met all
156
selection criteria and were included in the systematic review. However, only two studies had
157
usable outcome measures that could be combined into a meta-analysis. Details of the study
158
selection process are outlined in Figure 1. Each of these five studies was critically reviewed by 7
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
two investigators using a standard review form. All of the included studies were prospective and
160
were published either in or after the year 2006. All of the studies evaluated outcome pre and post
161
use of the CN-NINM device, however, there were no randomized controlled trials. The mean
162
sample size of the selected studies was 20 (SD ± 11.4) [range 9-40]. All of the studies were
163
conducted in adults. Three of the studies utilized control populations. A description of the studies
164
is provided in the evidentiary table (Table 2).35-39
SC
RI PT
159
165
The majority of the studies consisted of individuals with a chronic balance dysfunction due to
167
central or peripheral etiologies that affected balance, posture, and gait. Of these, the most
168
common peripheral conditions were gentamycin ototoxicity, and endolymphatic hydrops; the
169
most common central causes were Mal de Debarquement, cerebellar stroke, traumatic brain
170
injury and idiopathic causes. Other, less common, conditions included acoustic neuroma,
171
meningitis/encephalitis, and Parkinson’s Disease. Excluded conditions were pregnancy, mental
172
health disorders, myasthenia gravis, Charcot-Marie-Tooth disease, post-polio syndrome, Guillan-
173
Barré, fibromyalgia, chronic fatigue syndrome, herniated disc, and osteoarthritis of the spine.
174
Almost all of the subjects in the studies had previously completed standard vestibular
175
rehabilitation therapy and were at least one year from their acute condition. Most had well-
176
adapted compensatory strategies for coping with the debilitating effects of their condition.
177
Nonetheless, all subjects still had difficulty with standing, walking, or moving.
AC C
EP
TE D
M AN U
166
178
179
The assessments of study quality was performed using the Newcastle-Ottawa Quality
180
Assessment Scale for Nonrandomized Studies in Meta-Analysis and are shown in Table 3. Study 8
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
quality was variable among the articles with improvements in quality in later experiments. For
182
example, the earlier studies 38,39 did not have control groups compared to subsequent ones.35-37
183
Some of the factors that limited study quality included selection bias from referral populations,
184
lack of blinding of the investigators/assessors to the device and outcome, and the lack of
185
reporting of losses to follow-up or treatment compliance.
RI PT
181
SC
186
The outcomes used in the selected studies were established measures of balance and function
188
including the Dynamic Gait Index (DGI – a 24-point scale with higher scores representing better
189
ambulation), Dizziness Handicap Inventory (DHI – a 100-point self-assessment scale with higher
190
scores indicating greater perceived handicap from dizziness), Activities-specific Balance
191
Confidence Scale (ABC – a 100-point self-assessment scale with higher scores representing
192
greater perceived confidence in performing activities), and Sensory Organization Test (SOT -
193
uses dynamic posturography to measure postural sway in seconds with higher scores indicating
194
better balance). Additionally, functional MRI and other self-perception of impairment measures
195
were also employed. Each study used a variation of these outcome measures before and after
196
CN-NINM stimulation. Only two of the studies (Danilov 2007 and Wildenberg 2011)35,38 used
197
the same outcome tools with adequate data for the studies to be combined in a meta-analysis.
TE D
EP
AC C
198
M AN U
187
199
Both of the studies (n=40) included in the meta-analysis reported improvement in the three
200
outcome measures of Dynamic Gait Index (DGI), Activities-specific Balance Confidence Scale
201
(ABC), and Dizziness Handicap Inventory (DHI). Funnel plots were constructed to assess
202
systematic heterogeneity (Figure 2). For all three indices, heterogeneity was non-significant 9
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
(I2=0% and χ2<25%) (Figures 2 and 3) and funnel plots did not show any asymmetry so a fixed
204
effect model was used to assess effect size. Effect sizes were very similar between both studies.
205
The pooled results demonstrated a significant improvement in DGI post-intervention compared
206
to before the intervention with the CN-NINM device, with a mean difference 3.45 (95%CI 1.75,
207
5.15) and overall effect of Z=3.98; p<0.001 (Figure 3a). Similarly, the pooled results for the
208
ABC scale also indicated a significant improvement post-intervention, with a mean difference
209
16.65 (95%CI 7.65, 25.47) and overall effect of Z=3.64; p<0.001 (Figure 3b). For the DHI,
210
pooled results pre and post CN-NINM intervention showed a mean difference of -26.07 (95%CI
211
-35.78, -16.35) with an overall effect of Z=5.26; p<0.001 (Figure 3c).
M AN U
SC
RI PT
203
212
Although only two studies were pooled, the other three studies also demonstrated improvements
214
following treatment with CN-NINM. The study by Danilov et al. in 200639 was more of a
215
descriptive study that attained various positive observational results, such as greater inter-limb
216
coordination, endurance, and gait components that the subjects were not previously capable of.
217
The SOT results showed an average improvement of 50%. Retained benefits were also reported
218
in a subsequent article38 without continual usage of the CN-NINM device. A positive correlation
219
was established between the amount of time using the CN-NINM device and the length of time
220
subjects retained benefits. However, no correlation was found between usage time and change in
221
scores on the composite SOT, DHI, ABC or DGI. Among the lasting benefits recorded was
222
improvement in daily activities requiring balance and gait including but not limited to getting
223
dressed and traversing uneven terrain. No subjects reported adverse or negative side effects in
224
any of the studies reviewed.
AC C
EP
TE D
213
10
ACCEPTED MANUSCRIPT Systematic Review CN-NINM 225
Functional MRI was performed in three of the studies.35-37 In the 2010 study by Wildenberg et
227
al.37 a region within the pons exhibited sustained neuromodulation due to electrical tongue
228
stimulation by the CN-NINM on fMRI, but was unable to precisely identify the neuronal
229
structure involved. In a subsequent study,36 high-resolution imaging of the brainstem and
230
cerebellum was used to precisely localize sustained subcortical neuromodulation induced by CN-
231
NINM. The anterior cingulate cortex showed decreased activity after stimulation, while a region
232
within the pons had increased post-stimulation activity. These observations suggested that the
233
pontine region was the trigeminal nucleus and that tongue stimulation interfaced with the
234
balance-processing network within the pons. However, there were also improvements in
235
symptoms that appeared unrelated to the anatomical targets of the stimulation, suggesting that
236
CN-NINM induced wide-ranging neuroplastic changes that helped integrate diverse sensory
237
input.37,38 The investigators hypothesized that changes in cortical activity occurred through the
238
way information was processed from subcortical structures, through the thalamus, to the cortex.
239
This is consistent with a PET study indicating that the stimulation signal propagates through the
240
thalamus to multiple cortical regions.40
EP
TE D
M AN U
SC
RI PT
226
242
243
AC C
241
DISCUSSION
244
This systematic review summarizes the available evidence for the use of CN-NINM in treating
245
nervous system disorders. We identified five articles that prospectively examined the CN-NINM
246
device, two of which, had adequately described outcomes that could be combined for a meta11
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
analysis. These quasi-experimental observational studies used referral populations of nervous
248
system disorders with balance problems of different etiologies in order to examine the efficacy of
249
the device. Patients were treated with the CN-NINM device multiple times daily for a variable
250
length of time. The outcome measure of these studies consisted of an assortment of standardized
251
tests of balance and function, along with qualitative measures, that universally showed
252
improvements among test subjects.
SC
RI PT
247
253
There are a number of potential advantages to using CN-NINM for rehabilitation of nervous
255
system disorders. Firstly, the improvements seen in patients with chronic balance dysfunction
256
from various causes demonstrates the versatility of the CN-NINM device for different peripheral
257
and central conditions. Secondly, CN-NINM is non-invasive. Clinical application of implanted
258
nerve stimulators are frequently limited by their invasiveness and by their associated risks from
259
surgery, infections and tissue damage. Thirdly, fMRI has shown that the neuroplastic changes
260
induced after application of the CN-NINM device impact several regions of the brain. These
261
functional changes are retained after just 5 days of training with the device. Fourthly, the device
262
is portable and can be used at home, without constant medical supervision, given proper training.
263
This provides the opportunity to increase the availability and frequency of neurorehabilitation.
264
Most importantly, none of the studies have reported adverse effects from use of the device.
266
TE D
EP
AC C
265
M AN U
254
STUDY LIMITATIONS
267
12
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
Although these results are encouraging, there are important limitations that should be discussed.
269
Publication bias is a possibility given that studies reporting negative findings are less likely to be
270
published and more likely to be excluded from systematic reviews. However, we were able to
271
contact two of the authors to verify the existence of any unpublished data showing negative
272
results. There was no such data, so it is unlikely that publication bias influenced the conclusions.
273
In order to minimize selection bias three independent reviewers screened the literature for
274
possible study inclusion and all abstracts and primary manuscripts were assessed using
275
standardized eligibility criteria.
M AN U
SC
RI PT
268
276
Generally, the quality of a meta-analysis is a reflection of the included studies. There were small
278
sample sizes included in each of the studies reviewed and only two studies had adequately
279
described data with outcomes that could be pooled. This represents a total of only forty patients
280
in the meta-analysis. Accordingly, the small sample size limited the ability to conduct subgroup
281
analyses to identify patient characteristics that are most likely to benefit from this treatment.
282
Although the data was limited, one of the studies examined results in subgroups of age and
283
etiology (peripheral versus central) and showed little to no differences in treatment effects
284
between the groups. Another important limitation of the current literature on CN-NINM includes
285
the lack of appropriate placebo controlled groups. We can only speculate that placebo controlled
286
groups were not used due to the cost and the challenge of finding patients willing to enter a
287
lengthy study with the possibility of not being treated. Furthermore, the lack of blinding at
288
selection, during assessments, and of the patients themselves is another important factor that
289
could lead to bias in these studies. These will be important factors to consider in the design of
290
future studies. Other methodological issues consist of referral populations with very broad
AC C
EP
TE D
277
13
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
inclusion criteria and protocols that utilize varying lengths of training time with the CN-NINM
292
device. In certain experiments, the CN-NINM was used for a variable number of days among
293
patients depending on stamina and ability to complete follow-ups. Additionally, the length and
294
type of the follow-up assessments also varied within the same cohort. It will be important to
295
reduce this variability in treatment regimens in subsequent studies in order to establish optimal
296
treatment schedules and to maximize retained benefits.
SC
RI PT
291
297
Going forward, it will be necessary to maintain a high level of methodological rigor in assessing
299
this device, including the application randomized clinical controlled trials with larger sample
300
sizes that are powered to detect clinically relevant differences. It will be critical to include
301
standard diagnostic criteria, comprehensive measurement of outcomes conducted by blinded
302
examiners, high retention rates, measurement of intervention fidelity, the effect of adjuvant
303
therapies, and objective validated measures of change to monitor progress and improvement.
304
Despite these limitations, the studies reviewed show promising benefits to the use of the CN-
305
NINM device on balance, function and sensory-motor coordination. The ability to non-
306
invasively produce sustained modulation of neural activity holds promise as a new route for
307
therapy in individuals with neurological dysfunction.
309
TE D
EP
AC C
308
M AN U
298
CONCLUSION
310
14
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
This systematic review describes a number of studies using the CN-NINM device that appear to
312
improve balance and sensory-motor coordination in a myriad of peripheral and central
313
conditions. However, there are several limitations to the studies evaluated in this systematic
314
review that include small sample sizes, lack of randomization and control groups, absence of
315
blinding, use of referral populations, and variability in treatment schedules and follow-up rates.
316
Given these limitations, we recognize that a meta-analysis pooling only two studies is sub-
317
optimal and the results must be interpreted cautiously. Future studies of the CN-NINM device
318
will require very rigorously conducted randomized controlled studies with blinding and well-
319
defined outcome measures. In summary, further study of the CN-NINM device is required before
320
widespread clinical application.
321
TE D
322
M AN U
SC
RI PT
311
323
AC C
EP
324
15
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
REFERENCES
326
1. Young JA, Tolentino M. Neuroplasticity and its applications for rehabilitation. Am J Ther
327
2011;18:70-80.
328
2. Boudreau SA, Farina D, Falla D. The role of motor learning and neuroplasticity in designing
329
rehabilitation approaches for musculoskeletal pain disorders. Man Ther 2011;15:410-4.
330
3. De FP, Fellus J, Polito MZ, Thompson JW, Moser RS, DeLuca J. The new neuroscience
331
frontier: promoting neuroplasticity and brain repair in traumatic brain injury. Clin Neuropsychol
332
2009;23:1391-9.
333
4. Levin HS. Neuroplasticity and brain imaging research: implications for rehabilitation. Arch
334
Phys Med Rehabil 2006;87:S1.
335
5. Kuo MF, Grosch J, Fregni F, Paulus W, Nitsche MA. Focusing effect of acetylcholine on
336
neuroplasticity in the human motor cortex. J Neurosci 2007;27:14442-7.
337
6. Engineer ND, Moller AR, Kilgard MP. Directing neural plasticity to understand and treat
338
tinnitus. Hear Res 2012;295:58-66.
339
7. Kaufman M, Corner MA, Ziv NE. Long-term relationships between cholinergic tone,
340
synchronous bursting and synaptic remodeling. PLoS One 2012;7:e40980.
341
8. Roosevelt RW, Smith DC, Clough RW, Jensen RA, Browning RA. Increased extracellular
342
concentrations of norepinephrine in cortex and hippocampus following vagus nerve stimulation
343
in the rat. Brain Res 2006;1119:124-32.
344
9. Andrade DC, Borges I, Bravo GL, Bolognini N, Fregni F. Therapeutic time window of
345
noninvasive brain stimulation for pain treatment: inhibition of maladaptive plasticity with early
346
intervention. Expert Rev Med Devices 2013;10:339-52.
AC C
EP
TE D
M AN U
SC
RI PT
325
16
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
10. Villamar MF, Santos Portilla A, Fregni F, Zafonte R. Noninvasive brain stimulation to
348
modulate neuroplasticity in traumatic brain injury. Neuromodulation 2012;15:326-38.
349
11. Fregni F, Pascual-Leone A. Technology insight: noninvasive brain stimulation in neurology-
350
perspectives on the therapeutic potential of rTMS and tDCS. Nat Clin Pract Neurol 2007;3:383-
351
93.
352
12. Marangell LB, Martinez M, Jurdi RA, Zboyan H. Neurostimulation therapies in depression: a
353
review of new modalities. Acta Psychiatr Scand 2007;116:174-81.
354
13. Adeyemo BO, Simis M, Macea DD, Fregni F. Systematic review of parameters of
355
stimulation, clinical trial design characteristics, and motor outcomes in non-invasive brain
356
stimulation in stroke. Front Psychiatry 2012;3:88.
357
14. Goodman RR, Kim B, McClelland S, 3rd, Senatus PB, Winfield LM, Pullman SL, Yu Q,
358
Ford B, McKhann GM, 2nd. Operative techniques and morbidity with subthalamic nucleus deep
359
brain stimulation in 100 consecutive patients with advanced Parkinson's disease. J Neurol
360
Neurosurg Psychiatry 2006;77:12-7.
361
15. Fasano A, Daniele A, Albanese A. Treatment of motor and non-motor features of Parkinson's
362
disease with deep brain stimulation. Lancet Neurol 2012;11:429-42.
363
16. Benabid AL, Chabardes S, Mitrofanis J, Pollak P. Deep brain stimulation of the subthalamic
364
nucleus for the treatment of Parkinson's disease. Lancet Neurol 2009;8:67-81.
365
17. Handforth A, DeGiorgio CM, Schachter SC, Uthman BM, Naritoku DK, Tecoma ES, Henry
366
TR, Collins SD, Vaughn BV, Gilmartin RC, Labar DR, Morris GL, 3rd, Salinsky MC, Osorio I,
367
Ristanovic RK, Labiner DM, Jones JC, Murphy JV, Ney GC, Wheless JW. Vagus nerve
368
stimulation therapy for partial-onset seizures: a randomized active-control trial. Neurology
369
1998;51:48-55.
AC C
EP
TE D
M AN U
SC
RI PT
347
17
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
18. George MS, Rush AJ, Sackeim HA, Marangell LB. Vagus nerve stimulation (VNS): utility in
371
neuropsychiatric disorders. Int J Neuropsychopharmacol 2003;6:73-83.
372
19. Marangell LB, Rush AJ, George MS, Sackeim HA, Johnson CR, Husain MM, Nahas Z,
373
Lisanby SH. Vagus nerve stimulation (VNS) for major depressive episodes: one year outcomes.
374
Biol Psychiatry 2002;51:280-7.
375
20. Marangell LB, Suppes T, Zboyan HA, Prashad SJ, Fischer G, Snow D, Sureddi S, Allen JC.
376
A 1-year pilot study of vagus nerve stimulation in treatment-resistant rapid-cycling bipolar
377
disorder. J Clin Psychiatry 2008;69:183-9.
378
21. Nahas Z, Marangell LB, Husain MM, Rush AJ, Sackeim HA, Lisanby SH, Martinez JM,
379
George MS. Two-year outcome of vagus nerve stimulation (VNS) for treatment of major
380
depressive episodes. J Clin Psychiatry 2005;66:1097-104.
381
22. Park MC, Goldman MA, Carpenter LL, Price LH, Friehs GM. Vagus nerve stimulation for
382
depression: rationale, anatomical and physiological basis of efficacy and future prospects. Acta
383
Neurochir Suppl 2007;97:407-16.
384
23. Edwardson MA, Lucas TH, Carey JR, Fetz EE. New modalities of brain stimulation for
385
stroke rehabilitation. Exp Brain Res 2013;224:335-58.
386
24. Bandt SK, Anderson D, Biller J. Deep brain stimulation as an effective treatment option for
387
post-midbrain infarction-related tremor as it presents with Benedikt syndrome. J Neurosurg
388
2008;109:635-9.
389
25. Fitzgerald PB, Daskalakis ZJ. The use of repetitive transcranial magnetic stimulation and
390
vagal nerve stimulation in the treatment of depression. Curr Opin Psychiatry 2008;21:25-9.
391
26. Silberstein SD, Dodick DW, Saper J, Huh B, Slavin KV, Sharan A, Reed K, Narouze S,
392
Mogilner A, Goldstein J, Trentman T, Vaisma J, Ordia J, Weber P, Deer T, Levy R, Diaz RL,
AC C
EP
TE D
M AN U
SC
RI PT
370
18
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
Washburn SN, Mekhail N. Safety and efficacy of peripheral nerve stimulation of the occipital
394
nerves for the management of chronic migraine: results from a randomized, multicenter, double-
395
blinded, controlled study. Cephalalgia 2012;32:1165-79.
396
27. Mwenge GB, Rombaux P, Dury M, Lengele B, Rodenstein D. Targeted hypoglossal
397
neurostimulation for obstructive sleep apnoea: a 1-year pilot study. Eur Respir J 2012;41:360-7.
398
28. Milby AH, Halpern CH, Baltuch GH. Vagus nerve stimulation in the treatment of refractory
399
epilepsy. Neurotherapeutics 2009;6:228-37.
400
29. Hoppe C. Vagus nerve stimulation: Urgent need for the critical reappraisal of clinical
401
effectiveness. Seizure 2013;22:83-4.
402
30. Fregni F, Freedman S, Pascual-Leone A. Recent advances in the treatment of chronic pain
403
with non-invasive brain stimulation techniques. Lancet Neurol 2007;6:188-91.
404
31. Tyler ME, Braun JG, Danilov YP. Spatial mapping of electrotactile sensation threshold and
405
intensity range on the human tongue: initial results. Conf Proc IEEE Eng Med Biol Soc
406
2009;2009:559-62.
407
32. Tyler M, Danilov Y, Bach YRP. Closing an open-loop control system: vestibular substitution
408
through the tongue. J Integr Neurosci 2003;2:159-64.
409
33. Wells GA, Shea B, O'Connell D, Peterson J, Welch V, Losos M, Tugwell P. The Newcastle-
410
Ottawa Scale (NOS) for Assessing the Quality of Nonrandomized Studies in Meta-Analysis
411
http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp.
412
34. Oremus M, Oremus C, Hall GB, McKinnon MC. Inter-rater and test-retest reliability of
413
quality assessments by novice student raters using the Jadad and Newcastle-Ottawa Scales. BMJ
414
Open;2.
AC C
EP
TE D
M AN U
SC
RI PT
393
19
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
35. Wildenberg JC, Tyler ME, Danilov YP, Kaczmarek KA, Meyerand ME. Electrical tongue
416
stimulation normalizes activity within the motion-sensitive brain network in balance-impaired
417
subjects as revealed by group independent component analysis. Brain Connect 2011;1:255-65.
418
36. Wildenberg JC, Tyler ME, Danilov YP, Kaczmarek KA, Meyerand ME. High-resolution
419
fMRI detects neuromodulation of individual brainstem nuclei by electrical tongue stimulation in
420
balance-impaired individuals. Neuroimage 2011;56:2129-37.
421
37. Wildenberg JC, Tyler ME, Danilov YP, Kaczmarek KA, Meyerand ME. Sustained cortical
422
and subcortical neuromodulation induced by electrical tongue stimulation. Brain Imaging Behav
423
2010;4:199-211.
424
38. Danilov YP, Tyler ME, Skinner KL, Hogle RA, Bach-y-Rita P. Efficacy of electrotactile
425
vestibular substitution in patients with peripheral and central vestibular loss. J Vestib Res
426
2007;17:119-30.
427
39. Danilov YP, Tyler ME, Skinner KL, Bach-y-Rita P. Efficacy of electrotactile vestibular
428
substitution in patients with bilateral vestibular and central balance loss. Conf Proc IEEE Eng
429
Med Biol Soc 2006;Suppl:6605-9.
430
40. Ptito M, Moesgaard SM, Gjedde A, Kupers R. Cross-modal plasticity revealed by
431
electrotactile stimulation of the tongue in the congenitally blind. Brain 2005;128:606-14.
433
SC
M AN U
TE D
EP
AC C
432
RI PT
415
20
ACCEPTED MANUSCRIPT
Table Legend Table 1. Composition of search phrases for literature search strategy
RI PT
Table 2: Evidentiary table summarizing clinical studies on CN-NINM device
M AN U
SC
Table 3. Newcastle - Ottawa Quality Assessment Scale for Observational Studies The Newcastle-Ottawa Quality Assessment Scale was developed to assess the quality on nonrandomized studies. Assessments are intended to be incorporated into meta-analytic results. It is a star system based on three domains: 1) Selection of StudyGroups; 2) Comparability of Groups; and 3) Ascertainment of exposure/outcome.
Figure Legend Figure 1. Study Selection Process
TE D
Figure 2a, b, c. Funnel plots of comparisons between outcome measures pre and post intervention
AC C
EP
Figure 2a. Funnel plot of comparison: Dynamic Gait Index (DGI) Pre and Post Intervention with CN-NINIM. Figure 2b. Funnel plot of comparison: Activities-specific Balance Confidence Scale (ABC) Pre and Post Intervention with CN-NINIM. Figure 2c. Funnel plot of comparison: Dizziness Handicap Inventory (DHI) Pre and Post Intervention with CN-NINIM
Figure 3a, b, c. Forest plots of comparisons between outcome measures pre and post intervention with CN-NINIM Figure 3a. Forest plot of comparison: Dynamic Gait Index (DGI) Pre and Post Intervention with CN-NINIM. Figure 3b. Forest plot of comparison: Activities-specific Balance Confidence Scale (ABC) Pre and Post Intervention with CN-NINIM. Figure 3c. Forest plot of comparison: Dizziness Handicap Inventory (DHI) Pre and Post Intervention with CN-NINIM
ACCEPTED MANUSCRIPT
Table 1. Composition of search phrases for literature search strategy Terms included in nervous system disorders search phrase
Terms included in the treatment search phrase
Cranial nerve modulation
Central nervous system (CNS) disease/disorder Peripheral nervous system (PNS) disease/disorder Ataxia/Balance disorder Neuropathy Brain Injury Traumatic Brain Injury/TBI Neurological disease/disorder Brain/Cerebral/Cerebellar
Rehabilitation
M AN U
Therapy Recovery
SC
Treatment
TE D EP
Noninvasive Nerve stimulator Nerve modulator PoNS® BrainPort® Tongue simulator
AC C
Cranial nerve modulator
RI PT
Terms included in the cranial nerve modulation search phrase
ACCEPTED MANUSCRIPT
Table 2: Evidentiary table summarizing clinical studies on CN-NINM device
Electrical Tongue Stimulation Normalizes Activity Within the Motion-Sensitive Brain Network in BalanceImpaired Subjects as Revealed by Group Independent Component Analysis
Study #2 2011 Wilden berg et al.34
High-resolution fMRI detects neuromodulation of individual brainstem nuclei by electrical tongue stimulation in balance-impaired individuals
Study #3 2010 Wilden berg et al.35
Sustained cortical and subcortical neuromodulation induced by electrical tongue stimulation
Study #4 2007 Danilo v et al.36
Efficacy of electrotactile vestibular substitution in patients with peripheral and central vestibular loss
Study #5 2006 Danilo v et al.37
Efficacy of electrotactile vestibular substitution in patients with bilateral vestibular and central balance loss
Sample Size
Outcome Measures
Intervention
Results
Prospective Quasi-experimental With controls
12 subjects with symptoms of chronic balance disorder and 9 normal controls
-Measured Day 0 and Day 5 (Controls only assessed on Day 0) -Dynamic gait index (DGI) -Self-perception of impairment -Activities-specific Balance Composite scale -Dizziness Handicap Index (DHI) -fMRI (3T)
-CN-NINM was delivered only to the balance-impaired subjects over 9 sessions (two on days 1–4 and one on day 5) -During a stimulation session, subjects received continuous stimulation for 20 min while standing as still as possible with their eyes closed
Pre-intervention mean DGI score=19.0±4.3, mean DHI score=58.8±23.8 (lower score indicates less impairment), and an ABC score=57.9±24.4. Post-intervention mean DGI score=22.9±1.4 (p<0.005), mean DHI score=35.7±24.0 (p<0.0005), and an ABC score=75.3±21.2 (p<0.005). Global improvement (p< 0.005).
Prospective Quasi-experimental With controls
9 subjects with symptoms of chronic balance disorder and 9 normal controls
-Sensory Organization Testing (SOT) (a composite score that combines performance on six sensory conditions and measures overall postural stability [<70 abnormal]) -fMRI
-CN-NINM was delivered only to the balance-impaired subjects over 19 sessions (two on days 1–9 and one on day 10). Days 5 and 6 were separated by two rest days (no stimulation) -During a stimulation session, subjects received continuous stimulation for 20 min while standing as still as possible with their eyes closed
Pre-intervention Balance-impaired Subject #3 did not receive an SOT score pre-intervention due to significant stability impairment. 6/8 had SOT scores<70 Post-intervention Subject #3 received an SOT=62. 2/8 still had SOT <70. Mean improvement in SOT scores was 15.75±SEM 5.59 (p=0.026).
Prospective Quasi-experimental With controls
12 subjects with symptoms of chronic balance disorder and 9 normal controls
-Postural sway (MATLAB) (Only data from the anterior-posterior sway was used for analysis) -fMRI
-CN-NINM was delivered only to the balance-impaired subjects over 9 sessions (two on days 1–4 and one on day 5) -During a stimulation session, subjects received continuous stimulation for 20 min while standing as still as possible with their eyes closed
Pre-intervention The effect of optic flow on postural sway was greater in balance subjects than in controls. Post-intervention There was no difference in sway amplitude between balance subjects postintervention and controls. Balance subjects swayed more in response to the optic flow stimulus preintervention than post-intervention (p≤0.005).
Prospective Quasi-experimental
28 subjects with symptoms of chronic balance disorder
-Computerized Dynamic Posturography (CDP) -Sensory Organization Test (SOT) -Dynamic Gait Index (DGI) - Activities-specific Balance Confidence Scale (ABC) -Dizziness Handicap Inventory (DHI)
-CN-NINM was delivered to each balance-impaired subject twice daily for 34.5 days -Each stimulation session was 1-1.5 hours -During a stimulation session, subjects received a succession of shorter trial periods (1-5 minutes) and one 20 minute trial period -Subjects were to stand as still as possible while receiving stimulation
Prospective Quasi-experimental
40 subjects with symptoms of chronic balance disorder
-Sensory Organization Test (SOT) -Dynamic Gait Index (DGI) - Activities-specific Balance Confidence Scale (ABC) -Dizziness Handicap Inventory (DHI)
-CN-NINM was delivered to each balance-impaired subject over 9 sessions -Each stimulation session was 1.5-2 hours - During a stimulation session, subjects received continuous stimulation for 20 min while standing as still as possible with their eyes closed
Pre-intervention -mean DGI score=18.4±5.3 -mean DHI score=57.3±19.6 (lower score indicates less impairment -ABC score=61.7±20.4 -mean SOT score=48.5±18.3 Post-intervention -mean DGI score=21.5±3.1 (p<0.001) -mean DHI score=30.2±23.3 (p<0.001) (lower score indicates less impairment -ABC score=78.0±18.5 (p<0.001) -mean SOT score=65.0±17.1 (p<0.001) Pre-intervention -Unavailable Post-intervention -Average composite SOT improvement: 49.1% -All subjects ‘generally’ improved in functional transfer testing (DGI, ABC, and DHI), with the exception of four subjects exhibiting no change in the DGI
RI PT
Study #1 2011 Wilden berg et al.33
Design
SC
Title
M AN U
Study
AC C
EP
TE D
*FOLLOW UP ANALYSIS OF STUDY #1
ACCEPTED MANUSCRIPT
Table 3. Newcastle - Ottawa Quality Assessment Scale for Observational Studies
Outcome/Exposure (3 Stars)
**
*
*
**
*
*
**
*
*
*
-
*
-
RI PT
Comparability (2 Stars)
SC
Study #1 2011 Wildenberg et al.33 Study #2 2011 Wildenberg et al.34 Study #3 2010 Wildenberg et al.35 Study #4 2007 Danilov et al.36 Study #5 2006 Danilov et al.37
Selection (4 Stars)
M AN U
Study
* *
AC C
EP
GA Wells, B Shea, D O'Connell, J Peterson, V Welch, M Losos, P Tugwell. http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp
TE D
The Newcastle-Ottawa Quality Assessment Scale was developed to assess the quality on non-randomized studies. Assessments are intended to be incorporated into meta-analytic results. It is a star system based on three domains: 1) Selection of Study Groups; 2) Comparability of Groups; and 3) Ascertainment of exposure/outcome.
ACCEPTED MANUSCRIPT Figure 1. Study Selection Process 0 articles identified by review of Cochrane Database
50 articles identified by PUBMED/MEDLINE search
SC
M AN U
9 titles selected for further review after screening titles/abstracts
RI PT
41 articles did not meet inclusion based on title and abstract screening
AC C
EP
TE D
2 additional articles identified by review of selected bibliographies
1 article identified by contacting authors
12 articles had full text reviewed
7articles excluded - 4 did not fit the primary search criteria - 3 did not have any functional outcomes measured
5 primary studies included in this systematic review
2 studies with usable outcome data for metaanalysis
ACCEPTED MANUSCRIPT
Figure 2a, b, c. Funnel plots of comparisons between outcome measures pre and post intervention with CN-NINM
M AN U
SC
RI PT
2a. Funnel plot of comparison: Dynamic Gait Index (DGI) Pre and Post Intervention with CN-NINIM
AC C
EP
TE D
2b. Funnel plot of comparison: Activities-specific Balance Confidence Scale (ABC) Pre and Post Intervention with CN-NINIM
2c. Funnel plot of comparison: Dizziness Handicap Inventory (DHI) Pre and Post Intervention with CN-NINIM
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT Figure 3a, b, c. Forest plots of comparisons between outcome measures pre and post intervention with CN-NINIM
M AN U
SC
RI PT
3a. Forest plot of comparison: Dynamic Gait Index (DGI) Pre and Post Intervention with CN-NINIM
EP
TE D
3b. Forest plot of comparison: Activities-specific Balance Confidence Scale (ABC) Pre and Post Intervention with CN-NINIM
AC C
3c. Forest plot of comparison: Dizziness Handicap Inventory (DHI) Pre and Post Intervention with CN-NINIM
S0003-9993(14)00335-9
DOI:
10.1016/j.apmr.2014.04.018
Reference:
YAPMR 55823
To appear in:
ARCHIVES OF PHYSICAL MEDICINE AND REHABILITATION
Received Date: 6 January 2014 Revised Date:
21 March 2014
Accepted Date: 14 April 2014
Please cite this article as: Papa L, LaMee A, Tan CN, Hill-Pryor C, Systematic Review and MetaAnalysis of Noninvasive Cranial Nerve Neuromodulation for Nervous System Disorders, ARCHIVES OF PHYSICAL MEDICINE AND REHABILITATION (2014), doi: 10.1016/j.apmr.2014.04.018. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
Systematic Review and Meta-Analysis of Noninvasive Cranial Nerve Neuromodulation for Nervous System Disorders
RI PT
Linda Papa, MDCM, MSc1; Alexander LaMee 2; Ciara N. Tan1; Crystal Hill-Pryor, PhD3 1
Orlando Regional Medical Center, Orlando, Florida 2 University of Central Florida, Orlando, Florida 3 U.S. Army Medical Research and Materiel Command, Fort Detrick, Maryland
M AN U
Word Count: 3,189 (excluding abstract and references)
SC
March 19, 2014
Tables & Figures: 6 Number of Authors: 4 Number of Institutions: 3
Linda Papa, MDCM, MSc
TE D
Director of Academic Clinical Research and Attending Emergency Physician Department of Emergency Medicine Orlando Regional Medical Center 86 W. Underwood (S-200)
Tel.: 407-237-6329 Fax: 407-649-3083
AC C
[email protected]
EP
Orlando, Florida, 32806
Alexander LaMee
University of Central Florida Department of Biomedical Sciences [email protected]
Ciara N. Tan, BS Department of Emergency Medicine
1
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
Orlando Regional Medical Center [email protected]
Crystal Hill-Pryor, PhD
RI PT
Neurotrauma Research Deputy
Combat Casualty Care Research Program U.S. Army Medical Research and Materiel Command
SC
[email protected]
CORRESPONDING AUTHOR:
M AN U
Linda Papa, MDCM, MSc
Director of Academic Clinical Research and Attending Emergency Physician Department of Emergency Medicine Orlando Regional Medical Center 86 W. Underwood (S-200) Orlando, Florida, 32806 Tel.: 407-237-6329
TE D
Fax: 407-649-3083 [email protected]
EP
Key Words: Cranial nerve modulator, rehabilitation, central nervous system disorders,
PoNS™, neuroplasticity, balance disorders, neurological disorders, neurodegenerative
fMRI
AC C
disorders, traumatic brain injury, non-invasive neuromodulation, cranial nerve stimulator,
Grant Support: None
Author Disclosure Statement: None
2
ACCEPTED MANUSCRIPT Systematic Review CN-NINM 1
Systematic Review and Meta-Analysis of Noninvasive Cranial Nerve Neuromodulation for
2
Nervous System Disorders
3
ABSTRACT
5
Objective: To systematically review the medical literature and comprehensively summarize
6
clinical research done on rehabilitation with a novel portable and non-invasive electrical
7
stimulation device, called the cranial nerve noninvasive neuromodulator (CN-NINM) in patients
8
suffering from nervous system disorders. The CN-NINM induces processes of neuroplasticity by
9
noninvasive stimulation of four cranial nerves and targets the subcortical area of the brain. Data
10
Sources: PubMed®, MEDLINE® and the Cochrane Database from 1966 to March 2013. Study
11
Selection: Studies were included if they recruited adult patients with peripheral and central
12
nervous system disorders and were treated with CN-NINM and were assessed with objective
13
measures of function. Data Extraction: After title and abstract screening of potential articles,
14
full texts were independently reviewed to identify articles that met inclusion criteria. Data
15
Synthesis: The search identified twelve publications, five were critically reviewed and two
16
combined in a meta-analysis. There were no randomized controlled studies identified and the
17
meta-analysis was based on pre-post studies. Most of the patients were individuals with a chronic
18
balance dysfunction. The pooled results demonstrated significant improvements in 1) Dynamic
19
Gait Index (DGI) post-intervention with a mean difference 3.45 (1.75, 5.15)(p<0.001); 2) the
20
ABC scale with a mean difference of 16.65 (7.65, 25.47) (p<0.001); and 3) the DHI with
21
improvements of -26.07 (-35.78, -16.35)(p<0.001). Included studies suffered from small sample
22
sizes, lack of randomization, absence of blinding, use of referral populations, and variability in
23
treatment schedules and follow-up rates. Conclusions: Given these limitations, the results of the
AC C
EP
TE D
M AN U
SC
RI PT
4
1
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
meta-analysis must be interpreted cautiously. Further investigation using rigorous randomized
25
clinical controlled trials are needed to evaluate this promising rehabilitation tool for nervous
26
system disorders.
RI PT
24
27
Key Words:
29
Cranial nerve modulator; rehabilitation; nervous system disorders; neuroplasticity; non-invasive
30
neuromodulation
M AN U
SC
28
31
Abbreviations:
33
ABC - Activities-specific Balance Confidence Scale
34
CDP - Computerized Dynamic Posturography
35
CN-NINM – Cranial Nerve Non-Invasive NeuroModulation
36
CNS – Central Nervous System
37
DGI – Dynamic Gait Index
38
DHI - Dizziness Handicap Inventory
39
fMRI – Functional Magnetic Resonance Imaging
40
PNS – Peripheral Nervous System
41
SOT - Sensory Organization Test
42
TBI – Traumatic Brain Injury
AC C
EP
TE D
32
43
2
ACCEPTED MANUSCRIPT Systematic Review CN-NINM 44
INTRODUCTION
45
The goal of rehabilitation programs for patients suffering from nervous system disorders is to
47
achieve the highest level of functioning possible. Part of this rehabilitation involves adaptive
48
changes and dynamic remodeling of the structure and function of cells and tissues in response to
49
both endogenous and exogenous environmental stimuli.1,2 This adaptive ability of the nervous
50
system is referred to as “neuroplasticity.”1,3 Neuroplasticity allows the injured brain to remodel
51
or repair itself by recruiting brain regions that are not impaired, thus increasing neural
52
connections to these areas. This increased activation makes it possible for uninjured brain to
53
compensate for the functions of the injured parts.4 Neuroplasticity can also occur through
54
neuromodulation via neurotransmitters such as serotonin, dopamine, acetylcholine, and
55
histamine. These neurotransmitters or “neuromodulators” are secreted by a small group of
56
neurons and diffuse through large areas of the nervous system to regulate nervous system
57
activity.5-7,8 Neuromodulation occurs naturally in the brain but may be altered by extraneous
58
factors such as stress, medications or electrical stimulation.9
TE D
M AN U
SC
RI PT
46
EP
59
The use of external stimulation to change and regulate the internal electrochemical environment
61
of the brain is a rapidly growing area of research.10-13 Most electrical stimulation methods
62
currently being applied are invasive and include (1) deep brain stimulation (DBS) via implanted
63
microelectrodes in the basal ganglia, thalamus, and in the midbrain for treatment of Parkinson’s
64
disease and chronic pain14-16; (2) vagus nerve stimulation (VNS) via implanted electrodes on the
65
vagus nerve for suppression of epileptic seizures and depression;17-22 (3) subdural implantable
66
stimulators for stroke recovery;23,24 and (4) transcranial magnetic stimulation (TMS) via an
AC C
60
3
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
external magnet applied over the skull to stimulate the cerebral cortex for treatment of
68
depression.12,25 Moreover, peripheral nerve stimulation has also been shown to potentially
69
reduce pain due to chronic migraine headaches, but a common adverse effect is persistent pain at
70
the implant site.26 Similarly, hypoglossal neurostimulation has been used to improve symptoms
71
in patients with severe obstructive sleep apnea, but the side effects of surgical implantation limit
72
its use.27 Vagus nerve stimulation has seen positive results in the treatment of epileptic
73
conditions such as refractory epilepsy.28 Though generally regarded as a safe alternative to
74
intracranial surgery, one study reported that 51% of patients experienced side effects requiring
75
the VNS device be switched off or explanted.29 Non-invasive stimulation techniques are more
76
appealing yet not fully validated. Although advances have been made in treating chronic pain
77
through repetitive transcranial magnetic stimulation and transcranial direct current stimulation,
78
further study is still required.30
M AN U
SC
RI PT
67
TE D
79 80
A novel non-invasive and portable electrical stimulation device, called the cranial nerve
82
noninvasive neuromodulator (CN-NINM or Portable Neuromodulation Stimulator - PoNS™),
83
induces processes of neuroplasticity by noninvasive stimulation of four cranial nerves:
84
trigeminal, CN-V, facial, CN-VII, glossopharyngeal, CN-IX, and hypoglossal, CN-XII, that
85
innervate the human tongue. CN-NINM uses comfortable, superficial, electro-cutaneous
86
stimulation of afferent cranial nerve branches to the tongue via small surface electrodes. The
87
tongue is a natural candidate for electrical stimulation due, in part, to a high density of sensory
88
receptors and a high concentration of electrolytes in saliva. The CN-NINM targets the
89
subcortical part of the brain, including the brainstem and cerebellum, an area largely unreachable
AC C
EP
81
4
ACCEPTED MANUSCRIPT Systematic Review CN-NINM 90
by other existing neurostimulation systems.31 Using “white noise” patterns of superficial
91
electrical stimulation on the dorsal surface of the tongue, the impulses stimulate all receptors to a
92
depth of 200-400 microns in the tongue epithelia.32
RI PT
93
Through a systematic review of the medical literature, this article will comprehensively
95
summarize the clinical research done using the cranial nerve noninvasive neuromodulator (CN-
96
NINM or PoNS™) in the treatment of central and peripheral nervous system disorders. The
97
primary objective is to assess the effect of cranial nerve modulation on objective measures of
98
function and evaluate its potential as a rehabilitation tool for these disorders.
M AN U
SC
94
99 100
METHODS
101
A literature search of PubMed®, MEDLINE® and the Cochrane Database from 1966 to March
103
2013 was conducted using the MESH search terms cranial nerve modulator, non-invasive
104
neuromodulation, cranial nerve stimulator, tongue stimulator and tongue modulator and Portable
105
Neuromodulation Stimulator (PoNS™). Other terms also searched included central nervous
106
system disorders, neurological disorders, neurodegenerative disorders, brain injury and
107
expansions of these terms to match synonyms, subterms or derivatives (Table 1). These terms
108
were searched in all fields of publication (e.g. title, abstract, and key word). The search was
109
limited to the English language articles and to “human” studies. Studies were included if they
110
recruited adult patients with peripheral and central nervous system disorders and were treated
111
with non-invasive cranial nerve modulation and were assessed with objective measures of
112
function. Articles that did not use noninvasive cranial nerve modulation or did not treat nervous
AC C
EP
TE D
102
5
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
system disorders primarily, were excluded. In addition, the bibliographies and reference lists of
114
all articles and all review articles were evaluated for other potentially relevant articles.
115
Acceptable study designs included experimental studies, observational studies, and case control
116
studies with clearly defined outcomes measured. Review articles, opinion papers and editorials
117
were excluded. The abstracts of the publications were screened for relevance and in case of
118
uncertainty regarding the inclusion, the entire text of the article was read. Studies were defined
119
as prospective or retrospective according to whether the method of data collection and the end
120
points were defined before patient enrollment began. The full texts of the articles were then
121
pooled and reviewed by two different authors to identify articles that met inclusion criteria. Once
122
the relevant articles were selected they were reviewed using a standard review form. The
123
elements extracted from each study that were included in the standard review form were: 1)
124
study type; 2) study objective; 3) study setting; 4) inclusion/exclusion criteria; 5) outcome
125
measures; 6) sample size; 7) results; 8) conclusion; and 9) limitations. The review forms allowed
126
the reviewers to objectively assess the content of each article in a consistent fashion. A
127
composite evidentiary table was then constructed.
SC
M AN U
TE D
EP
128
RI PT
113
Studies were also assessed for quality. Two authors applied the Newcastle-Ottawa Quality
130
Assessment Scale for Nonrandomized Studies in Meta-Analysis to evaluate the overall quality of
131
evidence for the studies identified.33,34 This Scale was developed to assess the quality of
132
nonrandomized studies with its design, content and ease of use directed to the task of
133
incorporating the quality assessments in the interpretation of meta-analytic results. It uses a 'star
134
system' in three broad categories: 1) the selection of the study groups; 2) the comparability of the
135
groups; and 3) the ascertainment of either the exposure or outcome. A study can be awarded a
AC C
129
6
ACCEPTED MANUSCRIPT Systematic Review CN-NINM 136
maximum of 4 stars in the Selection category; 2 stars in the Comparability category; and a
137
maximum of 3 stars in the Exposure/Outcome category.
RI PT
138
All data were entered into RevMan software (version 5.2 for Windows; Copenhagen, Denmark).
140
The results of studies were pooled using fixed effects models after consideration of heterogeneity
141
among trials. For continuous outcomes we calculated the mean differences reported from
142
individual studies, and we pooled statistics as weighted mean differences with associated 95%
143
confidence intervals. The mean, SD, and sample size for each outcome (DGI, ABC, DHI) pre
144
and post intervention were used to calculate the weighted mean differences. We used χ2 and I2
145
statistics to test for heterogeneity (25%, 50%, and 75% representing low, moderate, and high
146
heterogeneity) as well as funnel plots. Funnel plots are a visual tool for assessing publication bias
147
(the association of publication probability with the statistical significance of study results) in
148
meta-analysis. They are essentially scatterplots of the treatment effects of each study (horizontal
149
axis) versus the study size (vertical axis). Publication bias is a concern if the funnel plots are
150
asymmetrical.
153
RESULTS
AC C
152
EP
151
TE D
M AN U
SC
139
154
The initial search identified fifty potential publications, twelve of which involved the use of
155
cranial modulation systems and had their full texts reviewed. Of these, five studies met all
156
selection criteria and were included in the systematic review. However, only two studies had
157
usable outcome measures that could be combined into a meta-analysis. Details of the study
158
selection process are outlined in Figure 1. Each of these five studies was critically reviewed by 7
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
two investigators using a standard review form. All of the included studies were prospective and
160
were published either in or after the year 2006. All of the studies evaluated outcome pre and post
161
use of the CN-NINM device, however, there were no randomized controlled trials. The mean
162
sample size of the selected studies was 20 (SD ± 11.4) [range 9-40]. All of the studies were
163
conducted in adults. Three of the studies utilized control populations. A description of the studies
164
is provided in the evidentiary table (Table 2).35-39
SC
RI PT
159
165
The majority of the studies consisted of individuals with a chronic balance dysfunction due to
167
central or peripheral etiologies that affected balance, posture, and gait. Of these, the most
168
common peripheral conditions were gentamycin ototoxicity, and endolymphatic hydrops; the
169
most common central causes were Mal de Debarquement, cerebellar stroke, traumatic brain
170
injury and idiopathic causes. Other, less common, conditions included acoustic neuroma,
171
meningitis/encephalitis, and Parkinson’s Disease. Excluded conditions were pregnancy, mental
172
health disorders, myasthenia gravis, Charcot-Marie-Tooth disease, post-polio syndrome, Guillan-
173
Barré, fibromyalgia, chronic fatigue syndrome, herniated disc, and osteoarthritis of the spine.
174
Almost all of the subjects in the studies had previously completed standard vestibular
175
rehabilitation therapy and were at least one year from their acute condition. Most had well-
176
adapted compensatory strategies for coping with the debilitating effects of their condition.
177
Nonetheless, all subjects still had difficulty with standing, walking, or moving.
AC C
EP
TE D
M AN U
166
178
179
The assessments of study quality was performed using the Newcastle-Ottawa Quality
180
Assessment Scale for Nonrandomized Studies in Meta-Analysis and are shown in Table 3. Study 8
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
quality was variable among the articles with improvements in quality in later experiments. For
182
example, the earlier studies 38,39 did not have control groups compared to subsequent ones.35-37
183
Some of the factors that limited study quality included selection bias from referral populations,
184
lack of blinding of the investigators/assessors to the device and outcome, and the lack of
185
reporting of losses to follow-up or treatment compliance.
RI PT
181
SC
186
The outcomes used in the selected studies were established measures of balance and function
188
including the Dynamic Gait Index (DGI – a 24-point scale with higher scores representing better
189
ambulation), Dizziness Handicap Inventory (DHI – a 100-point self-assessment scale with higher
190
scores indicating greater perceived handicap from dizziness), Activities-specific Balance
191
Confidence Scale (ABC – a 100-point self-assessment scale with higher scores representing
192
greater perceived confidence in performing activities), and Sensory Organization Test (SOT -
193
uses dynamic posturography to measure postural sway in seconds with higher scores indicating
194
better balance). Additionally, functional MRI and other self-perception of impairment measures
195
were also employed. Each study used a variation of these outcome measures before and after
196
CN-NINM stimulation. Only two of the studies (Danilov 2007 and Wildenberg 2011)35,38 used
197
the same outcome tools with adequate data for the studies to be combined in a meta-analysis.
TE D
EP
AC C
198
M AN U
187
199
Both of the studies (n=40) included in the meta-analysis reported improvement in the three
200
outcome measures of Dynamic Gait Index (DGI), Activities-specific Balance Confidence Scale
201
(ABC), and Dizziness Handicap Inventory (DHI). Funnel plots were constructed to assess
202
systematic heterogeneity (Figure 2). For all three indices, heterogeneity was non-significant 9
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
(I2=0% and χ2<25%) (Figures 2 and 3) and funnel plots did not show any asymmetry so a fixed
204
effect model was used to assess effect size. Effect sizes were very similar between both studies.
205
The pooled results demonstrated a significant improvement in DGI post-intervention compared
206
to before the intervention with the CN-NINM device, with a mean difference 3.45 (95%CI 1.75,
207
5.15) and overall effect of Z=3.98; p<0.001 (Figure 3a). Similarly, the pooled results for the
208
ABC scale also indicated a significant improvement post-intervention, with a mean difference
209
16.65 (95%CI 7.65, 25.47) and overall effect of Z=3.64; p<0.001 (Figure 3b). For the DHI,
210
pooled results pre and post CN-NINM intervention showed a mean difference of -26.07 (95%CI
211
-35.78, -16.35) with an overall effect of Z=5.26; p<0.001 (Figure 3c).
M AN U
SC
RI PT
203
212
Although only two studies were pooled, the other three studies also demonstrated improvements
214
following treatment with CN-NINM. The study by Danilov et al. in 200639 was more of a
215
descriptive study that attained various positive observational results, such as greater inter-limb
216
coordination, endurance, and gait components that the subjects were not previously capable of.
217
The SOT results showed an average improvement of 50%. Retained benefits were also reported
218
in a subsequent article38 without continual usage of the CN-NINM device. A positive correlation
219
was established between the amount of time using the CN-NINM device and the length of time
220
subjects retained benefits. However, no correlation was found between usage time and change in
221
scores on the composite SOT, DHI, ABC or DGI. Among the lasting benefits recorded was
222
improvement in daily activities requiring balance and gait including but not limited to getting
223
dressed and traversing uneven terrain. No subjects reported adverse or negative side effects in
224
any of the studies reviewed.
AC C
EP
TE D
213
10
ACCEPTED MANUSCRIPT Systematic Review CN-NINM 225
Functional MRI was performed in three of the studies.35-37 In the 2010 study by Wildenberg et
227
al.37 a region within the pons exhibited sustained neuromodulation due to electrical tongue
228
stimulation by the CN-NINM on fMRI, but was unable to precisely identify the neuronal
229
structure involved. In a subsequent study,36 high-resolution imaging of the brainstem and
230
cerebellum was used to precisely localize sustained subcortical neuromodulation induced by CN-
231
NINM. The anterior cingulate cortex showed decreased activity after stimulation, while a region
232
within the pons had increased post-stimulation activity. These observations suggested that the
233
pontine region was the trigeminal nucleus and that tongue stimulation interfaced with the
234
balance-processing network within the pons. However, there were also improvements in
235
symptoms that appeared unrelated to the anatomical targets of the stimulation, suggesting that
236
CN-NINM induced wide-ranging neuroplastic changes that helped integrate diverse sensory
237
input.37,38 The investigators hypothesized that changes in cortical activity occurred through the
238
way information was processed from subcortical structures, through the thalamus, to the cortex.
239
This is consistent with a PET study indicating that the stimulation signal propagates through the
240
thalamus to multiple cortical regions.40
EP
TE D
M AN U
SC
RI PT
226
242
243
AC C
241
DISCUSSION
244
This systematic review summarizes the available evidence for the use of CN-NINM in treating
245
nervous system disorders. We identified five articles that prospectively examined the CN-NINM
246
device, two of which, had adequately described outcomes that could be combined for a meta11
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
analysis. These quasi-experimental observational studies used referral populations of nervous
248
system disorders with balance problems of different etiologies in order to examine the efficacy of
249
the device. Patients were treated with the CN-NINM device multiple times daily for a variable
250
length of time. The outcome measure of these studies consisted of an assortment of standardized
251
tests of balance and function, along with qualitative measures, that universally showed
252
improvements among test subjects.
SC
RI PT
247
253
There are a number of potential advantages to using CN-NINM for rehabilitation of nervous
255
system disorders. Firstly, the improvements seen in patients with chronic balance dysfunction
256
from various causes demonstrates the versatility of the CN-NINM device for different peripheral
257
and central conditions. Secondly, CN-NINM is non-invasive. Clinical application of implanted
258
nerve stimulators are frequently limited by their invasiveness and by their associated risks from
259
surgery, infections and tissue damage. Thirdly, fMRI has shown that the neuroplastic changes
260
induced after application of the CN-NINM device impact several regions of the brain. These
261
functional changes are retained after just 5 days of training with the device. Fourthly, the device
262
is portable and can be used at home, without constant medical supervision, given proper training.
263
This provides the opportunity to increase the availability and frequency of neurorehabilitation.
264
Most importantly, none of the studies have reported adverse effects from use of the device.
266
TE D
EP
AC C
265
M AN U
254
STUDY LIMITATIONS
267
12
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
Although these results are encouraging, there are important limitations that should be discussed.
269
Publication bias is a possibility given that studies reporting negative findings are less likely to be
270
published and more likely to be excluded from systematic reviews. However, we were able to
271
contact two of the authors to verify the existence of any unpublished data showing negative
272
results. There was no such data, so it is unlikely that publication bias influenced the conclusions.
273
In order to minimize selection bias three independent reviewers screened the literature for
274
possible study inclusion and all abstracts and primary manuscripts were assessed using
275
standardized eligibility criteria.
M AN U
SC
RI PT
268
276
Generally, the quality of a meta-analysis is a reflection of the included studies. There were small
278
sample sizes included in each of the studies reviewed and only two studies had adequately
279
described data with outcomes that could be pooled. This represents a total of only forty patients
280
in the meta-analysis. Accordingly, the small sample size limited the ability to conduct subgroup
281
analyses to identify patient characteristics that are most likely to benefit from this treatment.
282
Although the data was limited, one of the studies examined results in subgroups of age and
283
etiology (peripheral versus central) and showed little to no differences in treatment effects
284
between the groups. Another important limitation of the current literature on CN-NINM includes
285
the lack of appropriate placebo controlled groups. We can only speculate that placebo controlled
286
groups were not used due to the cost and the challenge of finding patients willing to enter a
287
lengthy study with the possibility of not being treated. Furthermore, the lack of blinding at
288
selection, during assessments, and of the patients themselves is another important factor that
289
could lead to bias in these studies. These will be important factors to consider in the design of
290
future studies. Other methodological issues consist of referral populations with very broad
AC C
EP
TE D
277
13
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
inclusion criteria and protocols that utilize varying lengths of training time with the CN-NINM
292
device. In certain experiments, the CN-NINM was used for a variable number of days among
293
patients depending on stamina and ability to complete follow-ups. Additionally, the length and
294
type of the follow-up assessments also varied within the same cohort. It will be important to
295
reduce this variability in treatment regimens in subsequent studies in order to establish optimal
296
treatment schedules and to maximize retained benefits.
SC
RI PT
291
297
Going forward, it will be necessary to maintain a high level of methodological rigor in assessing
299
this device, including the application randomized clinical controlled trials with larger sample
300
sizes that are powered to detect clinically relevant differences. It will be critical to include
301
standard diagnostic criteria, comprehensive measurement of outcomes conducted by blinded
302
examiners, high retention rates, measurement of intervention fidelity, the effect of adjuvant
303
therapies, and objective validated measures of change to monitor progress and improvement.
304
Despite these limitations, the studies reviewed show promising benefits to the use of the CN-
305
NINM device on balance, function and sensory-motor coordination. The ability to non-
306
invasively produce sustained modulation of neural activity holds promise as a new route for
307
therapy in individuals with neurological dysfunction.
309
TE D
EP
AC C
308
M AN U
298
CONCLUSION
310
14
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
This systematic review describes a number of studies using the CN-NINM device that appear to
312
improve balance and sensory-motor coordination in a myriad of peripheral and central
313
conditions. However, there are several limitations to the studies evaluated in this systematic
314
review that include small sample sizes, lack of randomization and control groups, absence of
315
blinding, use of referral populations, and variability in treatment schedules and follow-up rates.
316
Given these limitations, we recognize that a meta-analysis pooling only two studies is sub-
317
optimal and the results must be interpreted cautiously. Future studies of the CN-NINM device
318
will require very rigorously conducted randomized controlled studies with blinding and well-
319
defined outcome measures. In summary, further study of the CN-NINM device is required before
320
widespread clinical application.
321
TE D
322
M AN U
SC
RI PT
311
323
AC C
EP
324
15
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
REFERENCES
326
1. Young JA, Tolentino M. Neuroplasticity and its applications for rehabilitation. Am J Ther
327
2011;18:70-80.
328
2. Boudreau SA, Farina D, Falla D. The role of motor learning and neuroplasticity in designing
329
rehabilitation approaches for musculoskeletal pain disorders. Man Ther 2011;15:410-4.
330
3. De FP, Fellus J, Polito MZ, Thompson JW, Moser RS, DeLuca J. The new neuroscience
331
frontier: promoting neuroplasticity and brain repair in traumatic brain injury. Clin Neuropsychol
332
2009;23:1391-9.
333
4. Levin HS. Neuroplasticity and brain imaging research: implications for rehabilitation. Arch
334
Phys Med Rehabil 2006;87:S1.
335
5. Kuo MF, Grosch J, Fregni F, Paulus W, Nitsche MA. Focusing effect of acetylcholine on
336
neuroplasticity in the human motor cortex. J Neurosci 2007;27:14442-7.
337
6. Engineer ND, Moller AR, Kilgard MP. Directing neural plasticity to understand and treat
338
tinnitus. Hear Res 2012;295:58-66.
339
7. Kaufman M, Corner MA, Ziv NE. Long-term relationships between cholinergic tone,
340
synchronous bursting and synaptic remodeling. PLoS One 2012;7:e40980.
341
8. Roosevelt RW, Smith DC, Clough RW, Jensen RA, Browning RA. Increased extracellular
342
concentrations of norepinephrine in cortex and hippocampus following vagus nerve stimulation
343
in the rat. Brain Res 2006;1119:124-32.
344
9. Andrade DC, Borges I, Bravo GL, Bolognini N, Fregni F. Therapeutic time window of
345
noninvasive brain stimulation for pain treatment: inhibition of maladaptive plasticity with early
346
intervention. Expert Rev Med Devices 2013;10:339-52.
AC C
EP
TE D
M AN U
SC
RI PT
325
16
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
10. Villamar MF, Santos Portilla A, Fregni F, Zafonte R. Noninvasive brain stimulation to
348
modulate neuroplasticity in traumatic brain injury. Neuromodulation 2012;15:326-38.
349
11. Fregni F, Pascual-Leone A. Technology insight: noninvasive brain stimulation in neurology-
350
perspectives on the therapeutic potential of rTMS and tDCS. Nat Clin Pract Neurol 2007;3:383-
351
93.
352
12. Marangell LB, Martinez M, Jurdi RA, Zboyan H. Neurostimulation therapies in depression: a
353
review of new modalities. Acta Psychiatr Scand 2007;116:174-81.
354
13. Adeyemo BO, Simis M, Macea DD, Fregni F. Systematic review of parameters of
355
stimulation, clinical trial design characteristics, and motor outcomes in non-invasive brain
356
stimulation in stroke. Front Psychiatry 2012;3:88.
357
14. Goodman RR, Kim B, McClelland S, 3rd, Senatus PB, Winfield LM, Pullman SL, Yu Q,
358
Ford B, McKhann GM, 2nd. Operative techniques and morbidity with subthalamic nucleus deep
359
brain stimulation in 100 consecutive patients with advanced Parkinson's disease. J Neurol
360
Neurosurg Psychiatry 2006;77:12-7.
361
15. Fasano A, Daniele A, Albanese A. Treatment of motor and non-motor features of Parkinson's
362
disease with deep brain stimulation. Lancet Neurol 2012;11:429-42.
363
16. Benabid AL, Chabardes S, Mitrofanis J, Pollak P. Deep brain stimulation of the subthalamic
364
nucleus for the treatment of Parkinson's disease. Lancet Neurol 2009;8:67-81.
365
17. Handforth A, DeGiorgio CM, Schachter SC, Uthman BM, Naritoku DK, Tecoma ES, Henry
366
TR, Collins SD, Vaughn BV, Gilmartin RC, Labar DR, Morris GL, 3rd, Salinsky MC, Osorio I,
367
Ristanovic RK, Labiner DM, Jones JC, Murphy JV, Ney GC, Wheless JW. Vagus nerve
368
stimulation therapy for partial-onset seizures: a randomized active-control trial. Neurology
369
1998;51:48-55.
AC C
EP
TE D
M AN U
SC
RI PT
347
17
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
18. George MS, Rush AJ, Sackeim HA, Marangell LB. Vagus nerve stimulation (VNS): utility in
371
neuropsychiatric disorders. Int J Neuropsychopharmacol 2003;6:73-83.
372
19. Marangell LB, Rush AJ, George MS, Sackeim HA, Johnson CR, Husain MM, Nahas Z,
373
Lisanby SH. Vagus nerve stimulation (VNS) for major depressive episodes: one year outcomes.
374
Biol Psychiatry 2002;51:280-7.
375
20. Marangell LB, Suppes T, Zboyan HA, Prashad SJ, Fischer G, Snow D, Sureddi S, Allen JC.
376
A 1-year pilot study of vagus nerve stimulation in treatment-resistant rapid-cycling bipolar
377
disorder. J Clin Psychiatry 2008;69:183-9.
378
21. Nahas Z, Marangell LB, Husain MM, Rush AJ, Sackeim HA, Lisanby SH, Martinez JM,
379
George MS. Two-year outcome of vagus nerve stimulation (VNS) for treatment of major
380
depressive episodes. J Clin Psychiatry 2005;66:1097-104.
381
22. Park MC, Goldman MA, Carpenter LL, Price LH, Friehs GM. Vagus nerve stimulation for
382
depression: rationale, anatomical and physiological basis of efficacy and future prospects. Acta
383
Neurochir Suppl 2007;97:407-16.
384
23. Edwardson MA, Lucas TH, Carey JR, Fetz EE. New modalities of brain stimulation for
385
stroke rehabilitation. Exp Brain Res 2013;224:335-58.
386
24. Bandt SK, Anderson D, Biller J. Deep brain stimulation as an effective treatment option for
387
post-midbrain infarction-related tremor as it presents with Benedikt syndrome. J Neurosurg
388
2008;109:635-9.
389
25. Fitzgerald PB, Daskalakis ZJ. The use of repetitive transcranial magnetic stimulation and
390
vagal nerve stimulation in the treatment of depression. Curr Opin Psychiatry 2008;21:25-9.
391
26. Silberstein SD, Dodick DW, Saper J, Huh B, Slavin KV, Sharan A, Reed K, Narouze S,
392
Mogilner A, Goldstein J, Trentman T, Vaisma J, Ordia J, Weber P, Deer T, Levy R, Diaz RL,
AC C
EP
TE D
M AN U
SC
RI PT
370
18
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
Washburn SN, Mekhail N. Safety and efficacy of peripheral nerve stimulation of the occipital
394
nerves for the management of chronic migraine: results from a randomized, multicenter, double-
395
blinded, controlled study. Cephalalgia 2012;32:1165-79.
396
27. Mwenge GB, Rombaux P, Dury M, Lengele B, Rodenstein D. Targeted hypoglossal
397
neurostimulation for obstructive sleep apnoea: a 1-year pilot study. Eur Respir J 2012;41:360-7.
398
28. Milby AH, Halpern CH, Baltuch GH. Vagus nerve stimulation in the treatment of refractory
399
epilepsy. Neurotherapeutics 2009;6:228-37.
400
29. Hoppe C. Vagus nerve stimulation: Urgent need for the critical reappraisal of clinical
401
effectiveness. Seizure 2013;22:83-4.
402
30. Fregni F, Freedman S, Pascual-Leone A. Recent advances in the treatment of chronic pain
403
with non-invasive brain stimulation techniques. Lancet Neurol 2007;6:188-91.
404
31. Tyler ME, Braun JG, Danilov YP. Spatial mapping of electrotactile sensation threshold and
405
intensity range on the human tongue: initial results. Conf Proc IEEE Eng Med Biol Soc
406
2009;2009:559-62.
407
32. Tyler M, Danilov Y, Bach YRP. Closing an open-loop control system: vestibular substitution
408
through the tongue. J Integr Neurosci 2003;2:159-64.
409
33. Wells GA, Shea B, O'Connell D, Peterson J, Welch V, Losos M, Tugwell P. The Newcastle-
410
Ottawa Scale (NOS) for Assessing the Quality of Nonrandomized Studies in Meta-Analysis
411
http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp.
412
34. Oremus M, Oremus C, Hall GB, McKinnon MC. Inter-rater and test-retest reliability of
413
quality assessments by novice student raters using the Jadad and Newcastle-Ottawa Scales. BMJ
414
Open;2.
AC C
EP
TE D
M AN U
SC
RI PT
393
19
ACCEPTED MANUSCRIPT Systematic Review CN-NINM
35. Wildenberg JC, Tyler ME, Danilov YP, Kaczmarek KA, Meyerand ME. Electrical tongue
416
stimulation normalizes activity within the motion-sensitive brain network in balance-impaired
417
subjects as revealed by group independent component analysis. Brain Connect 2011;1:255-65.
418
36. Wildenberg JC, Tyler ME, Danilov YP, Kaczmarek KA, Meyerand ME. High-resolution
419
fMRI detects neuromodulation of individual brainstem nuclei by electrical tongue stimulation in
420
balance-impaired individuals. Neuroimage 2011;56:2129-37.
421
37. Wildenberg JC, Tyler ME, Danilov YP, Kaczmarek KA, Meyerand ME. Sustained cortical
422
and subcortical neuromodulation induced by electrical tongue stimulation. Brain Imaging Behav
423
2010;4:199-211.
424
38. Danilov YP, Tyler ME, Skinner KL, Hogle RA, Bach-y-Rita P. Efficacy of electrotactile
425
vestibular substitution in patients with peripheral and central vestibular loss. J Vestib Res
426
2007;17:119-30.
427
39. Danilov YP, Tyler ME, Skinner KL, Bach-y-Rita P. Efficacy of electrotactile vestibular
428
substitution in patients with bilateral vestibular and central balance loss. Conf Proc IEEE Eng
429
Med Biol Soc 2006;Suppl:6605-9.
430
40. Ptito M, Moesgaard SM, Gjedde A, Kupers R. Cross-modal plasticity revealed by
431
electrotactile stimulation of the tongue in the congenitally blind. Brain 2005;128:606-14.
433
SC
M AN U
TE D
EP
AC C
432
RI PT
415
20
ACCEPTED MANUSCRIPT
Table Legend Table 1. Composition of search phrases for literature search strategy
RI PT
Table 2: Evidentiary table summarizing clinical studies on CN-NINM device
M AN U
SC
Table 3. Newcastle - Ottawa Quality Assessment Scale for Observational Studies The Newcastle-Ottawa Quality Assessment Scale was developed to assess the quality on nonrandomized studies. Assessments are intended to be incorporated into meta-analytic results. It is a star system based on three domains: 1) Selection of StudyGroups; 2) Comparability of Groups; and 3) Ascertainment of exposure/outcome.
Figure Legend Figure 1. Study Selection Process
TE D
Figure 2a, b, c. Funnel plots of comparisons between outcome measures pre and post intervention
AC C
EP
Figure 2a. Funnel plot of comparison: Dynamic Gait Index (DGI) Pre and Post Intervention with CN-NINIM. Figure 2b. Funnel plot of comparison: Activities-specific Balance Confidence Scale (ABC) Pre and Post Intervention with CN-NINIM. Figure 2c. Funnel plot of comparison: Dizziness Handicap Inventory (DHI) Pre and Post Intervention with CN-NINIM
Figure 3a, b, c. Forest plots of comparisons between outcome measures pre and post intervention with CN-NINIM Figure 3a. Forest plot of comparison: Dynamic Gait Index (DGI) Pre and Post Intervention with CN-NINIM. Figure 3b. Forest plot of comparison: Activities-specific Balance Confidence Scale (ABC) Pre and Post Intervention with CN-NINIM. Figure 3c. Forest plot of comparison: Dizziness Handicap Inventory (DHI) Pre and Post Intervention with CN-NINIM
ACCEPTED MANUSCRIPT
Table 1. Composition of search phrases for literature search strategy Terms included in nervous system disorders search phrase
Terms included in the treatment search phrase
Cranial nerve modulation
Central nervous system (CNS) disease/disorder Peripheral nervous system (PNS) disease/disorder Ataxia/Balance disorder Neuropathy Brain Injury Traumatic Brain Injury/TBI Neurological disease/disorder Brain/Cerebral/Cerebellar
Rehabilitation
M AN U
Therapy Recovery
SC
Treatment
TE D EP
Noninvasive Nerve stimulator Nerve modulator PoNS® BrainPort® Tongue simulator
AC C
Cranial nerve modulator
RI PT
Terms included in the cranial nerve modulation search phrase
ACCEPTED MANUSCRIPT
Table 2: Evidentiary table summarizing clinical studies on CN-NINM device
Electrical Tongue Stimulation Normalizes Activity Within the Motion-Sensitive Brain Network in BalanceImpaired Subjects as Revealed by Group Independent Component Analysis
Study #2 2011 Wilden berg et al.34
High-resolution fMRI detects neuromodulation of individual brainstem nuclei by electrical tongue stimulation in balance-impaired individuals
Study #3 2010 Wilden berg et al.35
Sustained cortical and subcortical neuromodulation induced by electrical tongue stimulation
Study #4 2007 Danilo v et al.36
Efficacy of electrotactile vestibular substitution in patients with peripheral and central vestibular loss
Study #5 2006 Danilo v et al.37
Efficacy of electrotactile vestibular substitution in patients with bilateral vestibular and central balance loss
Sample Size
Outcome Measures
Intervention
Results
Prospective Quasi-experimental With controls
12 subjects with symptoms of chronic balance disorder and 9 normal controls
-Measured Day 0 and Day 5 (Controls only assessed on Day 0) -Dynamic gait index (DGI) -Self-perception of impairment -Activities-specific Balance Composite scale -Dizziness Handicap Index (DHI) -fMRI (3T)
-CN-NINM was delivered only to the balance-impaired subjects over 9 sessions (two on days 1–4 and one on day 5) -During a stimulation session, subjects received continuous stimulation for 20 min while standing as still as possible with their eyes closed
Pre-intervention mean DGI score=19.0±4.3, mean DHI score=58.8±23.8 (lower score indicates less impairment), and an ABC score=57.9±24.4. Post-intervention mean DGI score=22.9±1.4 (p<0.005), mean DHI score=35.7±24.0 (p<0.0005), and an ABC score=75.3±21.2 (p<0.005). Global improvement (p< 0.005).
Prospective Quasi-experimental With controls
9 subjects with symptoms of chronic balance disorder and 9 normal controls
-Sensory Organization Testing (SOT) (a composite score that combines performance on six sensory conditions and measures overall postural stability [<70 abnormal]) -fMRI
-CN-NINM was delivered only to the balance-impaired subjects over 19 sessions (two on days 1–9 and one on day 10). Days 5 and 6 were separated by two rest days (no stimulation) -During a stimulation session, subjects received continuous stimulation for 20 min while standing as still as possible with their eyes closed
Pre-intervention Balance-impaired Subject #3 did not receive an SOT score pre-intervention due to significant stability impairment. 6/8 had SOT scores<70 Post-intervention Subject #3 received an SOT=62. 2/8 still had SOT <70. Mean improvement in SOT scores was 15.75±SEM 5.59 (p=0.026).
Prospective Quasi-experimental With controls
12 subjects with symptoms of chronic balance disorder and 9 normal controls
-Postural sway (MATLAB) (Only data from the anterior-posterior sway was used for analysis) -fMRI
-CN-NINM was delivered only to the balance-impaired subjects over 9 sessions (two on days 1–4 and one on day 5) -During a stimulation session, subjects received continuous stimulation for 20 min while standing as still as possible with their eyes closed
Pre-intervention The effect of optic flow on postural sway was greater in balance subjects than in controls. Post-intervention There was no difference in sway amplitude between balance subjects postintervention and controls. Balance subjects swayed more in response to the optic flow stimulus preintervention than post-intervention (p≤0.005).
Prospective Quasi-experimental
28 subjects with symptoms of chronic balance disorder
-Computerized Dynamic Posturography (CDP) -Sensory Organization Test (SOT) -Dynamic Gait Index (DGI) - Activities-specific Balance Confidence Scale (ABC) -Dizziness Handicap Inventory (DHI)
-CN-NINM was delivered to each balance-impaired subject twice daily for 34.5 days -Each stimulation session was 1-1.5 hours -During a stimulation session, subjects received a succession of shorter trial periods (1-5 minutes) and one 20 minute trial period -Subjects were to stand as still as possible while receiving stimulation
Prospective Quasi-experimental
40 subjects with symptoms of chronic balance disorder
-Sensory Organization Test (SOT) -Dynamic Gait Index (DGI) - Activities-specific Balance Confidence Scale (ABC) -Dizziness Handicap Inventory (DHI)
-CN-NINM was delivered to each balance-impaired subject over 9 sessions -Each stimulation session was 1.5-2 hours - During a stimulation session, subjects received continuous stimulation for 20 min while standing as still as possible with their eyes closed
Pre-intervention -mean DGI score=18.4±5.3 -mean DHI score=57.3±19.6 (lower score indicates less impairment -ABC score=61.7±20.4 -mean SOT score=48.5±18.3 Post-intervention -mean DGI score=21.5±3.1 (p<0.001) -mean DHI score=30.2±23.3 (p<0.001) (lower score indicates less impairment -ABC score=78.0±18.5 (p<0.001) -mean SOT score=65.0±17.1 (p<0.001) Pre-intervention -Unavailable Post-intervention -Average composite SOT improvement: 49.1% -All subjects ‘generally’ improved in functional transfer testing (DGI, ABC, and DHI), with the exception of four subjects exhibiting no change in the DGI
RI PT
Study #1 2011 Wilden berg et al.33
Design
SC
Title
M AN U
Study
AC C
EP
TE D
*FOLLOW UP ANALYSIS OF STUDY #1
ACCEPTED MANUSCRIPT
Table 3. Newcastle - Ottawa Quality Assessment Scale for Observational Studies
Outcome/Exposure (3 Stars)
**
*
*
**
*
*
**
*
*
*
-
*
-
RI PT
Comparability (2 Stars)
SC
Study #1 2011 Wildenberg et al.33 Study #2 2011 Wildenberg et al.34 Study #3 2010 Wildenberg et al.35 Study #4 2007 Danilov et al.36 Study #5 2006 Danilov et al.37
Selection (4 Stars)
M AN U
Study
* *
AC C
EP
GA Wells, B Shea, D O'Connell, J Peterson, V Welch, M Losos, P Tugwell. http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp
TE D
The Newcastle-Ottawa Quality Assessment Scale was developed to assess the quality on non-randomized studies. Assessments are intended to be incorporated into meta-analytic results. It is a star system based on three domains: 1) Selection of Study Groups; 2) Comparability of Groups; and 3) Ascertainment of exposure/outcome.
ACCEPTED MANUSCRIPT Figure 1. Study Selection Process 0 articles identified by review of Cochrane Database
50 articles identified by PUBMED/MEDLINE search
SC
M AN U
9 titles selected for further review after screening titles/abstracts
RI PT
41 articles did not meet inclusion based on title and abstract screening
AC C
EP
TE D
2 additional articles identified by review of selected bibliographies
1 article identified by contacting authors
12 articles had full text reviewed
7articles excluded - 4 did not fit the primary search criteria - 3 did not have any functional outcomes measured
5 primary studies included in this systematic review
2 studies with usable outcome data for metaanalysis
ACCEPTED MANUSCRIPT
Figure 2a, b, c. Funnel plots of comparisons between outcome measures pre and post intervention with CN-NINM
M AN U
SC
RI PT
2a. Funnel plot of comparison: Dynamic Gait Index (DGI) Pre and Post Intervention with CN-NINIM
AC C
EP
TE D
2b. Funnel plot of comparison: Activities-specific Balance Confidence Scale (ABC) Pre and Post Intervention with CN-NINIM
2c. Funnel plot of comparison: Dizziness Handicap Inventory (DHI) Pre and Post Intervention with CN-NINIM
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT Figure 3a, b, c. Forest plots of comparisons between outcome measures pre and post intervention with CN-NINIM
M AN U
SC
RI PT
3a. Forest plot of comparison: Dynamic Gait Index (DGI) Pre and Post Intervention with CN-NINIM
EP
TE D
3b. Forest plot of comparison: Activities-specific Balance Confidence Scale (ABC) Pre and Post Intervention with CN-NINIM
AC C
3c. Forest plot of comparison: Dizziness Handicap Inventory (DHI) Pre and Post Intervention with CN-NINIM