Baroreflex activation therapy for resistant hypertension: promise present or passed?

Arterial baroreceptors adapt to the prevailing level of blood pressure by resetting upward, requiring a higher threshold for stimulation [1]. Over time, progressive reductions in conduit arterial compliance will diminish further percussion wave induced mechanoreceptor stretch. Consequently, circumventing these adaptations by direct stimulation of afferent nerves arising from the carotid sinus is an appealing concept. However, that chronic stimulation of baroreceptor afferent nerves should elicit sustained reduction of blood pressure of individuals with longstanding hypertension may not be intuitively obvious. Bilateral carotid baroreceptor afferent nerve stimulation, via surgically implanted bipolar electrodes, has been shown to reduce, acutely, the SBP, efferent sympathetic discharge to skeletal muscle, plasma renin concentration, and to a lesser extent, the heart rate of individuals with drug-resistant hypertension [2,3]. However, in the spontaneously hypertensive rat, the reflex sympatho-inhibitory response to aortic baroreceptor nerve stimulation is attenuated due to central resetting [4]. Hence, the counter-argument, and source of equipoise for therapeutic deployment, is that if reflex resetting occurs within the central nervous system, as a neuroplastic adaptive response to sustained increases in afferent input [5], the sympatho-inhibitory vasodepressor response to baroreceptor nerve stimulation may attenuate, or vanish over time. Hence the value of a randomized trial evaluating the long-term durability of baroreflex activation therapy, and in particular one of contemporary technology, involving unilateral placement of a single unipolar electrode on a carotid sinus wall. The first pivotal randomized controlled trial of BAT for drug-resistant hypertension evaluated the Rheos System (CVRx Inc, Minneapolis, Minnesota, USA) [6]. Stimulating leads connected to a pulse generator were tunneled subcutaneously to engage each carotid sinus. Participants (265) were randomly allocated, 1 month post implant to either immediate BAT continuing for 12 months, or deferred BAT, which commenced stimulation 6 months after randomization and continued a further 12 months. Stimulation parameters were increased incrementally over time according to a protocol-defined algorithm with support from field clinical engineers. Overall, there was a mean reduction in systolic pressure of 35 mmHg, compared with preimplant values, but accompanied by a notable number of adverse safety events. However, the mean decrease in SBP at 6 months was not significantly different between groups, due to a substantial decline in the deferred group; nor was the proportion of participants whose SBP fell at least 10 mmHg at 6 months compared with baseline statistically significant. In those with 12 months of stimulation, fewer hypertensive crises were reported and in 88 of responders, a systolic reduction was sustained for up to 12 months. In an uncontrolled follow-up aggregating 383 patients from three trials, blood pressure reductions were sustained for up to 6 years, with greater reductions observed in those with heart failure with preserved ejection fraction [7]. Uncontrolled studies of unilateral BAT, via the Barostim neo device (CVRx Inc), have also documented blood pressure lowering, sustained up to 2 years [8] with acute fluctuations upon deactivation [9]. In a trial involving 16 patients treated for an average of 2.7 years who were randomized to 4 weeks of blinded withdrawal, mean home blood pressure did not change, automated office readings rose by 10/4 mmHg on average, and ambulatory blood pressure by 10/8 mmHg, the latter two variables statistically significant, but much less than the initial blood pressure drop with stimulation [10]. Importantly, as Biaggioni recently highlighted, no randomized trials to determine the safety and efficacy of this second-generation technology for the treatment of hypertension have been published [11]. In the current issue of the Journal, the Cologne group who authored these latter studies now report the findings of a randomized controlled trial commenced over a decade ago involving 16 patients (aged 67 ± 8.5 years, six women) diagnosed with resistant hypertension (prescribed at least three drugs from different antihypertensive classes, including one diuretic), who were scheduled for generator replacement after a median of 50 months (IQR 38–77 months) of unilateral BAT [12]. These individuals consented to a protocol involving randomization to 4 weeks of device deactivation or 4 weeks of continued stimulation over the 8 weeks preceding surgery. Participants were implanted between 2011 and 2017, enrolled between 2016 and 2019, and assessed between 2017 and 2020. Heart failure with reduced ejection fraction was an exclusion criterion. Heart failure with preserved ejection fraction was not. Criteria (e.g., an increase in office blood pressure of ≥ 30 mmHg within 5 min of deactivation; ≥ 30/15 mmHg increase in home blood pressure readings over 3 consecutive days) were established, a priori, to mandate device reactivation on safety grounds. A baseline visit with office blood pressure recordings acquired with the device active was scheduled at approximately 12 weeks before the surgical date; thereafter, medication changes were discouraged. ‘Blinded’ clinic (average of ≥ 3 readings) and telemetered home measures (three per day) were evaluated. Antihypertensive therapy (mean 5.4 agents) was stable throughout the study period. In two patients, in-office deactivation was associated with a surge in SBP, obliging immediate reactivation for safety. In the remainder, there was an immediate, but nonsignificant rise in average values for systolic and diastolic pressure. Clinic SBP was 8 ± 14 (mean ± SD) higher in the absence of baroreceptor stimulation (P = 0.06). In 13 with available paired data, home SBP was 5 ± 7 mmHg (P < 0.02) higher in the absence of baroreceptor stimulation. Average DBP was 3 mmHg lower (P < 0.02). Heart rate was identical under both conditions. One week appeared sufficient to document the maximum off-response: readings acquired 1 and 4 weeks after deactivation were similar. The authors’ interpretation was that baroreceptor activation therapy indeed had durable blood pressure lowering effect. As supportive data, they highlighted substantially lower average SBP at the time of enrolment, relative to preimplantation values (147 ± 27 vs. 171 ± 21 mmHg; P < 0.01; ≥ 5 mmHg lower in 12/16) plus surges in blood pressure, in a few, upon deactivation. However, overall, only nine of the cohort enrolled (60%) were judged to be responders to BAT, based upon systolic pressure increasing by 5 mmHg or more following deactivation, that is, a response analogous to single-drug therapy. The authors describe theirs as an exploratory study, and acknowledge that is it limited by the small number of participants enrolled over a 3-year period (vs. 265 in the pivotal Rheos trial [6]), selection bias (participants perceiving benefit of BAT declined randomization), experience confined to two centres, reliance on intermittent home and clinic readings rather than ambulatory blood pressure recordings (hence absence of 24-h data) as had been acquired in their prior study [13], and phenotypic heterogeneity: of the 16 with resistant hypertension studied, seven were deemed to have also heart failure with preserved ejection fraction. For logistical reasons, further recruitment became infeasible after 2019. Heart rates were similar in the stimulated and deactivated states. Unlike reflex gain for sympathetic vasoconstrictor outflow, which is preserved in drug-sensitive primary hypertension [14], that for heart rate is attenuated [14,15], suggesting less potential for afferent nerve stimulation to augment efferent vagal discharge. The investigators were unable to identify potential predictors of the blood pressure response to BAT. One possibility for such obscurity may be between-participant heterogeneity of central reflex resetting and hence responsiveness to acute or chronic afferent stimulation. Dell’Oro et al.[16] note that in contrast to drug-responsive primary hypertension, the baroreflex regulation of sympathetic nerve firing is attenuated, by more than 30% on average, in patients with drug-resistant hypertension. In experimental preparations, Chapleau et al.[5] demonstrated that baro-reflex mediated inhibition of sympathetic nerve firing is maintained during pulse phase afferent activity, suggesting pulse pressure as a potential modulator. In contrast, continuous nonphasic baroreceptor discharge or a rapid (>1.5 Hz) pulse phasic discharge disinhibited efferent sympathetic discharge. Such disinhibition was amplified by age. These latter findings may be relevant to current BAT stimulation algorithms (10–100 Hz); sudden release of afferent restraint on such disinhibition might underlie the investigators’ observation that blood pressure surges minutes after the device is switched off, in a few to concerning levels, regardless of time elapsed since implantation. If central reflex resetting emerges over time but is modifiable by pulsatile input, this could have therapeutic implications for with other potential interventions, such as endovascular dilatation of the carotid sinus [17]. The present data by Ernst et al.[12], acquired via a unique study design, are modestly positive, with a blood pressure effect in the clinically meaningful range, but will this and any future conceived randomized clinical trials of this nature increase the appetite of clinicians to recommend and the willingness of patients to accept this invasive procedure, which encompasses a permanently implanted circuit, anticipation of future generator replacement, and potential safety implications if deactivation triggers acute surges, to aid the management of their hypertension? It is perhaps telling that at these investigators’ own experienced site, renal denervation has superseded baroreflex activation therapy as the predominant procedure performed for device management of hypertension. ACKNOWLEDGEMENTS Conflicts of interest There are no conflicts of interest.