What Price For Blood Pressure Control?

Clinicians responsible for hypertensive patients recognize that some hypertension is resistant to drug therapy and therapeutic goals are not reached. While many treatment failures undoubtedly reflect suboptimal antihypertensive drug therapy, high costs or poor adherence to medications, others represent genuine resistance to effective medical regimens. Developing powerful new tools to expand treatment options for resistant hypertension has long been a “Holy Grail” in this field.

Most authors define resistant hypertension as blood pressures (BPs) that remain above goal despite adherence to 3 drugs at full doses, one of which is a diuretic. ¹ The most recent National Health and Nutrition Education Survey (NHANES) indicates that just over 53% of patients were at levels <140/90 mm Hg. ² In a larger percentage of older patients, this standard is not achieved with drug therapy. Achieving lower BPs commonly recommended for diabetic patients and for those with chronic kidney disease remains elusive. Recent guidelines recognize that most such patients require complex drug regimens that carry substantial cost and may produce poorly tolerated side effects.

How, then, should we position newly developed surgically implantable devices that lower BP by electrical stimulation of carotid baroceptors? The past few years have brought a flurry of developments in this regard. Contrary to views that suggest pressure‐mediated stimulation of baroceptors acts mainly to modulate short‐term swings in arterial pressures, recent studies demonstrate that sustained electrical stimulation of nerves emanating at the carotid body can lower both systolic and diastolic BPs indefinitely. ³ BPs in both normal and experimental hypertension can be reduced by 20 to 30 mm Hg associated with a fall in circulating norepinephrine levels and sympathetic nerve traffic. The renal nerves do not appear to be required for these effects, at least for pressure reduction in normal animals. ⁴ Technical advances have led to successful deployment of bilateral carotid body–stimulating electrodes in humans, triggered by a subcutaneous battery pack that can be programmed externally. ⁵ The commercial entity producing this product is based in Minneapolis, MN, (Rheos, CVRx) and identifies this device as the Rheos Baroreflex hypertension therapy system. ⁶ Although clinical trials on a large scale are only just beginning, initial studies of 16 patients followed for more than 2 years (mean 33 months) have been presented as the Device‐Based Therapy of Hypertension (DEBuT‐HT study). Systolic BP was lowered by 35±8 mm Hg and diastolic BP by 24±6 mm Hg. ³ For some patients, this was accompanied by reduction in antihypertensive drug therapy requirements and reversal of left ventricular hypertrophy. These results have been impressive and exceed antihypertensive effects commonly reported as the basis for Food and Drug Administration drug approval.

These benefits are not likely to come cheaply. The Rheos trial currently entails bilateral carotid body electrode placement with intraoperative “mapping” of the hemodynamic response to stimulation for optimal placement. Hence, placement of this device entails anesthesia, surgical exposure and exploration of both carotid sinuses, and perioperative testing and care. Anticipated costs for the device and implantation range between $20,000 and $30,000. ⁷ The pivotal trial now in progress will compare randomized periods of stimulator “ON” and “OFF” to evaluate efficacy, durability, and safety of this device. Up to now, reported adverse effects have been limited mainly to local tissue stimulation and/or adjacent nerve dysfunction, usually related to the magnitude of applied voltage. There have been at least two significant device infections leading to explantation of the electrodes and/or stimulator. ⁶ The antihypertensive efficacy of the Rheos device has been impressive. Sustained BP reductions have been observed that allow withdrawal of some medications. No major electrolyte or cardiovascular disturbances (such as changes in heart rate) have been reported up to now. Will the benefits of this approach over the long‐term outweigh the risks and costs?

In this issue of The Journal of Clinical Hypertension, Young and colleagues from the University of Rochester report results of formal Markov modeling to address the cost‐effectiveness of this implantable device. They present a detailed consideration of a “base case” (age 50) typical of their initial trials with risk levels comparable to the Framingham studies, assuming a best medical therapy systolic BP of 180 mm Hg and a reduction in BP by 20 mm Hg with the Rheos device. Their model assumes risk reduction for cardiovascular adverse events, such as stroke and heart failure, that have been observed in treatment trials with antihypertensive drug therapy, and costs of surgical implantation at about $20,000 (2007 US dollars). Risk reduction is translated into quality‐of‐life‐years gained (quality‐adjusted life‐year [QALY]) as a result of sustained effective BP reduction. The measure presented then represents the cost per life‐year gained, or in the case of comparative or additional benefits, the additional (or incremental) cost per life‐year (ICER). This resulted in an overall incremental cost‐effectiveness ratio (ICER) of $64,000 per QALY. If the efficacy of the device actually lowered BP by 30 to 35 mm Hg (as reported in DEBuT‐HT) the ICER fell to less than $34,300 per QALY. Such a cost is often considered acceptably cost‐effective compared with other interventions, such as colonoscopy and kidney transplant. If one considers a higher‐risk, slightly older hypertensive population such as those enrolled in the Anglo‐Scandinavian Cardiac Outcomes Trial–Blood Pressure–Lowering Arm (ASCOT‐BPLA), the ICER was only $26,700 per QALY. Such an intervention might be cost‐effective down to systolic BP levels as low as 158 mm Hg.

How does one interpret such incremental costs? Sometimes this is simply presented as a comparison with other procedures, such as implantable cardiac defibrillators or deep brain stimulation for Parkinson’s disease as the authors suggest. ⁷ Conventional analyses require ICER levels below $50,000 to be widely accepted. Whether these figures are truly relevant to individual patients is questionable. The authors do provide sensitivity analyses that support increased cost‐effectiveness in specific populations at highest risk for complications from untreated hypertension. These include women more than men, particularly those with higher starting systolic BP. Interestingly, the net gain with older age groups is limited and this procedure actually loses cost‐effectiveness as ages move above 60 years old.

One may challenge the quantitative assumptions in Markov models. Nonetheless, this approach does provide a standardized frame of reference to consider cost‐effectiveness. The careful analysis by Young and colleagues underscores the potential for this remarkable device to reduce morbidity and mortality for resistant hypertension. It does so by avoiding devastating cardiovascular events, such as stroke, albeit while incurring substantial (initial) costs related to treatment. The authors acknowledge that such devices are not for everyone and the overall ICER for the base case is “indeterminate” at more than $64,000 per QALY. The actual costs for this device are not yet known with certainty. The value of the Young study derives from the assembly of reasonable sensitivity analyses that allow prediction of the patient characteristics and the required therapeutic effect that most likely will determine its clinical value. While cost is only one dimension that determines antihypertensive therapy, it is more important now than ever. Becoming familiar with analyses such as the cost‐effectiveness model here will be essential for regulators of health care delivery, patients, and clinicians alike.