Despite recent advances in antihypertensive pharmacotherapy, hypertension continues to remain uncontrolled in a large percentage of the hypertensive population. Adding to this concern are recent projections suggesting that up to 50% of adults in developed countries will meet the standard definition of hypertension by 2025. Some of the most challenging patients are those with resistant hypertension, defined as a blood pressure (BP) above target despite use of at least 3 antihypertensive drugs, one of which is a diuretic, all taken at maximal (or highest‐tolerated) doses. The exact prevalence of treatment‐resistant hypertension (TRH) continues to be debated, but recent studies using clinic‐measured BP suggest that 20% to 30% of patients with hypertension are affected. With its associated comorbidities of obesity, diabetes, and chronic kidney disease, the prevalence of TRH is expected to increase in the coming decades.
Observational studies have shown that adding specific antihypertensive drugs (such as chlorthalidone) or classes of drugs (such as mineralocorticoid receptor–blocking agents) helps to control BP in patients with TRH. More recently, device‐based therapy has been studied. One such investigational strategy has been designed to inhibit the neurogenic component of hypertension, which is often abnormal in these individuals. The Rheos Pivotal Trial (Rheos System, CvRx, Inc, Minneapolis, MN) was the first large‐scale, manufacturer‐sponsored, randomized, double‐blind, parallel‐design clinical trial performed in the United States to assess the safety and efficacy of baroreflex activation therapy (BAT) delivered through a surgically implanted device in patients with TRH. The device consists of a pulse stimulation generator placed subcutaneously on the anterior chest wall with bilateral electrodes that are tunneled to each carotid sinus. It works by delivering an exogenous source of energy to the carotid baroreceptors, interpreted in the central medulla as a rise in BP. The brain then sends sympatho‐inhibitory signals to the blood vessels, heart, and kidneys resulting in a reduction of BP. To be eligible for the study, patients had to have resistant hypertension (≥3 antihypertensive medications, including a diuretic) for at least 1 month with at least 1 outpatient, in‐office systolic BP ≥160 mm Hg and a diastolic BP ≥80 mm Hg. In addition, patients had to be a surgical candidate for device implantation without significant carotid stenosis and have an average 24‐hour ambulatory systolic BP ≥135 mm Hg without significant orthostatic hypotension. Forty‐nine centers consented and screened 590 patients between March 2007 and November 2009. Each center was allowed to implant up to 2 nonrandomized patients prior to initiating the randomized portion of the trial to become familiar with the procedure. A total of 265 patients were randomized and followed for 12 months (2 patients in group A for every 1 patient in group B) with the device implanted by either a vascular, cardiothoracic, or neurosurgeon. It was turned off in all patients for the first month. The 181 patients in group A had the device activated at 1 month (time zero) (immediate BAT activation) while the 84 in group B continued to have the device turned off until 6 months into the trial (BAT deferred), with blinding maintained for a total follow‐up of 12 months. Once the device was activated, stimulation was adjusted according to a protocol‐defined algorithm with best practices shared across centers. Patients and investigators remained blinded to treatment until after the 12‐month visit. Outpatient BP was measured in the office using the BpTRU automated device (VSM Medtech Ltd, Vancouver, Canada) taken within 4 to 6 hours of the most recent dose of antihypertensive medication with the investigator out of the room. Programmed to take 6 measurements at 1‐minute intervals, the first measurement was discarded with the device reporting the average of the last 5 measurements. This was done to minimize any white‐coat effect. Investigators were allowed to change antihypertensive medications in both arms of the study during the course of the trial. There were 5 prespecified co‐primary end points, 2 for efficacy (acute efficacy, sustained efficacy) and 3 for safety (procedural safety, BAT safety, and device safety). All five end points had to be met in order to meet the primary end point. Secondary end points were prespecified and included mean change in office systolic BP and a comparison of immediate vs deferred BAT. Sample size was calculated to appropriately power all co‐primary end points. A pre‐established statistical analysis plan was developed instructing the data monitoring committee (DMC) to perform an interim statistical analysis at 6‐month intervals remaining blinded to treatment group. The 2 randomized groups were well matched for clinical baseline and demographic characteristics (mean age, 53 years; 61% men; 80% Caucasian). On average, patients had received 5.2±1.7 medications for 21±8 months, with more than 90% taking a diuretic. A total of 91% were either taking a β‐blocker or sympatholytic agent. Only 1 patient (in group A) was lost to follow‐up. The DMC recommended and all patients were allowed to complete the 6‐month visit in a blinded fashion. After all patients had been enrolled and implanted, but before 95 patients had completed the 6‐month visit, the DMC advised the sponsor that the trial was unlikely to attain significance for the acute efficacy analysis. The acute responder analysis at 6 months (defined as at least a 10‐mm Hg drop in systolic BP at month 6) yielded a 54% response rate in group A and 46% in group B, which failed to meet significance with a 20% predefined superiority margin (P=.97). For sustained efficacy at 12 months (defined as a reduction from month 0 to month 12 of at least 10 mm Hg in systolic BP and to remain at least 50% of that seen at month 6), 88% of patients in group A responded, surpassing the prespecified objective performance criterion (OPC) of 65% (P<.001). The rate of freedom from procedural complications within 30 days after implantation was 75%, short of the prespecified OPC of 82% (P=1.00). Procedural complications were most often related to carotid sinus lead placement, and most (76%) resolved completely. There were no procedure‐related deaths. The rate of freedom from adverse BAT‐related events between day 30 and month 6 was 92% in group A and 89% in group B (P<.001 for noninferiority). The rate of freedom from major hypertension‐ and device‐related adverse events from day 30 to month 12 was 87%, surpassing the prespecified OPC of 72% (P<.001). Secondary end points including the mean change in systolic BP at 6 months from month 0 was 16±29 mm Hg for group A and 9±29 mm Hg for group B (P=.08). Mean decrease in systolic BP at 12 months from month 0, at which point group A had received 12 months of BAT and group B had received 6 months of BAT was 25±32 mm Hg for group A and 25±31 mm Hg for group B. An ancillary analysis was performed on the percentage of patients achieving a systolic BP ≤140 mm Hg at 6 months (P=.005) but with no difference seen at 12 months (P=.70) when both groups had received BAT for at least 6 months. In an effort to allow more direct comparisons to drug therapies, a post hoc analysis looking at the change from the pre‐implant period rather than from time zero was performed due to unexpected differences between the systolic BP values during this period, which now showed a significant difference at 6 months (P=.03) but no difference at 12 months (P=.57). The Rheos trial was the first large‐scale device trial in the United States in patients with TRH. The trial did not meet 2 of the 5 prespecified coprimary end points, both in the short‐term analysis (for efficacy and safety). While BAT was associated with mean reductions in systolic BP of up to 35 mm Hg with more than 50% of patients able to reduce systolic BP to ≤140 mm Hg, future studies need to address design flaws in the present study that caused it to be stopped early in an effort to better understand whether a clinical benefit exists for future BAT device technology1. Bisognano JD, Bakris G, Nadim M, et al. Baroreflex activation therapy lowers blood pressure in patients with resistant hypertension: results from the double‐blind, randomized, placebo‐controlled Rheos Pivotal Trial. J Am Coll Cardiol. 2011;58:765–773.
Comment
Many forms of hypertension have a significant neurogenic component, including primary hypertension, obesity‐associated hypertension, hypertension associated with sleep apnea, renal hypertension, pregnancy‐associated hypertension, as well as TRH. In these conditions, BP elevation is initiated and sustained in part by activation of the sympathetic nervous system (SNS). A decrease in parasympathetic and/or an increase in SNS activity is associated with an increase in peripheral vascular resistance and a reduction in renal blood flow and leads to sodium and water retention, which all contribute to chronic hypertension. Abnormalities of the baroreflex have long been recognized in those with hypertension. As acute elevations in BP occur, arterial afferent baroreceptors fire rapidly in an attempt to offset the elevation in BP. As sustained elevations become more chronic, the ability of the carotid sinus baroreceptor to respond diminishes over time with a new threshold for activation established. In an effort to “reset” the baroreflex via exogenous stimulation, a surgically implantable device for the treatment of resistant hypertension has been developed. The Rheos (CVRx, Inc) System administers BAT via electrical stimulation of the carotid baroreceptors. The present trial was negative and did not meet 2 of the 5 prespecified co‐primary end points for the short‐term efficacy and short‐term safety analysis. These possibly relate to design flaws in the trial. In terms of the short‐term efficacy, the larger‐than‐expected reduction in systolic BP during the first 6 months of the trial in group B when the device was turned off can be attributed to allowing investigators to increase the dose or add antihypertensive medications to this very group being used as the comparator group to those receiving device‐based BAT (group A). This minimized any ability to achieve a significant BP difference between groups. Furthermore, taking BP readings 4 to 6 hours after medication was received rather than at the trough period might have contributed to a further reduction in BP, exaggerating the benefits that added antihypertensive medication allowed in the protocol.
The short‐term procedural safety measure, comparing the procedure‐related or system‐related adverse event‐free rate for events occurring within 30 days failed to meet the performance criterion of 82%. This is most likely due to the historical figure used from implantable cardioverter‐defibrillator trials that was higher than the adverse event profile from carotid endarterectomy trials, which are much more similar to those found in the present BAT trial.
So what is the clinician to make of these findings? Excepting the clinical design flaws as already mentioned, the results remain encouraging news for a group of patients with few choices and substantial morbidity and mortality. While this technology is investigational (ie, as yet unavailable) it continues to evolve. Subsequent studies in larger, more‐diverse cohorts are needed. However, the device seems to represent a novel, viable approach to resistant hypertension. Ultimately, the benefits of the device will have to be weighed against the cost and invasive nature of the procedure. However, as the device is further tested, with future improvements in its technology, it may allow a reduction in the number of antihypertensive medications required to control the most treatment‐resistant patients. Until those clinical trials occur, BAT with the present Rheos system, is not yet ready for prime time.