Acute Response to Unilateral Unipolar Electrical Carotid Sinus Stimulation in Patients with Resistant Arterial Hypertension

Abstract

Bilateral bipolar electrical carotid sinus stimulation acutely reduced muscle sympathetic nerve activity (MSNA) and blood pressure (BP) in patients with resistant arterial hypertension but is no longer available. The second-generation device utilizes a smaller unilateral unipolar disk electrode to reduce invasiveness while saving battery life. We hypothesized that the second-generation device acutely lowers BP and MSNA in treatment-resistant hypertensive patients.

Eighteen treatment-resistant hypertensive patients (9 women/9 men, 53±11 years, 33±5 kg/m²) on stable medications have been included in the study. We monitored finger and brachial BP, heart rate (HR), and MSNA.

Without stimulation, BP was 165±31/91±18 mmHg, HR was 75±17 bpm, and MSNA was 48±14 bursts/min. Acute stimulation with intensities producing side effects that were tolerable in the short term elicited interindividually variable changes in systolic BP (SBP: −16.9±15.0 mmHg, range: 0.0 to −40.8 mmHg, p=0.002), HR (−3.6±3.6 bpm, p=0.004), and MSNA (−2.0±5.8 bursts/min, p=0.375). Stimulation intensities had to be lowered in 12 patients to avoid side effects at the expense of efficacy (SBP: −6.3±7.0 mmHg, range: 2.8 to −14.5 mmHg, p=0.028; HR: −1.5±2.3 bpm, p=0.078; comparison against responses with side effects). Reductions in diastolic BP and MSNA (total activity) were correlated (r²=0.329, p=0.025).

In our patient cohort, unilateral unipolar electrical baroreflex stimulation acutely lowered BP. However, side effects may limit efficacy. The approach should be tested in a controlled comparative study.

Introduction

Devices reducing blood pressure (BP) through stimulation of afferent baroreflex nerves have been developed decades ago but failed mainly for technical reasons.^1–3 More recently, implantable devices electrically stimulating the carotid sinus have been developed and clinically tested in patients with severe treatment resistant hypertension. The first-generation device (Rheos™) applied bilateral localized field stimulation through bipolar electrodes in tripolar configuration placed around the carotid sinuses. In carefully conducted animal experiments^4,5 and an uncontrolled clinical trial on resistant hypertension,⁶ the treatment lowered BP. A mechanistic substudy showed that BP reduction was mediated through sympathetic inhibition and that normal baroreflex regulation remained intact with carotid sinus stimulation.⁷ In the controlled phase of a subsequent clinical trial, electrical carotid sinus stimulation lowered BP; however, the predefined endpoint acute efficacy defined as proportion of ≥10 mm Hg systolic BP responders was not significantly different between groups.⁸ The second-generation device (neo™), which is approved and clinically applied in Europe, utilizes a small unilateral unipolar disk electrode to decrease invasiveness and to improve battery life. Patients with resistant arterial hypertension implanted with the new device showed BP reductions in an uncontrolled clinical trial.⁹ A controlled clinical trial is currently ongoing in the US (NCT01679132). Electrical fields produced by disk-shaped unipolar electrodes differ markedly from electrical fields produced by bipolar circumferential electrodes,¹⁰ which together with modified stimulation settings could affect the efficacy in engaging carotid baroreceptors. The side-effect profile due to electrical stimulation of surrounding anatomical structures could also differ between unipolar and bipolar electrode designs. Therefore, we tested the hypothesis that unilateral unipolar carotid sinus stimulation elicits acute reductions in sympathetic vasoconstrictor tone and BP in patients with treatment resistant hypertension.

Methods

Patients

We included patients with treatment-resistant hypertension who had been implanted with the CVRx® Barostim neo™ System for baroreflex activation therapy. All patients included in our study had previously been evaluated and treated at specialized hypertension clinics in Germany. Patients with mental inability, drug or alcohol addiction, secondary forms of hypertension, and pregnant or breast-feeding women have been excluded. The ethics committee of Hannover Medical School approved our study and all patients gave written informed consent.

Electrical baroreflex stimulation

The second generation carotid sinus stimulator (neo™, CVRx, Inc., Minneapolis, MN) consists of a programmable pulse generator that is implanted subcutaneously near the collar bone, a unipolar disk electrode that is sutured unilaterally to the carotid sinus wall (Figure 1), and a lead connecting generator and electrode. The device delivers constant current as rectangular pulses. Impulse width, intensity, and frequency can be programmed transcutaneously.

Figure 1. Stimulation electrodes.

First- (Rheos™) and second-generation (neo™) carotid sinus stimulation electrodes. First-generation bipolar electrodes in tripolar configuration had to be placed around the carotid sinuses on both sides of the neck. The smaller second-generation unipolar electrode investigated in our study, is sutured directly to the carotid sinus wall, preferentially on the right side of the neck.

We subjected patients to alternating on/off protocols with different stimulation intensities that have been individually determined by stepwise adjustments of stimulation intensity. The timing of the protocol was guided by automated BP measurements whose interval was set at 2 minutes. Immediately after each second measurement we switched the stimulation to the opposite state. Thus, each on- or off-period, respectively, lasted for approximately 4 minutes. We did not inform the patients about the switching times.

Cardiovascular and sympathetic measurements

We conducted the experiments after an overnight fast in the morning hours while patients remained in the supine position. Electrocardiogram, finger BP (Finometer® MIDI, Finapres Medical Systems, the Netherlands), and thoracic impedance in order to monitor breathing movements (Niccomo™, medis GmbH, Ilmenau, Germany) were continuously recorded. Brachial oscillometric BP measurements (Dinamap® Pro 100, GE Healthcare, Waukesha, Wisconsin) were performed twice during stimulator-on and stimulator-off periods. Muscle sympathetic nerve activity (MSNA) was recorded from the right peroneal nerve (Nerve Traffic Analyzer 662C3, Biomedical Engineering Department, University of Iowa, Iowa City, IA, USA) as described previously.¹¹ Data were analog-to-digital converted and analyzed using a program written by one of the authors (AD). We determined the following MSNA parameters from the integrated nerve signal: Burst frequency, i.e. the number of MSNA bursts per minute (bursts/min), burst incidence, i.e. the number of bursts per 100 heart beats (bursts/100 heart beats), as well as total MSNA, i.e. the area under the bursts per minute as arbitrary units per minute (au/min). We assessed three parameters of spontaneous cardiac baroreflex sensitivity (BRS) for 5-min resting periods using cross-spectral analysis between SBP and RR-interval variabilities in the low-frequency band (BRS_lf)¹² and the sequence technique (BRS_us, BRS_ds).^13,14 After instrumentation, subjects rested for at least 20 minutes to achieve a stable baseline with the stimulator on or off. This implies that patients may have been studied after prolonged baroreflex activation. Then, we identified stimulation intensities eliciting side effects that were tolerable for the time of the experiment and intensities that could be applied chronically. Afterwards, we started the alternating on/off protocol. In 9 patients, we were able to obtain on/off-protocol data with both of these stimulation intensities.

Statistical Analysis

Individual data originating from the alternating on/off protocol are medians. Group data are expressed as means±sd. Responses to electrical carotid sinus stimulation were grouped on the basis of side effects (present/absent). We tested group responses to the stimulation using one-sample tests and response differences between the groups (t and Wilcoxon signed rank test considering data distribution). Additionally, for patients with response data during both conditions, i.e. with and without stimulation side effects, we compared their responses using the paired t test. The relationship between measurements was assessed by correlation analysis. A value for p<0.05 was considered significant.

Results

We studied 18 treatment-resistant hypertensive patients (9 women/9 men). Baseline characteristics are given in Table 1. In three patients, we could not find a stable microneurographic recording position in the peroneal nerve. Sympathetic outflow was elevated in our patients when compared with normotensive controls taking age, gender, and body mass index into account.^15–18

Table 1.

Patients’ baseline characteristics (n=18)

Parameter	mean ± sd
Age [years]	53.5 ± 10.6
BMI [kg/m2]	33.4 ± 5.2
SBP [mmHg]	163 ± 22
DBP [mmHg]	93 ± 15
HR [bpm]	74.8 ± 15.0
MSNA [bursts/min]	51.1 ± 16.4
MSNA [bursts/100 hb]	65.2 ± 13.4
MSNA [au]	2.99 ± 1.17
Medications	7.1 ± 1.4

BMI = Body mass index; SBP/DBP = Systolic/diastolic blood pressure; HR = Heart rate; MSNA = Muscle sympathetic nerve activity; Medications = Number of antihypertensive medication classes

Without electrical carotid sinus stimulation, BP was 165±31/91±18 mmHg, HR was 75±17 bpm, and MSNA was 48±14 bursts/min while patients were on antihypertensive medications (Table 1). Figure 2 shows original finger BP, heart rate (HR), and MSNA tracings during decremental stimulation intensities in a patient responding to carotid sinus stimulation. The relationship between stimulation intensity and occurrence of jaw pain in another responder is given in Figure 3. Individual changes in BP, HR, and MSNA obtained during the alternating on/off protocols are given in Table 2: Stimulation with intensities producing side effects that were tolerable in the short term changed systolic BP (SBP) −16.9±15.0 mmHg (p=0.002). However, the response showed large interindividual variability ranging between 0.0 and −40.8 mmHg (Figure 4, middle column). Pulse pressure (PP) changed −10.6±11.0 mmHg (p=0.007), HR changed −3.6±3.6 bpm (p=0.004), and MSNA changed −2.0±5.8 bursts/min (p=0.375). Twelve patients reported jaw or neck pain, globus or swallowing sensation, coughing, or voice problems. We had to decrease stimulation intensities in these patients to avoid side effects with chronic treatment. The alternating on/off protocol with reduced stimulation intensity (Figure 4, right column) changed SBP −6.4±7.0 mmHg (range: 2.8 to −14.5 mmHg, p=0.003), PP −4.2±8.4 mmHg (p=0.073), HR −1.5±2.3 bpm (p=0.023), and MSNA 0.0±3.5 bursts/min (p=0.701). Avoiding intolerable stimulation-related side effects reduced efficacy in terms of BP response (SBP: p=0.028/p=0.042, PP: p=0.285/p=0.098, HR: p=0.078/p=0.090, MSNA burst frequency: p=0.286/p=0.243; paired/unpaired comparison). For comparison, we also included data on acute BP responses from a previous study in patients implanted with the bilateral bipolar carotid sinus stimulator Rheos™ (Figure 4, left column). Reductions in diastolic BP and total MSNA during the on/off protocol were correlated (r²=0.329, p=0.025; Figure 5). To visualize intraindividual variabilities in the BP response to repeated acute electrical carotid sinus stimulation, we grouped patients according to their median acute BP response. In each response group, we plotted all acute responses during on/off testing (Figure 6). We did not observe breath holding or similar maneuvers that may have influenced the results. Furthermore, acute responses to electrical carotid sinus stimulation were similar in patients who had been chronically stimulated (ΔSBP: −17.8 mmHg with side effects, −5.3 mmHg without side effects) and in patients in whom the stimulator had been off before testing (ΔSBP: −15.6 mmHg with side effects, −8.9 mmHg without side effects). Furthermore, responses to electrical carotid sinus stimulation in terms of SBP or MSNA reduction were not correlated with prevailing SBP, MSNA, or HR with the stimulator off (correlations: ΔSBP × total MSNA: r²=0.024, p=0.579; ΔSBP × SBP: r²=0.072, p=0.282; ΔSBP × HR: r²=0.077, p=0.265; Δtotal MSNA × total MSNA: r²=0.001, p=0.923; Δtotal MSNA × SBP: r²=0.001, p=0.936; Δtotal MSNA × HR: r²=0.027, p=0.558). Likewise, response magnitude cannot be predicted on the basis of spontaneous cardiac baroreflex sensitivity parameters (correlations: ΔSBP × BRS_lf: r²=0.001, p=0.923; ΔSBP × BRS_us: r²=0.001, p=0.892; ΔSBP × BRS_ds: r²=0.021, p=0.570).

Figure 2. Original recording.

Finger blood pressure (FBP), heart rate (HR), and muscle sympathetic nerve activity (MSNA) in a patient responding to electrical carotid sinus stimulation. Shaded areas indicate stimulation intervals. Darkness level indicates stimulation intensity. Stimulator settings are given in the top. Each time the stimulator was on, BP, HR, and MSNA decreased in a stimulation-intensity related fashion. Please note that the baseline of the MSNA recording increased with stimulation intensity likely indicating stray currents. The lowermost trace represents the enlarged view of the framed portion of the MSNA recording. Electrical baroreflex stimulation was switched on at 0 seconds: Note the marked reduction in MSNA burst frequency and the effect of stray currents.

Figure 3. Relationship between stimulation intensity, systolic blood pressure (SBP), and side effects in one patient.

The stimulator settings are shown in the top. SBP decreased in a stimulation-intensity related fashion. Double and single diamonds indicate strong and moderate jaw pain, respectively. Due to the pain, stimulation intensity had to be reduced which entailed lessened efficacy. The equality sign "=" denotes unchanged parameter settings.

Table 2.

Responses to electrical baroreflex stimulation taking into account side effect (SE) occurrence (+ yes/− no)

	ΔSBP [mmHg]		ΔDBP [mmHg]		ΔHR [bpm]		ΔMSNA [bursts/min]		ΔMSNA [bursts/100 hb]		ΔMSNA [au]
Pat. ID	+SE	−SE	+SE	−SE	+SE	−SE	+SE	−SE	+SE	−SE	+SE	−SE
01	−26.0	−12.3	−10.0	−1.5	−4.4	−7.0	−0.4	−3.60	3.9	−1.0	−0.06	−0.35
02	−40.8		−12.8		−4.5		3.6		8.1		0.29
03		1.5		−10.0		1.5		2.00		−1.8		−1.04
04	−12.5	2.8		−2.5		−1.8
05	−12.5	−14.5	−7.0	−3.8	−1.6	−2.5	−1.0	−3.03	−1.4	−3.7	−0.05	−0.41
06		−9.3		−5.5		−4.0		0.00		2.6		−0.09
07	−6.0	2.3	5.0	1.3	−1.2	0.3	−6.0	5.50	−6.9	9.1	−0.23	0.03
08		−13.5		−5.0		−3.1		3.75		4.8		0.29
09	−37.0	−9.5	−5.0	1.0	−4.1	−0.3	−2.0	−1.00	5.7	0.3	0.78	0.17
10		−13.5		−1.0		−1.6		0.04		1.1		0.18
11		1.5		−2.8		−0.6		−5.01		−6.0		−0.04
12	−36.0	−12.5	−18.3	−5.0	−10.0	−2.7	−16.4	3.00	−20.0	9.5	−1.07	−0.08
13	−24.5	−13.0	−13.0	−5.0	−11.5	−3.0
14	0.0	0.0	−1.5	−1.0	0.0	2.5	4.0	5.00	6.7	7.1	0.21	0.18
15	−5.5		−5.0		−1.7		0.0		3.3		−0.05
16	−6.5		−2.9		−1.3		−2.0		−1.6		0.26
17	−6.3	−6.5	−4.3	−2.0	−1.1	−0.8	−0.4	−1.91	−0.5	−0.6	−0.06	0.04
18	−1.5	1.3	−0.5	−0.3	−1.7	−0.1						−0.35

Figure 4. Individual changes in systolic blood pressure (SBP) with electrical carotid sinus stimulation.

Leftmost data points represent data from a previous study⁷. Middle and right columns show present investigational data. Upper p values refer to group responses against null (one-sample tests). Lower p values refer to group differences, i.e. the comparison of stimulation effects with vs. without side effects. In 9 patients measurements were obtained under both conditions (dashed lines). In the remaining 9 patients, data were obtained with side effects (3 patients) or without side effects only (6 patients). Baroreflex stimulation was less effective without side effects irrespective of whether or not whole groups or only paired data were compared statistically.

Figure 5. Sympathetic activity and blood pressure.

Association between changes in muscle sympathetic nerve activity and blood pressure elicited by electrical carotid sinus stimulation. Dots represent all patients with successful nerve recordings (n=15).

Figure 6. Patients’ response variability with electrical carotid sinus stimulation.

Based on their individual median acute response, we grouped patients in three groups. Then, we generated boxplots from all acute responses within each group. The most prominent outliers (closed circles) represent measurements in the same patient with striking spontaneous BP surges. Apart from that the distributions imply that response variability is primarily caused by the underlying BP variability rather than unstable responsiveness to electrical carotid sinus stimulation.

With chronic electrical carotid sinus stimulation with intensities producing minimal or no side effects, office BP decreased (ΔSBP: −15.5±14.6 mmHg, p=0.009; ΔDBP: −9.1±12.8 mmHg, p=0.137; ΔPP: −6.4±10.7 mmHg, p=0.090; ΔHR: −5.6±12.9 mmHg, p=0.201; n=10).

Discussion

The main finding of our study is that electrical carotid sinus stimulation using a unilateral unipolar electrode produces interindividually variable acute reductions in BP, HR, and MSNA in patients with resistant arterial hypertension. Reductions in BP and MSNA were related to each other. The finding supports the idea that sympathetic inhibition contributes to BP lowering through electrical carotid sinus stimulation. HR reductions with electrical carotid sinus stimulation were limited even in some patients with a BP and MSNA response. Possible explanations include different baroreceptor stimulation thresholds for vasomotor and cardiac responses¹⁹ or selective stimulation of baroreflex afferents controlling different effectors.^20–22 Two thirds of our patients (12 of 18) experienced stimulation-related side effects such that stimulation intensity had to be reduced. In these patients, lowering of stimulation intensity to chronically tolerable levels was associated with marked reduction in the acute efficacy. Nevertheless, office BP improved with chronic treatment. Whether or not chronic reduction in office BP resulted from electrical carotid sinus stimulation cannot be ascertained in the absence of a control group. In fact, we previously showed that acute responses to electrical carotid sinus stimulation predict chronic ambulatory BP responses.⁷ Our findings are important for the clinical use of electrical carotid sinus stimulators, which are approved in the European Union, and for the design of future devices.

The pathophysiology of essential hypertension is heterogeneous. The same state-of-affairs is likely true for severely affected patients not sufficiently responding to standard antihypertensive combination therapy. In these patients with treatment resistant hypertension, sympathetic activation has been implicated in maintaining hypertensive BP levels.²³ We previously observed that the contribution of sympathetic activity to BP assessed through near complete pharmacological ganglionic blockade markedly differs between patients with arterial hypertension.²⁴ Surgical sympathectomy also elicited variable blood pressure reductions in hypertensive patients and some did not respond at all.²⁵ Therefore, among patients with resistant hypertension a certain proportion of true non-responders to a treatment reducing sympathetic activity can be expected.²⁶ However, in our patients the overall acute response to stimulation that was tolerable in the long term was surprisingly small and more than a third of patients did not respond at all. While we observed non-responders in our previous study applying bilateral bipolar electrical carotid sinus stimulation,⁷ reductions in BP and MSNA were more robust.

It is likely that a substantial part of the modest response in our patients is explained by insufficient carotid baroreceptor engagement. Both, electrode placement and electrode design could be causative. We studied patients who had been implanted by vascular surgeons with extensive experience in implanting such devices. Moreover, we adhered to all recommendations for implantation provided by the device manufacturer. Finally, we did not observe obvious differences between patients implanted in different centers. In any case, if successful implantation were more operator dependent than with the bipolar electrode, this would pose a limitation for more widespread clinical use. We suggest that the unipolar design of the electrode may not be ideal. The electrical current travels between electrode and pacemaker located in the upper chest such that baroreceptor stimulation may be limited by stray currents eliciting side effects through off-target electrical stimulation. We noticed a peculiarity as exemplified in Figure 2 that supports the idea: The lower lane shows integrated MSNA. Although there is evident suppression of sympathetic bursts with strong stimulation (leftmost, darkest gray area) the MSNA trace is offset to a higher level (see dashed line for comparison). The offset is smaller with less intense stimulation (see middle and right gray area). The observation suggests presence of stray currents reaching down to the popliteal space where they are picked up by the microneurography electrode pair.

The anatomy of the carotid sinus and baroreceptor distribution is variable and may not always be covered by the electrical field produced by a small disc-shaped electrode. Indeed, electrode design substantially affects the properties of the stimulating electric field.²⁷ The unipolar design allows for greater penetration of the electrical current into the media of the carotid artery where the baroreceptors are located.¹⁰ On the other hand, current spread to adjacent structures may precipitate side effects.28 Smaller electrode spacing may increase the threshold for unwanted side effects without compromising the desired response.²⁹ Development of multichannel electrodes may further improve stimulation selectivity.³⁰ Finally, engagement of carotid chemoreceptors, which are known to regulate sympathetic activity and BP,³¹ could differ between electrode designs. In fact, carotid body chemoreceptors directly suppress baroreflex responses.^32,33

Electrical carotid sinus stimulation can lower sympathetic activity and BP in patients with resistant arterial hypertension. However, turning this interesting approach into a treatment that can be introduced into routine clinical care is challenging. Implantation of the first-generation device was more invasive because the electrode was placed on both sides and more extensive dissection of the carotid artery was required. Furthermore, battery life was rather short. Despite these limitations, the treatment was shown to reduce BP in a controlled clinical trial.⁸ Implantation of the second generation device is less invasive and battery life has been improved.

Our findings in patients with treatment-resistant arterial hypertension cannot be simply extrapolated to other potential indications. Recent clinical trials utilizing unilateral unipolar electrical carotid sinus stimulation in heart failure patients yielded promising results.^34–36 However, reconfirmed efficacy data from controlled clinical trials is mandatory regardless of the indication. Finally, considering the invasiveness and costs of electrical carotid sinus stimulation, we should find ways to identify patients most likely to benefit.

Perspectives

In our patient cohort, unilateral unipolar electrical baroreflex stimulation elicited interindividually variable reductions in sympathetic activity and blood pressure in patients with resistant hypertension. However, side effects may limit efficacy. Pending results from controlled clinical trials, the approach should not be applied in routine clinical care. Our study suggests that careful physiological profiling can aid in assessing antihypertensive treatments and could be incorporated at an earlier stage in clinical development. The fact that compared to medications, device-based treatments require much less rigorous clinical testing in many countries should be scrutinized. The same is true for technological modifications, which – as shown in our study – may alter efficacy and side-effects.

Novelty and Significance.

What Is New?

• Electrical carotid sinus stimulation using a unipolar unilateral device elicits interindividually variable reductions in vasoconstrictor sympathetic activity and arterial pressure in patients with refractory hypertension.

• A considerable proportion of patients experiences side effects necessitating reduction of stimulation intensity.

What Is Relevant?

• Titrating carotid sinus stimulation intensity to chronically tolerable levels may be associated with marked reduction in efficacy.

• The efficacy of new or substantially modified antihypertensive devices has to be confirmed in properly controlled clinical trials.

Summary

Electrical carotid sinus stimulation has been developed as non-pharmacologic option for the management of treatment-resistant arterial hypertension. Shortcomings of first-generation devices led to the development of a new approach utilizing unilateral unipolar stimulation with a much smaller electrode. Using high-fidelity cardiovascular phenotyping, we confirmed the blood pressure lowering potential and the contribution of baroreflex-mediated sympathoinhibition. However, introduction of this less invasive approach may have sacrificed efficacy, partly because stimulation intensity has to be reduced in order to diminish side effects. The device is approved in Europe and routinely implanted but has not been tested in controlled clinical trials. Our study challenges this practice.

Abbreviations

au

arbitrary units (sympathetic activity)

BP

blood pressure (arterial)

bpm

beats per minute

BRS

cardiac baroreflex sensitivity

lf

low-frequency band

us

up sequences

ds

down sequences

DBP

diastolic blood pressure

FBP

finger blood pressure (arterial)

HR

heart rate

MSNA

muscle sympathetic nerve activity

PP

pulse pressure

SBP

systolic blood pressure

SE

side effects