Improved Control of Resistant Hypertension With Device‐Mediated Electrical Carotid Sinus Baroreflex Stimulation

A 36‐year‐old African American woman was referred for evaluation of refractory hypertension with blood pressure (BP) readings in the range of 200 to 230 mm Hg systolic and 100 to 110 mm Hg diastolic. She had been hypertensive for more than a decade with severe elevations in BP during the previous 4 years. Her hypertension was unsuccessfully treated with combination therapy of multiple medications. The patient had been obese all of her life, despite counseling and attempts at dietary and lifestyle modifications. She also had a history of obstructive sleep apnea treated with bi‐level positive airway pressure (BiPAP) for 6 months without apparent effect on BP. At presentation, the patient was asymptomatic except for occasional frontal headaches and dysnea with exertion, which forced her to accept disability as a nurses' aide. A strong maternal family history of hypertension included a grandmother who developed end‐stage renal disease secondary to hypertension. There was no history of alcohol consumption, smoking, or recreational drug use. Current medications included furosemide 160 mg twice daily, metolazone 5 mg once daily, potassium chloride 30 mEq 3 times a day, 2 clonidine transdermal patches 0.3 mg/24 h applied weekly, amlodipine 10 mg/d, metoprolol XL 200 mg/d, lisinopril 40 mg/d, telmisartan 160 mg/d, minoxidil 10 mg twice daily, and isosorbide 30 mg/d. Addition and uptitration of peripheral α₁‐receptor or aldosterone receptor antagonists failed to lower her BP level. On physical examination, the patient was 154.9 cm tall and weighed 119.7 kg (body mass index 49.9). Her BP was 220/120 mm Hg in both arms measured with a large cuff, and her pulse rate was 70 to 80 beats per minute. No supraclavicular or postcervical fat pads, striae, bruits, or edema were noted. All peripheral pulses were full, and no bruits were heard over the cervical, epigastric, or femoral vessels. Radial and femoral upstrokes were simultaneous. Fundoscopic examination revealed arteriolar narrowing but without arteriovenous crossing change. Urinalysis results were normal except for microalbuminuria. Serum creatinine level was 1.1 mg/dL with an estimated glomerular filtration rate of 72 mL/min/1.73 m². Serum potassium level was 4.1 mEq/L and serum CO₂ level was 26 mEq/L. Thyroid‐stimulating hormone, free T4, and all other routine biochemical test results were normal. Echocardiography demonstrated left ventricular hypertrophy. Carotid ultrasonography revealed no abnormalities.

CASE DISCUSSION

This patient was diagnosed with resistant hypertension. Her BP exceeded 160/90 mm Hg when measured with a large BP cuff; white coat hypertension was ruled out with 24‐hour ambulatory BP monitoring. The latter revealed a mean pressure of 188/114 mm Hg with a mean arterial pressure of 141 mm Hg.

No historical or physical examination findings suggested secondary hypertension. Nonetheless, a reasonable screening workup to exclude secondary causes of hypertension was undertaken. Doppler ultrasonography of the kidneys was negative for renal artery stenosis. Given the low pretest probability of renal artery disease, neither magnetic resonance angiography nor renal arteriography was pursued. The serum aldosterone‐to‐renin ratio was 16. A 24‐hour urine test produced the following findings: total creatinine 1.6 g/d; metanephrines 194 µg/d; norepinephrine 17 µg/d; epinephrine 13 µg/d; vanillylmandelic acid 4.3 µg/d; dopamine 400 µg/d; and aldosterone 7 µg/d. All results were within normal range. Results of the 24‐hour urine test also showed a reasonable, although less than ideal, dietary sodium level of 185 mmol/d or 4.2 g/d. Morning plasma cortisol level was <1 µg/mL after a 1‐mg dose of dexamethasone was given the night before. The patient appeared to be on a more‐than‐adequate regimen of medications with agents that inhibit renin‐angiotensin access (lisinopril, telmisartan, and metoprolol), sympatholytic agents (clonidine, metoprolol), and agents that appropriately address sodium‐volume/calcium dysmetabolism (furosemide, metolazone, and amlodipine) at seemingly adequate doses.

She had no history of use of drugs such as nonsteroidal anti‐inflammatory drugs, pseudoephedrine, or other over‐the‐counter medications that would attenuate the effectiveness of any of the antihypertensives. While one might doubt the patient's sworn adherence to her medication regimen, she required increasing doses of diuretics after being placed on minoxidil and expressed concern about further increases in the minoxidil dose after developing hirsutism. These factors plus the evidence of adherence to a 4‐g/d sodium diet (based on the 24‐hour urine test) added credence to her self‐proclaimed compliance with her medical regimen. As noted, the patient did have obstructive sleep apnea and applied a BiPAP device on a nightly basis for at least 6 months; however, no decrease in the level of BP was observed at her office visits or by the patient at home using a self‐monitoring device.

An important lifestyle‐related factor was the patient's obesity. A body mass index of nearly 50 heightened suspicions that obesity‐related mechanisms were contributing to her hypertension. Obesity and associated insulin resistance can lead to hypertension via several different mechanisms, with over‐activity of the sympathetic nervous system playing a central pathogenic role. ¹ Sodium retention can occur as a result of enhanced sympathetic nervous system and renin‐angiotensin system effects on the kidney; however, antihypertensive therapy simultaneously directed at blocking the sympathetic nervous system, the renin‐angiotensin system, and extracellular volume expansion proved ineffective. Visceral obesity has also been associated with increased aldosterone levels. ² Our patient failed to respond to aldosterone receptor antagonist, however.

While bariatric surgery was considered, the patient's elevated BP precluded this. Given the failure of conventional antihypertensive therapy to control the patient's BP and the risks of associated end‐organ damage, novel therapies offered by clinical trials were considered. Enrollment into a trial of an endothelin receptor antagonist (Darusentan; Myogen, Westminster, CO) was considered, but exclusion criteria disqualified her. Conversely, the patient qualified for entry into a study of device‐mediated electrical carotid sinus baroreflex stimulation (Rheos Baroreflex Activation System; CVRx, Minneapolis, MN). Direct electrical stimulation of the carotid nerve to activate the baroreflex and reduce BP had been demonstrated in human studies, ³ , ⁴ , ⁵ but the approach had considerable technologic limitations. BP reduction by means of baroreflex activation occurs by attenuation of central sympathetic traffic and downstream reduction in sympathetic activity to the kidneys, down‐regulating the renin‐angiotensin system with resultant decreases in salt and water retention and angiotensin II mediated vasoconstriction. ⁶ , ⁷ In 2004, Lohmeier colleagues ⁸ demonstrated sustained reduction in both heart rate and BP with chronic baroreflex stimulation via electrodes implanted about both carotid arteries in dogs. Greater understanding of electrophysiology and improvements in implantable technology have resulted in refinements to baroreflex‐stimulating devices to control BP. ⁹ Recent trials of a more sophisticated device in humans have demonstrated promising results in patients with resistant hypertension. European and US feasibility clinical trials are evaluating the Rheos System in patients with drug‐resistant hypertension (Figure 1). Early results for the first 12 European and 10 US patients have been reported. ¹⁰ , ¹¹ After 3 months of active Rheos therapy, systolic BP was reduced by an average of 24 mm Hg (Europe) and 22 mm Hg (US). The implants were well tolerated and there were no unanticipated serious adverse events related to the system or procedures.

Figure 1.

Rheos Baroreflex Activation System.

The Rheos device was surgically implanted in our patient; intraoperative testing produced a progressive fall in BP in response to increasing electrical stimulation of the carotid sinuses (Figure 2). The device then was turned off and remained dormant for 1 month to preclude occurrence of any surgical complications that could confound results. At the end of 1 month, the device was turned on and a stepwise increase in electrical stimulation was delivered to the carotid sinuses with a target of reducing BP by 10% per month. The patient continued taking her usual antihypertensive medications after implantation. The office cuff BP was 204/142 mm Hg preimplantation and 165/108 mm Hg postimplantation but before device activation. After device activation, office cuff pressures were generally improved, but not significantly so (Table). Evaluable 24‐hour ambulatory BP recordings (defined by protocol as >70% successful recordings) were available only for the 1‐, 3‐, 6‐, and 12‐month post‐device activation time points, but all data are included for completeness (Table, Figure 3). Mean ambulatory BP readings were significantly improved (P<.001) over the course of baroreflex stimulation. The improvement in BP was not attributable to change in weight, which remained unaltered over the course of the study. After 1 year of active carotid baroreflex stimulation, the patient remained on the original medications except for minoxidil and telmesartan. With resolution of her headaches and cardiovascular symptoms, she returned to work full time as a nurses' aide.

Figure 2.

Initial dose response of systolic blood pressure (BP) level following activation of Rheos System.

Table.

Blood Pressure (BP) Levels Before and Up to 12 Months After Implantation of the Rheos Baroreflex Activation System

Time, mo	OC BP, mm Hg	AMB BP, mm Hg	Voltage
−1	204/142	188/114 (54%)	‐
0	165/108	182/108 (73%)	‐
1	230/130	206/128 (91%)	3–4
2	160/128	181/110 (58%)	4–5
3	133/68	137/79 (82%)	1.5
6	200/97	151/84 (70%)	2–4
9	207/99	152/91 (60%)	5.5
12	131/78	127/77 (77%)	7
Abbreviations: AMB BP, ambulatory BP; OC BP, office cuff BP. Time is marked from the point when the Rheos System was activated (immediately following Time 0). Per protocol, several OC BP measurements were taken separated by 1 to 3 minutes until the readings varied by <5 mm Hg. The last 2 were averaged and recorded. Percent of successful recordings is shown for each value in parentheses. Voltage was the amount of current applied to the carotid baroreceptor leads from the pulse generator. For several months, the device was programmed to supply different amounts of voltage during periods of the day when BP was higher, documented on previous AMB BP recordings.

Figure 3.

Ambulatory blood pressure measurements over the course of the study period. Open squares represent the systolic pressure, filled squares represent diastolic pressure, and filled circles represent the calculated mean pressure: (2 × diastolic + systolic)/3. At 1 month, the pressure is higher than baseline (P<.05). At 3 months, the pressures are significantly below the initial pressures (P<.001), and this is sustained through the remaining months of the trial. The data are corrected for multiple comparisons (Dunnett method).

The response to baroreflex stimulation lends credence to the suspected pathophysiologic underpinnings of resistant hypertension in this patient. The reasons previous pharmacologic sympatholysis failed to bring about adequate control in this patient are not understood but may relate to mechanisms overriding central α₂‐agonism and peripheral α₁‐antagonism.

CONCLUSIONS

Hypertension is a major cause of and contributor to stroke, heart disease, and kidney disease. Despite the development of numerous medications to treat hypertension over the past half‐century, its adequate treatment continues to be a major problem in the United States. The Third National Health and Nutrition Examination Survey (NHANES III) found that only 34% of hypertensive individuals reach a BP <140/90 mm Hg. ¹² Device‐mediated electrical carotid sinus baroreflex stimulation offers the potential to bridge this gap in antihypertensive control. It may ultimately facilitate our ability to achieve lower BP targets and attenuate cardiovascular and renal disease in populations at greatest risk. The device is currently being tested in the blinded Rheos Pivotal Trial to determine its efficacy in resistant hypertension compared with medical therapy alone. While the initial data are promising, we will need to await results of these phase II/III studies before application to the broader hypertensive community can be made.

Acknowledgments:

The authors wish to thank Mark Shelly, MD, for statistical assistance and Colin Sloand for data collection.