Study design and overview

The DAN-VNS study is composed of four work packages (WP) designed to explore the effects of tVNS on DAN-induced GI symptoms in people with diabetes (figure 2).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Overview of the diabetic autonomic neuropathy (DAN)-vagus nerve stimulation (VNS) study. GI, gastrointestinal.

Work package I

In WP-I, the effect of short-term, high-intensity tVNS treatment will be investigated. Participants will be allocated to either active or sham treatment and will be instructed to self-administer the treatment four times a day for 7 days.

Work package II

Following a wash-out period of minimum 2 weeks, participants will continue in WP-II, in which the effect of long-term, medium-intensity treatment with tVNS will be investigated. Participants will follow the same treatment as allocated in WP-1. In WP-II, self-administration of the treatment will be performed twice a day for 8 weeks.

Follow-up

A follow-up visit will be conducted 7 days after end of WP-II including blood samples and ECG, which will be reviewed by a medical doctor in order to ensure participant safety after cessation of the intervention. Drop-outs will likewise be encouraged to participate in a follow-up visit after trial discontinuation.

Work package III

In WP-III, the acute endocrine and metabolic response to short-term tVNS will be investigated in a randomised, explorative, cross-over study. Enriched enrolment will include 15 responders from WP-II, defined as those having the greatest decrease in GI symptoms in response to active tVNS. Participants will be randomised to the order of active or sham treatment conducted at two study days separated by a wash-out period of minimum 2 weeks. During each of the two study days, a semisolid test meal will be ingested, and tVNS will be administered by study personnel five times (at baseline and every hour thereafter) during a 4 hour session.

Work package IV

In order to provide a comparable baseline dataset, 40 age and gender-matched healthy controls will be recruited at the site in Aalborg, Denmark for a cross-sectional, descriptive study.

Study participants

Participants will be included at three different locations in Denmark: at Steno Diabetes Center North Jutland (Aalborg), at Steno Diabetes Center Copenhagen and at Steno Diabetes Center Aarhus. In total, 120 participants will be included for WP-I and WP-II with approximately 40 at each study site. In contrast, WP-III will be conducted in Copenhagen, and WP-IV in Aalborg.

Recruitment and inclusion

Potentially eligible participants will be identified at each study site after approval of the treating physician. The healthy controls will be recruited locally in Aalborg from an established volunteer database or advertisements. Eligible individuals will be provided with both oral and written information regarding the study. Before consenting to participation, subjects are encouraged to consider their decision for 24 hours. Furthermore, consent can be withdrawn at any point during the study period with no consequences for future treatment. Subjects confirm their willingness to participate by signing the informed consent form after which they are allowed to enter the trial. Screening and inclusion of candidates will be performed by a medical doctor.

Inclusion criteria include either a verified diagnosis of type 1 diabetes for a minimum of 1 year and stable hyperglycaemic medication ensuring that participants as a minimum have received the given treatment (long acting and fast acting insulin or insulin pump with dosing adjustments according to regimens) for at least 3 months prior to study entrance, or a verified diagnosis of type 2 diabetes for a minimum of 1 year, for which the patient receives no or stable (at least 3 months prior to study entrance) hyperglycaemic medication. Additionally, GI symptoms will be evaluated using the Gastroparesis Cardinal Symptom Index (GCSI) (see online supplemental material 1) and the Gastrointestinal Symptom Rating Scale (GSRS) (see online supplemental material 2). Inclusion in the study require a combined weighted GCSI and GSRS score of minimum 2.3. The justification of the cut-off value was based on the calculated weighted mean in several cohorts of healthy subjects plus one (GCSI) and two (GSRS) SDs.21–26 Furthermore, to confirm the diagnosis of DAN, participants must have either (1) minimum one abnormal cardiac autonomic reflex test (heart rate variability (HRV) in rest, in response to expiration:inspiration, lying to standing position and Valsalva manoeuvre) assessed with the validated VAGUS device,27 28 or (2) decreased electrochemical resistance assessed in both hands and feet with the SUDOSCAN device, using a cut-off value <50 µS at hand and <70 µS in the feet,29 or (3) a autonomic symptom score >16 assessed by the validated Composite Autonomic Symptom Score (COMPASS-31) questionnaire29–31 (see online supplemental material 3). A complete list of inclusion and exclusion criteria can be found in table 1. Lists of screened and included subjects will be kept at a secure location in order to ensure confidentiality.

Participants can be discontinued from the trial based on medical judgement or if they are considered non-compliant. Drop-outs will be mirror randomised by new participants in order to ensure sufficient power of the study.

Intervention

The study intervention consists of tVNS, which will be applied by a battery-powered handheld tVNS device (GammaCore, ElectroCore LLC, Basking Ridge, New Jersey, USA). The GammaCore is registered as a class IIa medical device, and has been CE marketed (CE 571753) for the following indications in adults: primary headache (migraines, cluster headache, and hemicranias continua) and medication overuse headache.32 The GammaCore device provides electrical stimulation of the cervical region of the vagal nerve by producing a proprietary low‐voltage electrical signal comprising a 5‐kHz sine wave burst lasting for 1 millisecond (five sine waves, each lasting 200 microseconds). Bursts are repeated once every 40 ms (25 Hz), generating a 24 V peak voltage and 60 mA peak output current. On stimulation, the device is positioned at the cervical course of the vagal nerve, situated anterior to the sternocleidomastoid muscle and lateral to the carotid artery. Stimulation is provided through two steel contact electrodes, which, prior to use, are covered with conductive gel (Sigma gel, Parker Laboratories, New Jersey, USA). Based on the manufacturer’s recommendations, each stimulation in the DAN-VNS study will consist of two consecutive bilateral tVNS doses, which corresponds to 2×120 s. Intensity of the stimulation can be adjusted by the user in the range of 1–40 (arbitrary units) via the digital user interface. The stimulation has been optimised, in regard to intensity (24 V peak voltage and 60 mA peak output current), pulse duration (1 ms), and frequency (5 kHz), to predominantly target afferent Aβ-fibres. Hereby, interference with cardiovascular innervating efferent C-fibres is limited and potential adverse effects such as bradycardia and bronchoconstriction are limited.20

Participants will be instructed to use the highest tolerable intensity setting in order to obtain maximum vagal stimulation. Placebo intervention will be applied by a sham GammaCore device (ElectroCore LLC, Basking Ridge, New Jersey, USA), which is identical to the active device in regard to appearance and application. In contrast to the electrical stimulation provided by the active device, the sham device provides a stimulation sound in order to mimic the active treatment. However, to eliminate the possibility of participants being able to distinguish between the active and sham stimulation, a parallel study design, in which all subjects remain in their allocated treatment group throughout the entire study, was chosen for WP-I and WP-II.

During the instruction session, correct placement of the device will be marked on the cervical region with a permanent marker. Participants are encouraged to refresh these marks at regular intervals in order to ensure tVNS/sham compliance. Furthermore, before the GammaCore is handed out, a thorough instruction of the use of the device and of compliance and daily doses will be given to ensure that the participants acknowledge the importance of receiving sufficient self-administered stimulation. In addition, this is emphasised in the handed-out Danish version of the user manual, which encourages stimulation within certain time windows, for example, morning between 06:00 and 07:00.

tVNS will be applied as an add-on treatment, in the sense that any habitual medication of participants will be continued without any changes. Habitual medication must be stable prior to inclusion in the study. During the study period, no new medication must by introduced to participants.

In WP-I and II, tVNS/sham treatment will be self-administered as two bilateral doses conducted four or two times a day, respectively, with approximately six- or 12 hours intervals. Participants will note time and intensity level of each stimulation or lack of such in a provided diary. In WP-III, tVNS/sham treatment will be applied by study personnel.

Outcomes and experimental procedures

An overview of outcomes and methods of assessment is presented in table 2 and tables 3–5. Experimental procedures are described below.

Patient-reported outcomes

Upper GI symptoms will be evaluated by the Patient Assessment of Upper GI Symptom Severity Index, which is a validated tool developed for measurement of symptom severity in patients with gastroparesis and includes the GCSI covering three subgroups: nausea/vomiting, bloating and postprandial fullness33 (see online supplemental material 1). Predominantly lower GI symptoms will be evaluated by the GSRS questionnaire, which was developed and validated for patients with luminal GI symptoms, covering five subgroups: reflux, abdominal pain, indigestion, diarrhoea, constipation34 (see online supplemental material 2). Additionally, the COMPASS-3130 will be applied for quantitative self-assessment of autonomic dysfunction symptoms (see online supplemental material 3).

Cardiovascular and sudomotor function

Several aspects of cardiovascular function will be assessed in order to investigate the presence of cardiac autonomic neuropathy in participants. (1) Cardiac vagal tone will be measured during 5 min of rest using a non-invasive ECG monitor (eMotion Faros, Mega Electronics, Kuopio, Finland). (2) Cardiovascular reflexes will be evaluated in response to three standardised tests: (a) posture change (supine to standing), (b) deep breathing respiratory cycle and (c) the Valsalva manoeuvre using the Vagus device (Medicus Engineering, Aarhus, Denmark). Based on age-dependent cut-off values, the device provides a cardiovascular autonomic neuropathy (CAN) score ranging from 0 to 3, indicating the level of CAN as being either absent, borderline or definitive.35 (3) Five days continuous Holter monitoring will be applied for extraction of HRV measures (time and frequency domain) of study participants using the ePatch device (BioTelemetry Technology Aps, Hørsholm, Denmark). The HRV measurements will be performed in accordance with the standards defined by The Task Force of the European Society of Cardiology and The North American Society of pacing and Electrophysiology. Additionally, sudomotor function has been suggested as an early marker of small fibre neuropathy.36 The electrochemical reaction mediated by chloride ions after stimulation of sweat glands will be assessed in both hands and feet by a non-invasive device (SUDOSCAN, Impeto Medical, California, San Diego, USA).37

Resting brain electrical activity

Electroencephalography (EEG) will be recorded in resting state for 5 min during which the subject is instructed to keep the eyes closed. The recording will be conducted with 40 electrodes by a direct current amplifier (NuAmp, Neuroscan, El Paso, USA). As previously reported by this group, resting EEG will be used to model connectivity between brain centres as well as to identify dominating centres of brain activity.38 The data will be preprocessed offline. First, notch filtered from 49 to 51 Hz, followed by a bandpass filter from 1 to 70 Hz. Afterwards, time intervals with clear artefacts in the majority of channels due to heavy muscle activity, and so on, will be removed along with interpolation of channels with abnormal traces or high amplitudes. The signal will lastly be re-referenced to the average reference before further analysis. The spectral indencies of the data will be analysed using a continuous wavelet analysis using a Morlet wavelet function with a bandwidth parameter of 10 Hz and a wavelet centre frequency of 1 Hz. Source localisation will be performed by the standardised low-resolution brain electromagnetic tomography.

Quantitative sensory testing

Sensory function will be assessed by quantitative sensory testing. First, pain tolerance threshold to bone and muscle pressure will be performed on the tibia bone and quadriceps muscle. A handheld pressure algometer (Type 2, Somedic Production AB, Sweden), applied at a rate of 30 kPa/s, will be used. Second, a cold pressor test will be applied. Participants will be instructed to immerse the left hand in circulating ice-chilled water (2℃) for 120 s or until pain becomes intolerable. During the immersion, pain intensity will be rated regularly on a visual analogue scale. Furthermore, immediately after the hand is withdrawn from the ice water, a pain test stimulus (quadriceps muscle pressure) will be applied, allowing for quantification of the conditioned pain modulation capacity, which provides insights to the function of descending inhibitory pain pathways.39

Brain imaging and GI tract function

Structural and functional assessments of brain and GI tract will be explored by 3 T and 1.5 magnetic resonance imaging (Signa HDxt, General Electrics, Milwaukee, Wisconsin, USA). Brain networks during rest and concentrations of brain metabolites will be investigated by functional MRI and magnetic resonance spectroscopy, respectively. Stomach volume, adaptive capacity and motility, small bowel motility and colonic volume will be assessed at regular time points after intake of a 400 mL standardised liquid meal (1.5 kcal per mL).

Glycaemic control

Continuous glucose monitoring will be applied in order to be able to measure real-time fluctuations in blood glucose and estimate ‘Time in Range’. Participants will be equipped with the FreeStyle Libre device (Abbot Diabetes Care Denmark, Copenhagen Ø, Denmark), which allows for continuous glucose monitoring via a portable system for 14 days before changing of needle and patch.40

Circulating inflammatory biomarkers

Serum collected from fasting blood samples will be aliquoted in appropriate volumes and stored in a biobank at −80℃. The complete set of serum samples will be analysed after study end for levels of inflammatory biomarkers such as TNF-α, IL-2, IL-4, IL-6 and IL-10 by multiplexing technology.

Gastrointestinal transit

A wireless motility capsule (SmartPill, Given Imaging, Yokneam Illit, Israel) will be used to evaluate the whole gut transit time as well as segmental for example, stomach, small bowel and large bowel transit times. After consumption, the indigestible capsule traverses through the GI tract while continuously measuring temperature, pH and pressure in the gut lumen.41 Such parameters can be used as proxies for the motor function of the enteric nervous system.

Postprandial levels of endocrine and metabolic biomarkers

Following an overnight fast, blood samples from participants in WP-III will be collected at regular intervals during the study session. In order to ensure similar glycogen storage status and macronutrient balance, participants will be instructed to avoid strenuous physical exercise and alcohol consumption 48 hours prior to WP-III study days. Levels of pancreatic polypeptides and gut hormones (gastrin, cholecystokinin, glucose-dependent insulinotropic polypeptide, glucagon-like peptide 1 and 2, oxyntomodulin, neurotensin, fibroblast growth factor 19) will be assessed after study completion.

Appetite and satiety sensation

Hunger, satiety, fullness and prospective food consumption will be reported by WP-III participants on a Visual Analogue Scale at regular time intervals.

Gallbladder volume

Gallbladder emptying will be investigated by volumetric measurements using ultrasonography (LOGIQ E9, GE Healthcare, Wauesha, Wisconsin, USA).

Resting energy expenditure

Following 30 min of rest, resting energy expenditure will be measured for 15 min by resting gas exchange monitoring (CCM Express, Indirect Calorimeter, Medgraphics, St. Poul, Minnesota, USA).

Ad libitum meal response

In order to assess voluntary food intake, participants in WP-III will be provided with an ad libitum pasta Bolognese meal and the ingested amount will be registered.

Randomisation and blinding

Included participants will be allocated a unique anonymised identification number kept throughout the entire study. Even though path-dependency may exist, we judged that a centrally generated block-randomisation list for WP-I and WP-II, generated by use of the website www.randomization.com was preferable in order to control seasonal and geographical bias. The randomisation list will be devised by the sponsor and kept in a sealed envelope until last participant last visit in WP-II. Furthermore, sealed envelopes, containing treatment allocation for each participant, will be kept in a secure location to allow for individual unblinding in case of medical emergencies. To ensure blinding of the study personnel, sponsor will be responsible for assigning active/sham tVNS medical devices to study sites. Nevertheless, blinding of participants and study personnel may be complicated by the fact that the sham device is unable to mimic the active stimulation in regard to associated fascial and cervical muscle contractions. Therefore, a parallel-group design was chosen for WP-I and II. Additionally, to optimise blinding of study personnel during WP-I and II, an unblinded and otherwise non-involved person at each study site will be responsible for the training of participants in the use of the tVNS GammaCore device. Furthermore, participants will be instructed to restrain from sharing any information regarding stimulation and associated physical sensations to other study personnel.

After completion of WP-I and WP-II, the unblinding of treatment allocation will be conducted by the sponsor. Likewise, sponsor will prepare a blinded data set divided in ‘Treatment A’ and ‘Treatment B’, in order to allow for blinding of data analysts. Additionally, the sponsor will identify responders of long-term active treatment of tVNS (WP-II) for identification of those subjects that are eligible for WP-III.

For participants entering WP-III, a new centrally generated randomisation code for the order of active/sham treatment in the cross-over design of WP-III will be assigned. Again, study personnel and participants will be blinded to the randomisation of treatment order, and even though blinding may be complicated, this aspect is regarded as acceptable given the fact that the primary outcome of WP-III consists of objective measures.

Harms and adverse events

Based on evidence of safety data regarding tVNS from previous studies,19 42 we do not expect the intervention to cause harm or substantial discomfort for participants. Transient side effects during stimulation including shortness of breath, hoarseness and twitching of facial muscles, are, however, listed as potential adverse reactions to GammaCore use (GammaCore User Manual). Furthermore, rare cases of syncope have been reported.43 Participants experiencing unacceptable levels of side effects will be instructed to reduce the level of stimulation intensity.

Adverse events, regardless of severity, will be collected regularly during trial conduct and recorded in the case report form. Investigators are responsible for immediate (within 24 hours) reporting of any serious adverse events to the sponsor. Furthermore, on a yearly basis, sponsor will notify relevant authorities of all registered serious adverse events. Causality of serious adverse events with the intervention will be assessed by a physician, and all serious adverse device effects will be reported to the manufacturer. Based on interim analyses of serious adverse device effects, sponsor will be able to terminate the trial if the safety of study participation is considered to be insufficient. During participation in the trial, subjects are covered by the patient insurance of the respective study site.

Sample size calculation

The study is powered to detect a minimal difference of 30% between the active and the sham tVNS group on the GCSI score 1 week (WP-I) or 2 months (WP-II) after study initiation. Based on a baseline score of 2.0±1.0 (mean±SD),33 60 subjects in each arm are required to provide a statistical power of 90% with the use of a two-sided significance level of 0.05. A power of 90% was chosen in order to account for secondary endpoints. A drop-out rate of up to 30% is estimated in the patient group. In order to achieve the required sample size, new subjects will be mirror randomised to account for participants withdrawing from the study.

Data collection and data management

Data will be collected by highly experienced research personnel, all of which have been trained in the principles of good clinical practice (GCP). Experimental data will be recorded directly and managed on a secure electronic platform (REDCap Electronic Data Capture Tool, hosted at Aalborg University Hospital, Aarhus University, and The Capital Region of Denmark). Filled-out questionnaires will be transferred to REDCap by two individual entry persons in order to limit the risk of errors in the process. Digitalised data are stored for up to 5 years after trial end. For each participant, a physical Case Report Form, containing all physical source material, will be kept in a secure and locked location for 5 years.

Serum samples will be stored at a research biobank for a maximum of 10 years. Samples will be anonymised and labelled with protocol number, subject randomisation ID and date of sample collection.

Monitoring

Monitoring of the trial will be conducted prior to, during and after study conduction by the GCP Unit (GCP Unit, Aalborg, Aarhus, and Copenhagen, Denmark) in order to ensure compliance with study protocol and national standards. The monitoring will include complete inspection of signed informed consent forms and adverse events logs. Likewise, investigators and study sites will allow for any quality assurance audits by relevant regulatory authorities.

Data analysis

The initial analysis of primary outcomes will be by intention to treat regardless of adherence to study protocol. Secondary outcomes will be analysed per protocol, in the sense that only subjects, who have completed the specific experimental setup, will be included. In WP-I and WP-II and WP-III, both primary and secondary outcomes will be evaluated as changes from baseline by appropriate statistical tests including paired samples t-test and equivalent non-parametric tests. Analyses of WP-III data will include a linear mixed effects model with fixed effects for period and treatment, and random effects for sequence and subject nested in sequence. To account for multiple testing in the analysis of secondary outcomes, the Bonferroni Correction or other applicable corrections will be applied. Missing data will be handled by applicable statistical methods including multiple imputation. Data collected from drop-outs will be included in the data analysis if appropriate (eg, data from completed WP-I).

After primary analysis of study specific outcomes, the complete anonymised data set will be available for additional analyses by the respective research groups at each study site.

Patient and public involvement

The original idea behind the DAN-VNS study was based on reports from patients with DAN-induced GI symptoms, who requested a simple, safe, and effective treatment option for their condition. The primary objective of the study was chosen to be patient-reported outcomes in order to put the subjective effect of tVNS in focus rather than objective measures, which may not correlate with patient evaluation and satisfaction with the treatment. After study completion, participants will receive information regarding own study arm allocation (active/sham) as well as overall findings in the form of a written report. Additionally, study results will be disseminated to relevant patient associations. No public involvement was included in the design phase of the study.

Ethics and dissemination

The study protocol (version 1.4) has been approved by the Danish Health and Medicines Authority (CIV-19-07-029105) and the North Denmark Region Committee on Health Research Ethics (N-20190020). Any amendments to the protocol will be implemented after approval of the mentioned authorities. Sponsor is responsible for informing all study sites of any such changes.

The conduction of the study will be in accordance with the protocol, the standard “Clinical investigation of medical devices for human subjects – Good Clinical Practise” (DS/EN ISO 14155:2012), designated Standard Operating Procedures, any applicable local regulatory requirements and laws, as well as the principles of the World Medical Association, Declaration of Helsinki amended by the 52nd General Assembly, Edinburgh, Scotland, October 2000, clarified by the General Assembly in Washington 2002, Tokyo 2004, Seoul 2008, and Fortaleza 2013.

Trial results, positive, negative or inconclusive, will be published in relevant international peer-reviewed journals and presented at national and international conferences. Authorship of published material will be decided by the steering committee. Additional dissemination of results will be initiated after approval of the principal investigator at each study site. Likewise, sponsor may grant access to the final data set on request.