Noninvasive low-level tragus stimulation attenuates inflammation and oxidative stress in acute heart failure

Purpose Acute decompensated heart failure is associated with inflammation, oxidative stress, and excess sympathetic drive. It is unknown if neuromodulation would improve inflammation and oxidative stress in acute heart failure. We, therefore, performed this proof-of-concept study to evaluate the effects of neuromodulation using noninvasive low-level Tragus stimulation on inflammation and oxidative stress in ADHF. Methods 19 patients with ejection fraction < 40% were randomized to neuromodulation- 4 hours twice daily (6 AM-10 AM and 6 PM-10 PM) (n = 8) or sham stimulation (n = 11) during hospital admission. All patients received standard-of-care treatment. Blood samples were collected at admission and discharge. Serum cytokines were assayed using standard immunosorbent techniques. Reactive oxygen species inducibility from cultured coronary endothelial cells exposed to patient sera was determined using dihydrodichlorofluorescein probe test (expressed as fluorescein units). Results Compared to sham stimulation, neuromodulation was associated with a significant reduction of circulating serum Interleukin-6 levels (−78% vs −9%; p = 0.012). Similarly, neuromodulation led to reduction of endothelial cell oxidative stress, in the neuromodulation group (1363 units to 978 units, p = 0.003) compared to sham stimulation (1146 units to 1083 units, p = 0.094). No significant difference in heart rate, blood pressure or renal function were noted between the two groups. Conclusion In this proof-of-concept pilot study, in acute systolic heart failure, neuromodulation was feasible and safe and was associated with a reduction in systemic inflammation and attenuation of cellular oxidative stress. Clinical trial: NCT02898181


Introduction
Acute decompensated heart failure (ADHF) is the most common cause of hospitalization in patients with heart failure. 1 Despite major advances in the management of chronic heart failure, limited treatment options are available for ADHF patients.Pharmacological management with diuretics, correction of precipitating factors, and supportive measures including vasodilators or inotropes remain the mainstay for ADHF. 2 Even with the most modernized treatment, ADHF carries poor outcome with recurrent hospitalizations and increased mortality. 2,3 onsidering the burden of ADHF and the limitations of currently available therapies, there is an unmet need for novel treatment strategies for ADHF.
A dysregulated autonomic nervous system, characterized by higher sympathetic and blunted parasympathetic response is the hallmark of heart failure. 4Although the enhanced sympathetic activity helps to maintain the end organ perfusion, in the long term it is associated with progressive and irreversible deterioration of cardiac function.Pro-in ammatory cytokines and oxidative stress contribute to disease progression by inducing contractile dysfunction, cell death, apoptosis, and brosis. 5,6 adrenergic receptor mediated stimulation of intracellular cyclic adenosine monophosphate and subsequent downstream kinase activation is one principal cause of maladaptive contractile, metabolic, and structural remodeling of the ventricle in heart failure. 7Hemodynamic stress, neurohormonal activation, and hypoperfusion also contribute to deranged cardiac metabolism, abnormal cardiac redox state, and cardiac and systemic in ammation.
The tragus of the external ear is innervated by the auricular branch of the vagus nerve.Low-level tragus stimulation (LLTS) is an emerging noninvasive neuromodulation strategy that is found to be associated with bene cial remodeling of the cardiac autonomic nervous system and restoration of sympathovagal balance. 8-10LLTS is found to ameliorate in ammation, remodeling in heart failure with preserved ejection fraction. 9,11,12 Hwever, the role of LLTS in ADHF has not been evaluated.The current study evaluated the effects of LLTS on biomarkers of in ammation and oxidative stress in patients with ADHF.
Here, we harnessed the bene cial effects of the neuro-immune axis to potentially reduce in ammation and oxidative stress in ADHF.

Study protocol
This single-center, prospective, double-blind, randomized study was conducted at the University of Oklahoma Medical Center.Patients who were hospitalized with ADHF with left ventricular ejection fraction (LVEF) < 40% were recruited.Patients with any of the following criteria were excluded from the study: acute coronary syndrome, complex congenital heart disease, recurrent vasovagal syncope, history of unilateral or bilateral vagotomy, sick sinus syndrome, cardiac conduction abnormalities including 2nd, or 3rd -degree atrioventricular block, bifascicular block or prolonged 1st degree AV block (PR > 300ms), pregnancy, prisoners, or end-stage renal disease on dialysis.The study was approved by the local Institutional Review Board.After informed consent, included patients were randomly assigned (using a computer generated model allocation on REDCAP) to either active LLTS or sham stimulation which was continued for the duration of their index hospital admission.The baseline clinical data including physical examination, routine laboratory investigation, electrocardiography, and echocardiography data were recorded.Patients were then instructed about the process of using the device and were asked to repeat the process themselves under direct supervision, to ensure the correct use of the device.This demonstration process was performed by a clinical coordinator who was not blinded to random allocation.All other investigators were blinded to the allocation arm.Blood samples were collected on the day of admission before the initiation of the treatment protocol and on the day of discharge.All procedures and protocols were approved by the Institutional Review Board at the University of Oklahoma.Informed consent was obtained from each participant.

Stimulation Protocol
Active LLTS was performed using electrodes attached to the tragus of the ear, which is innervated by the auricular branch of the vagus nerve. 13The PARASYM® (Parasym Ltd, London, United Kingdom) TENS Unit, was used for LLTS.The device was connected to a clip electrode that was attached to the external ear.In the active group, the ear clip electrode was attached to the tragus.In the sham stimulation group, the clip electrode was attached to the ear lobule where no vagal innervation is present.The unit was set at a pulse width of 200 µs, and a pulse frequency of 20 Hz.These parameters were chosen based on our previous published work in diastolic dysfunction and paroxysmal atrial brillation.The amplitude was titrated to the level of sensory threshold, typically in the range of 10-50 mA.The stimulation strength was gradually increased until the patient experienced mild discomfort and was then decreased by 1mA below the threshold of stimulation.LLTS was applied for 8 hours daily (4 hours twice daily, between 6 AM-10 AM and 6 PM -10 PM), and the device was used throughout the index hospital admission.

Serum in ammatory markers
Blood samples were centrifuged within 30 minutes of collection (4000g for 10 minutes), and serum/plasma was immediately stored in aliquots at -80°C.To determine systemic in ammation, serum levels of IL-6 were measured.As a proof of concept we also analyzed certain other markers in in ammation: CRP, TNFα, soluble intercellular adhesion molecule 1 (sICAM-1), and vascular cell adhesion molecule 1 (sVCAM-1) using commercially available enzyme linked immunosorbent assay kits (Thermo Fisher Scienti c™, USA)

Oxidative stress
Primary human coronary endothelial cells (HCAECs) based bioassay was used to evaluate the oxidative stress-inducing capacity of patient sera.Detector HCAECs were cultured in an Endothelial Basal Medium supplemented with 10% fetal calf serum until the time of serum treatment, as described. 14For treatment, fetal calf serum was replaced with patient sera (10%; for 48 h), following published protocols. 14Sera collected before and after treatment was distributed in a 96-well plate in quadruplicates, samples, and from each patient were measured in the same plate.Production of reactive oxygen species (ROS) by cultured HCAECs was measured by ow cytometry after treatment with sera of ADHF patients.Intracellular production of ROS induced by serum factors was measured in detector HCAECs using 2,7 dihydrodichloro uorescein (DCF) at 3 micromol concentration. 15

Statistical Analysis:
The primary outcome of the study was the reduction in in ammatory activity and ROS (cytosolic peroxide content) during hospital admission.We did not perform power calculations in this rst phase of feasibility trial.Categorical data were presented as percentages.For continuous variables, the distribution of data was assessed using the Shapiro-Wilk test and variance homogeneity with F-test.Normally distributed continuous data were presented as mean ± SD whereas median(IQR) values were used to present skewed data.The serum in ammatory and oxidative stress markers were compared between two groups (baseline LLTS vs baseline sham control or baseline vs predischarge) using the ANOVA test and within group changes were analyzed using Wilcoxon sum rank test at 5% signi cance.

Results
Study Population 20 patients were enrolled in the study, but 19 patients completed the study, and data from these 19 patients were used for analysis.Of these 19 patients, 8 patients (42%) were randomized to the intervention group, and 11 (52%) patients were randomized to the sham stimulation (control) group.The demographics and baseline characteristics of both groups are presented in Supplemental Table 1.No adverse events were reported by patients and device was well tolerated by all subjects.One patient left the hospital against medical advice and hence follow up data was not available.

LLTS oxidative in human coronary endothelial cells
Cultured HCAEC were exposed to the serum and subsequent measurement of the uorescent intensity of reactive oxygen species using sensitive uorescent probes

Effects of LLTS on in ammation:
In our pilot, proof-of-concept study in acute heart failure, neuromodulation technique using LLTS was associated with a reduction in serum IL-6 but also favorably altered other cytokines (CRP, TNF-α).The reduction of systemic in ammation by LLTS in ADHF is a novel nding of our study.In addition, neuromodulation using LLTS led to favorable modulation of human coronary endothelial cell stress.
7][18] The acute increase in cytokines may modulate the course of acute heart failure by negatively in uencing ventricular systolic and diastolic function, systemic and pulmonary uid homeostasis, and promoting prothrombotic and proapoptotic milieu.Serum cytokine levels are also reported to correlate with the degree of ventricular remodeling and adverse prognosis. 16, 170][21] IL-6, in acute myocarditis, also plays a crucial role in the transition from acute to chronic phase by promoting macrophage recruitment, and persistent elevation of IL-6 is known to predict the risk for recurrent hospitalization and non-recovery of ventricular function. 21 22The severity of myocardial damage and remodeling correlates with the degree of in ammation.In a canine model of ischemic heart failure, a greater reduction of in ammation with LLTS translated into more reverse remodeling and improvement of contractile function. 23chanisms underlying vagal-mediated in ammatory modulation have been described previously.In experimental models of systolic heart failure, restoration of sympathovagal balance by chronic vagal stimulation is associated with inhibition of cytokine release and improvement of ventricular function. 23,24 vagally mediated 'in ammatory re ex' is known to control the immune response and in ammation during pathogenic invasion as well as during tissue injury, the parasympathetic vagal system being the integral component of both afferent and efferent system in this communication. 25LLTS is thought to relay afferent information to the dorsal vagal complex and dorsal motor nucleus through the auricular branch of the vagus nerve. 26Modulation of the brainstem stem vagal centers is thought to modulate the interaction between the sympathetic and parasympathetic systems with a net increase in efferent vagal out ow to the heart and periphery, speci cally the splenic tissue.This anti-in ammatory re ex is partly mediated via α7 nicotinic acetylcholine receptor receptors (α7nAchR) on splenic macrophages thereby modulating the peripheral generation of in ammatory cytokines. 11,27 n summary, the vagomimetic and antiadrenergic action of LLTS, via the brain-cardiac-splenic axis, may explain the attenuation of in ammation in our study.LLTS-mediated reduction in systemic in ammation may suggest and uncover a novel, non-pharmacological intervention that may aid in treating a pathophysiological axis in ADHF.
Effect of LLTS on oxidative stress: This oxidation is also heavily dependent on Cytochrome c (residing on the inner membrane of the mitochondria) and redox active transition metals. 32In a previous study by Csiszar et.al., endothelial cells exposed to increased in ammatory stress (such as increased circulating cytokines-TNF-α) with subsequent increased oxidative stress and ROS generation, speci cally peroxide, led to higher DCF values. 30,33 ncreased circulating cytokines perhaps exposed the HCAEC's to an in ammatory milieu leading to increased ROS generation and subsequent increased DCF uorescence.To our knowledge, this is the rst study in ADHF to determine the extent of induced human coronary endothelial ROS quanti cation, using DCF based assays, speci c for cytosolic peroxide content.
An aberrant redox balance, characterized by enhanced pro-oxidant ROS generation and /or reduced activity of antioxidant mechanisms, plays a major role in the pathophysiology of heart failure. 6Increased ROS may further induce contractile, electrical, and structural remodeling in HF. 6 Serum ROS level correlates with prognosis and severity HF. 34 Reduction of oxidative stress is reported to improve in ammation, provide cytoprotectant effects, and ameliorate post-infarction remodeling. 35Adrenergic stimulation promotes ROS production by inducing a reductive redox as well as by stimulation of nicotinamide oxidases by β2-adrenergic activity. 36,37 ow-level vagal stimulation is reported to reduce ROS generation and improve antioxidant capacity in experimental models of myocardial infarction and heart failure. 38,39 oth antiadrenergic and cholinomimetic action of LLTS might also contribute to the in vivo inhibition of ROS production by augmentation of muscarinic activation(M2 receptors), anti-beta2 activity, downstream improvement of metabolic balance, as well as reduction of nicotinamide oxidase activity. 31,34,35 Ltations: Our study carries several limitations, a small sample size being the rst limitation.Second, the endpoints were focused on surrogate markers rather than clinical outcomes.However, this pilot study is hypothesis generating on the effects of LLTS on core pathogenic pathways in ADHF.Future prospective studies with large sample sizes are warranted to determine the effects of LLTS on clinical endpoints in ADHF.Third, although heart rate variability parameters are considered markers of cardiac autonomic activity, the effect of short-term LLTS on heart rate variability parameters has not been evaluated in our study.Our ongoing extended study (NCT02898181) of LLTS in acute heart failure (HFrEF and HFpEF) will further evaluate the effects on heart rate variability, in ammation, heart failure clinical biomarkers.In this ongoing study the duration of LLTS is now reduced to 2 hours/day.

Conclusions
LLTS during hospital admission is safe in the acute decompensated heart failure patients with reduced ejection fraction and is associated with the mitigation of systemic in ammatory activity.Endothelial level oxidative stress activity, speci cally cytosolic peroxide content, was also decreased with LLTS in ADHF, perhaps directly linked to attenuation in systemic in ammation.Further studies are underway to corroborate and extend these ndings.

Declarations Figures
Effect of LLTS on serum in ammatory markers: Serum in ammatory cytokine levels were assessed using high sensitivity ELISA.Changes from baseline to discharge values were expressed as percentage of the baseline value (discharge-baseline/baseline x 100).Changes were compared between LLTS and control.*p<0.05,Wilcoxon-test
with LLTS compared to sham.LLTS did not adversely affect heart rate, blood pressure, and renal function during hospitalization.This is the rst study to demonstrate favorable effects of LLTS on in ammation and oxidative stress in ADHF.
Discussionrst-in-human, proof-of-concept, pilot study demonstrated the feasibility and safety of LLTS in ADHF and showed improvement in systemic in ammation, as evident in IL6.Similarly, changes were noted in oxidative stress 31,30ary endothelium is an important source of ROS generation in HF and our study demonstrated an attenuation of endothelial oxidative stress with neuromodulation via LLTS.Reactive oxygen species are primarily comprised of superoxide (O2 − ), hydroxyl (OH − ) and peroxide(H 2 O 2 ).These (speci cally superoxide and hydroxyl) are principally generated in mitochondria and cytosol from electron leak through electron transport chain and oxidation of reduced nicotinamide adenine dinucleotide phosphate by its corresponding oxidase enzyme.28Superoxide,anextremelyunstablecompoundwithhalf-life in nano seconds, is subsequently metabolized to H 2 O 2 by superoxide dismutase.Due to rapid reduction by superoxide dismutase and very short half-life, superoxide is far more unstable compared to H 2 O 2 and thereby rendering it as a less effective direct oxidative biomarker to measure.28Cytosolicperoxide is primarily dependent on diffusion of mitochondrial H 2 O 2 into cytosol.29,30Dihydrodichlorouoresceindiacetate is widely used to evaluate cellular oxidative stress.DCF passes through plasma membrane and this non uorescent compound is ultimately oxidized to a uorescent compound dichloro uroscein (DCF), via a process that involves various ROS, but more speci c to H 2 O 2 .31