Renal Denervation for Lowering Blood Pressure: Comments on the AHA Scientific Statement

Hypertension is unequivocally an established risk factor for cardiovascular (CVD) -renal disease as well as premature mortality. Importantly, pharmacological blood pressure (BP) lowering reduces these events for patients with and without pre-treatment CVD across a broad range of SBP levels – even at levels not currently recommended for treatment.¹ Until just recently, patients with hypertension and their clinicians had two major strategies for lowering BP – diet/lifestyle interventions and pharmacological BP lowering with over 100 FDA approved drugs. The therapeutic armamentarium for lowering BP has now expanded to include devices with FDA approval in late 2023 of two renal denervation systems as adjunctive therapy for drug-treated patients without adequate BP control. The FDA has an established precedent of accepting sustained BP reductions as a surrogate for cardiovascular (CVD) disease risk reduction. Accordingly, most antihypertensive drugs, as well as these renal denervation systems, do not have long-term clinical trial hard endpoint data showing their clear clinical benefit on clinical CVD risk reduction. Also, as BP targets have moved incrementally lower (<130/80 mm Hg for most patients), it is no longer practical to study single drug or device-based interventions in isolation because this low target BP will not be attainable with this strategy in most patients.

Renal denervation can be accomplished using radiofrequency (RF), ultrasound, or injection of dehydrated alcohol into renal arteries. Presumably, this is a one-time procedure based on the multi-year follow-up data currently available. Renal denervation using a rigorous sham-controlled study design has been shown to lower BP, in drug naïve patients with hypertension as well as in those taking antihypertensive drug therapy, on par with a single antihypertensive drug.² The procedure takes approximately one hour via femoral artery access.

10 Key Take-Aways about Renal Denervation for Lowering BP

The scientific statement is well-written, comprehensive, and does a commendable job of rationally interpreting the clinical trial and registry data while contextualizing them with appropriate interpretational caveats.

1. Renal denervation, using radiofrequency and ultrasound, appears to be as effective as a single antihypertensive agent for lowering BP; most RDN-treated patients (24%) do not achieve daytime or home BP < 135/85 mm Hg though more do than the sham-controls (12%). RDN randomized sham-controlled trials were typically 2 – 3 months in duration.
2. Renal denervation results in durable BP lowering over multi-year follow-up. However, some patients experience an initial BP response to RDN then, over time, experience a rise in BP back toward baseline. RDN also appears safe with no difference in major adverse events or changes in kidney function between RDN and sham-controls. There does not appear to be an increment in the incidence of renal artery stenosis post-RDN. Iodinated contrast is used in this procedure.
3. Accurately determining the RDN BP lowering effect in drug-treated hypertensives has been difficult because of the large reductions in BP amongst the sham controls that appears to be driven by differential changes between the RDN and sham-control groups in post-randomization medication adherence.
4. The FDA approval of the Symplicity Spyral System (Medtronic; radiofrequency) and the Paradise System (Recor; ultrasound) is for drug-treated patients with hypertension with inadequate control of BP.
5. Institutions performing RDN should assemble multi-disciplinary teams inclusive of practitioners with specialty training/expertise in the field (hypertension specialists and proceduralists) to screen patients, make appropriate referrals, and manage them post-procedure.
6. There should be careful screening of potentially eligible patients to demonstrate both office and home BP elevations along with screening for potentially treatable secondary causes of hypertension – especially primary aldosteronism (PA) and sleep apnea.
7. Amongst potentially eligible drug-treated hypertensives, priority should be given to those at the highest CVD risk who experience the greatest (absolute) benefit from BP lowering. Examples of high-risk patients include those with resistant hypertension, known CVD-renal disease and subclinical target-organ injury (left ventricular hypertrophy). Other groups of patients likely to benefit from RDN include those with medication intolerance(s), or those who are unwilling or incapable of adequately adhering to their prescribed medication regimen resulting in inadequate BP control.
8. Practitioners should engage in shared decision-making with their patients regarding their interest in and suitability for RDN. Most drug-treated hypertensives (70%) do not express an interest in undergoing RDN for BP control; however, predictors of preferring RDN included younger age, male sex, higher home or office BP readings, being prescribed more antihypertensive medication, CV morbidities, experiencing medication side effects and poor medication adherence. Also, patients with hypertension appear to have overly optimistic expectations of the RDN procedure highlighting the need for practitioners to help align their expectations with likely outcomes.
9. Approximately 60 - 70% of ultrasound RDN-treated patients manifest a daytime or office SBP response > 5 mm Hg. Predictors of the magnitude of BP response post-RDN intervention are few. Predictors of greater response include higher baseline BP and orthostatic hypertension. Arterial stiffness, the presence of renal artery anatomic variants, and lower pre-RDN BP predicts a lesser BP response to RDN.
10. RDN catheters have improved over time which has allowed more complete renal denervation by treating not only the main renal artery but also its branches and accessory arteries. Renal artery anatomy is an important consideration in patient suitability for RDN as those with significant renal artery stenosis, renal artery aneurysms, kidney tumors and renal artery branches that cannot be cannulated by contemporary catheters should not undergo RDN.

Limitations

Though the BP lowering efficacy has been established and appears to be durable – at least for several years, there are patients who experience initial BP lowering followed by a rise in BP back toward baseline. Animal models have shown regrowth of renal sympathetic nerves post-RDN. However, it is unclear if this phenomenon occurs in humans, and if it does, whether it substantively contributes to the post-RDN rise in BP. The scientific statement highlights as well patient subgroups where there are limited data, even over the short-term for RDN. Patients with eGFR < 40 ml/min/1.73 m² as well as patients with stage 1 hypertension have been infrequently included in RDN randomized sham-controlled trials to date. Also, the optimal pharmacological drug regimen for patients undergoing RDN is presently not known.

An important limitation that impacts practitioner discussions with patients and the shared decision-making process is the lesser confidence in accurately estimating the RDN-treatment effect in drug-treated patients with uncontrolled BP. Direct measurement of drug metabolites in one RDN sham-controlled trial³ showed that, amongst US participants only, that post-randomization increases in drug exposure occurred more in the sham-control group than RDN and that reductions in drug exposure occurred more in the RDN than sham-control group; such changes confound the unbiased estimation of the true RDN BP lowering effect and bias between group differences toward the null. This will also impact our understanding of long-term BP trends since this is dependent not only on the durability of the sympathetic denervation but also on the adherence to/changes in the pharmacological drug regimen. Future validation of biomarkers, other than BP that are more specific to the physiological interruption of renal sympathetic nervous system tone, are needed.

Implications for Patient Identification, Evaluation and Treatment

This promising new device-based technology will almost assuredly force transdisciplinary collaboration between hypertension specialist, proceduralist, and primary care physicians to make determinations regarding the suitability of drug-treated patients with hypertension but uncontrolled BP. A model for this collaboration already exists in the Oncology world – tumor boards. These newly created transdisciplinary teams will be charged with ensuring the comprehensive evaluation of patients prior to giving their approval for this procedure. In the absence of an authoritative expert consensus recommendation to guide the pre-RDN patient evaluation, the components of this evaluation will likely be dictated by third party payors. The identification of secondary causes of hypertension, per se, will not always preclude such patients from undergoing RDN. For example, RDN has been shown to reduce the apnea-hypopnea index in hypertensive patients with sleep apnea.⁴ Furthermore, there is evidence from animal models of chronic obstructive sleep apnea that RDN reduces atrial fibrillation via depressing augmented sympathetic nervous system activity.⁵ Primary aldosteronism (PA) is more common than it was previously thought to be⁶ and is highly prevalent in patients with resistant hypertension and sleep apnea.⁷ The majority of patients with proven PA will not be candidates for unilateral adrenalectomy because they have bilateral adrenal hypersecretion of aldosterone. These PA patients will receive medical management for PA and, like sleep apnea patients, will be viable candidates for RDN. Patient acceptance of RDN, despite the fact that it appears to be a one-time (though invasive) procedure appears to be limited.