First randomised sham controlled trial of vagal nerve stimulation for heart failure treatment fails

1663

NECTAR-HF, the first randomised, sham-controlled trial of vagal nerve stimulation for the management of heart failure has failed to show significant difference in cardiac remodelling between vagal nerve stimulation and a sham procedure-meaning that the study has not met its pre-specified six-month primary efficacy endpoint. Results were presented at a Hot Line session of the European Society of Cardiology congress (ESC; 30 August – 3 September, Barcelona, Spain). At the same congress, results of ANTHEM-HF, another small feasibility study exploring vagal nerve stimulation in heart failure showed positive outcomes.

Principal investigator and presenter of NECTAR-HF, Faiez Zannad (l’Institut Lorrain du Coeur et des Vaisseaux Louis Mathieu, Vandoeuvre-lès-Nancy, France) said that the trial’s failure to show a cardiac benefit of vagal nerve stimulation was “unexpected” given the results from pre-clinical studies and the initial open-label pilot study (De Ferrari et al, Eur Heart J 2011;32:847-855) that have indicated that vagal nerve stimulation may be a beneficial treatment for the management of heart failure.


The aim of NECTAR-HF (Neurocardiac therapy for heart failure) was to evaluate whether right-sided vagal nerve stimulation was safe and might attenuate cardiac remodelling in patients with systolic heart failure using a randomised, sham-controlled designed study. Zannad commented that patients were enrolled if they had New York Heart Association (NYHA) class II-III, ejection fraction ≤35%, had left ventricular ejection diastolic diameter of ≥5.5cm, and had been on optimal treatment for at least 30 days prior to enrolment.


In the study, 95 patients who had been implanted with the vagal nerve stimulation device (Precision, Boston Scientific) were randomised to be “on” therapy (63) or to be “off” therapy (32) at 24 centres across Western Europe. Zannad explained that patients had the device implanted-by neurosurgeons or vascular surgeons-in their neck, near the right vagus nerve, and connected to a pulse generator implanted under the skin of the chest. “This therapy has been used for treatment of epilepsy so there is a big experience when it comes to safety by neurosurgeons,” he noted.

He also said that, due to lost patient data, the results presented at this time were for 59 of the on therapy patients and 28 of the off therapy patients (a matched cohort). He added that they represented “only the first six months of the randomised part of the protocol” and that “after six months, every patient was switched on.” The primary endpoint was reduction in left ventricular end-systolic diameter (LVESD).

Results

Regarding the results, Zannad said that there were no significant differences in the mean LVESD between baseline and six months follow-up point in either group (change in LVESD from baseline to six months was −0.04±0.25cm in the therapy group compared with −0.08±0.32cm in the control group (p=0.60)). Additional echocardiographic parameters of left ventricular end diastolic dimension, left ventricular end systolic volume, left ventricular end diastolic volume, left ventricular ejection fraction, peak V02, and N-terminal pro-hormone brain natriuretic peptide failed to show superiority compared to the control group.

He also reported that the safety results were “acceptable” for both the on therapy patients and the off therapy patients [complete results of primary safety will be assessed at 18 months] and that there were some significant improvements in quality of life seen with vagal nerve stimulation. However, he said, these findings “should be interpreted with caution given the imperfect patient level blinding.” He explained that although patients were blinded to their group assignment, meaning they were not told if they were receiving active treatment, it is possible that those in the active treatment group were able to feel the sensation of stimulation. If this was the case, patients could have been vulnerable to a placebo effect, knowing that they were receiving active therapy.


He concluded: “Although robust pre-clinical data showed the benefits of vagal nerve stimulation, NECTAR-HF, the first vagal nerve stimulation randomised sham controlled trial, failed to demonstrate a successful clinical translation of vagal nerve stimulation therapy to the primary endpoint of cardiac remodelling.” He added that “inclusion of an appropriate control group is crucial, and a randomised study like NECTAR-HF should be the benchmark for future studies of novel device therapies for the treatment of heart failure.”


John Camm (St George’s University of London, London, UK), discussant of the trial, commented: “NECTAR-HF was a very well framed hypothesis with an innovative sham-control design but unfortunately the blinding proved inadequate and was well documented by the investigators. This lack of adequate blinding may have led to improvements in events that were subjective such as a NYHA classification, quality of life and perhaps to events that were controlled by physicians informed by patients such as hospitalisation or within trial therapeutic decisions.”

He also said that with a “relatively small trial it is not surprising that there were some small differences between the groups.”

The negative results of NECTAR-HF, said Camm, raises many questions: “Firstly we must ask, were the right patients selected for this trial? Do we need to have some kind of assessment of whether the patients may respond to the stimulation and whether the ventricle is capable of responding? Was the trial design optimum? Was six months enough? Was the sample size which was originally 96 and then reduced to 87 with paired data at the conclusion of the trial sufficient? Is the discrepancy between ventricular remodeling results and the clinical events and the heart rate variability and quality of life relevant, does it tell us anything about this system of management? Was the non-R wave linked stimulation without putative afferent block optimum for changing sympatho-vagal balance? Where the stimulation parameters (duty cycle, frequency, amplitude, etc) optimum? Is stimulation induced bradycardia (missing with this form of stimulation) an important component of the therapy (therapeutic or as a marker of effect)? And, did the excess of post-randomisation clinical events in the control group (15.6% vs 11.1%) distort the remodelling findings because of missing data or treatment changes?


Camm mentioned two other ongoing trials that are also investigating vagal nerve stimulation for heart failure treatment: INOVATE-HF (CardioFit, Biocontrol Medical) and ANTHEM-HF (Cyberonics IPG: Model 103). He said that these trials are not using the same stimulation parameters. The comparative groups, primary endpoints and duration of follow-up are different.


NECTAR-HF is ongoing; the six-month results have been published in the European Heart Journal.

ANTHEM-HF vagal nerve stimulation study shows significant improvement in cardiac function of heart failure patients

Also, at a Hot Line session of the ESC congress, Inder Anand (University of Minnesota Medical School, Minneapolis, USA), presented preliminary results from the ANTHEM-HF study. Results have shown significant improvements of cardiac function and symptoms in patients with chronic heart failure treated with the Cyberonics Vagus Nerve Stimulation Therapy system regardless of whether the device is implanted on the left or right vagus nerve.


The ANTHEM-HF (Autonomic neural regulation therapy to enhance myocardial function in heart failure) study was published simultaneously in the Journal of Cardiac Failure.


Anand told delegates: “Left-sided vagal nerve stimulation, which could be combined with devices, has not been evaluated in heart failure patients, and the effects of left and right vagal nerve stimulation have not been directly compared.” Therefore, ANTHEM-HF was designed to evaluate the Cyberonics Vagus Nerve Stimulation system with left or right vagal nerve stimulation for the treatment of chronic symptomatic heart failure.


The prospective, multicentre study included 60 heart failure patients (aged approximately 51 years) with left ventricular ejection fraction ≤40%, NYHA class II/III, who were receiving optimal and stable pharmacological therapy.

Anand commented that the Cyberonics Vagus Nerve Stimulation system is currently used for epilepsy treatment and that over 100,000 left-sided implants have been performed.

In the ANTHEM-HF study, Anand said, patients were randomised to receive the device implanted on either the left (n=31) or right (n=29) side and stimulation to the vagus nerve was titrated over a 10-week period to determine the best-tolerated intensity. Following titration, vagal nerve stimulation was then delivered for six months at an amplitude of 2.0 (±0.6) mA, and a constant frequency of 10Hz.

Anand said that the results of the study showed significant improvement (mean 4.5%) from baseline in left ventricular ejection fraction (LVEF) among all patients, with no statistically significant differences between left- and right-sided vagal nerve stimulation. Improvement in left ventricular end systolic volume (LVESV) was not statistically significant.

He also noted that there was also a mean improvement of 56 metres in the 6-Minute Walk Test, but this improvement was significantly less with left- compared to right-sided vagal nerve stimulation (mean 34 vs. 77 meters). Minnesota Living with Heart Failure Questionnaire scores also improved by a mean of 18 points, with no significant difference between left- and right-sided vagal nerve stimulation.

There was a similar rate of device-related adverse events in both groups, including transient mild dysphonia (voice alteration), cough, and oropharyngeal pain, which resolved during the study. “These are consistent with device-related non-serious adverse events previously reported in epilepsy patients treated with vagal nerve stimulation,” noted Anand.


One patient with a history of carotid atherosclerosis suffered an embolic stroke during device implantation and died three days later.”It is possible that manipulation of the common carotid artery in the neck during dissection of the vagus nerve caused plaque disruption or dislodged a thrombus,” said Anand. “Avoidance of this procedure in patients with severe obstructive carotid disease is likely to minimise such occurrences.”

ANTHEM-HF is the first clinical study to compare the feasibility and tolerance of left- and right-sided ART, and compare safety and efficacy measures, he concluded. “This preliminary assessment shows promising results which need to be confirmed in a larger, controlled trial,” he said.