Author + information
- Received February 21, 2020
- Revision received April 1, 2020
- Accepted April 13, 2020
- Published online June 8, 2020.
- Felix Mahfoud, MDa,∗ (, )@FelixMahfoud,
- Giuseppe Mancia, MD, PhDb,
- Roland Schmieder, MDc,
- Krzysztof Narkiewicz, MD, PhDd,
- Luis Ruilope, MD, PhDe,
- Markus Schlaich, MDf,
- Robert Whitbourn, MDg,
- Andreas Zirlik, MDh,
- Thomas Zeller, MDi,
- Philipp Stawowy, MDj,
- Sidney A. Cohen, MD, PhDk,l,
- Martin Fahy, MSk and
- Michael Böhm, MDa
- aDepartment of Internal Medicine, Saarland University Hospital, Homburg/Saar, Germany
- bUniversity of Milano-Bicocca and Policlinico di Monza, Monza, Italy
- cDepartment of Nephrology and Hypertension, University Hospital Erlangen, Erlangen, Germany
- dDepartment of Hypertension and Diabetology, Medical University of Gdansk, Gdansk, Poland
- eDepartment of Cardiovascular Risk, Hospital Universitario 12 de Octubre and CIBERCV and School of Doctoral Studies and Research, Universidad Europea de Madrid, Madrid, Spain
- fDepartment of Medicine, Dobney Hypertension Centre, School of Medicine–Royal Perth Hospital Unit, The University of Western Australia, Perth, Western Australia, Australia
- gDepartment of Cardiology, Saint Vincent’s Hospital, Melbourne, Victoria, Australia
- hDepartment of Cardiology, Medizinische Universität Graz, Graz, Austria
- iDepartment of Angiology, Universitäts-Herzzentrum Freiburg, Bad Krozingen, Germany
- jDepartment of Cardiology, Deutsches Herzzentrum Berlin, Berlin, Germany
- kCoronary and Structural Heart Division, Medtronic PLC, Santa Rosa, California
- lDepartment of Cardiology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania
- ↵∗Address for correspondence:
Prof. Dr. Felix Mahfoud, Saarland University Hospital, Department of Cardiology, Angiology and Intensive Care Medicine, Kirrberger Straße 100, IMED, Homburg/Saar 66421, Germany.
Background Renal denervation (RDN) is under investigation for treatment of uncontrolled hypertension and might represent an attractive treatment for patients with high cardiovascular (CV) risk. It is important to determine whether baseline CV risk affects the efficacy of RDN.
Objectives The purpose of this study was to assess blood pressure (BP) reduction and event rates after RDN in patients with various comorbidities, testing the hypothesis that RDN is effective and durable in these high-risk populations.
Methods BP reduction and adverse events over 3 years were evaluated for several high-risk subgroups in the GSR (Global proSpective registrY for syMPathetic renaL denervatIon in seleCted IndicatIons Through 3 Years Registry), an international registry of RDN in patients with uncontrolled hypertension (n = 2,652). Comparisons were made for patients age ≥65 years versus age <65 years, with versus without isolated systolic hypertension, with versus without atrial fibrillation, and with versus without diabetes mellitus. Baseline cardiovascular risk was estimated using the American Heart Association (AHA)/American College of Cardiology (ACC) atherosclerosis cardiovascular disease (ASCVD) risk score.
Results Reduction in 24-h systolic BP at 3 years was −8.9 ± 20.1 mm Hg for the overall cohort, and for high-risk subgroups, BP reduction was −10.4 ± 21.0 mm Hg for resistant hypertension, −8.7 ± 17.4 mm Hg in patients age ≥65 years, −10.2 ± 17.9 mm Hg in patients with diabetes, −8.6 ± 18.7 mm Hg in isolated systolic hypertension, −10.1 ± 20.3 mm Hg in chronic kidney disease, and −10.0 ± 19.1 mm Hg in atrial fibrillation (p < 0.0001 compared with baseline for all). BP reduction in patients with measurements at 6, 12, 24, and 36 months showed similar reductions in office and 24-h BP for patients with varying baseline ASCVD risk scores, which was sustained to 3 years. Adverse event rates at 3 years were higher for patients with higher baseline CV risk.
Conclusions BP reduction after RDN was similar for patients with varying high-risk comorbidities and across the range of ASCVD risk scores. The impact of baseline risk on clinical event reduction by RDN-induced BP changes could be evaluated in further studies. (Global proSpective registrY for syMPathetic renaL denervatIon in seleCted IndicatIons Through 3 Years Registry; NCT01534299)
Three recent sham-controlled trials verified the short-term (2 to 6 months) safety and efficacy of renal denervation (RDN) in patients with uncontrolled hypertension and relatively few comorbidities (1–3). However, whether patients with comorbidities associated with increased sympathetic activity or with overall higher cardiovascular risk have a differing blood pressure (BP)–lowering response following RDN is unknown. It is also uncertain whether the durability of the BP-lowering effect of RDN could be limited due to disease progression or differences in response in these subpopulations (4). Finally, the rates of clinical events as well as longer-term safety of RDN are not well described.
The GSR (Global proSpective registrY for syMPathetic renaL denervatIon in seleCted IndicatIons Through 3 Years Registry) is an ongoing, multicenter, international single-arm trial with planned enrollment of 3,000 patients with varying cardiovascular (CV) risk who are followed up to 3 years after RDN. We evaluated short- and long-term BP reduction, clinical events, and adverse event rates after RDN in patients with various comorbidities and baseline CV risk to test the hypothesis of whether RDN is effective and durable in these high-risk populations.
The design of the GSR (NCT01534299) has been previously published (5). In this international, prospective, single-arm registry, patients are enrolled with uncontrolled hypertension and/or conditions associated with sympathetic nervous system activation. Uncontrolled hypertension was defined as BP above recommended levels (regardless of therapy) according to published local guidelines at the time of enrollment. Sympathetic nervous system activation was defined as conditions associated with increased sympathetic nervous system activity, including diabetes, congestive heart failure, chronic kidney disease, obstructive sleep apnea, or arrhythmias. The study was approved by the institutional review board or ethics committee at each enrolling site, and the trial adhered to the Declaration of Helsinki.
All patients are treated with the Symplicity RDN system (Medtronic, Santa Rosa, California) using either the Symplicity Flex or Symplicity Spyral catheter. The primary objective of the study was to document long-term safety and effectiveness of RDN in a real-world patient population. Follow-up is recommended for 3 years post-RDN. Key endpoints measured in this prospective registry include BP measurements and changes in medications. All protocol-defined safety events were adjudicated by an independent clinical events committee (Cardiovascular Research Foundation, New York, New York). Selective monitoring was applied to ensure the quality of the data.
Post hoc analyses were completed for various high-risk subgroups, including: elderly patients (age ≥65 years) and patients with atrial fibrillation (AF), diabetes mellitus (DM) type II, severe treatment resistant hypertension (office systolic blood pressure [SBP] >150 mm Hg despite prescription of ≥3 antihypertensive medications), chronic kidney disease (estimated glomerular filtration rate <60 ml/min/1.73 m2), and isolated systolic hypertension (ISH) (baseline office SBP ≥140 mm Hg and diastolic BP <90 mm Hg). Baseline atherosclerotic cardiovascular disease (ASCVD) risk scores were calculated for patients with available office SBP measurements, antihypertensive medications, and cholesterol measurements, as well as diabetic and smoking status (6). Efficacy was assessed by serial office and, where available, ambulatory BP measurements over the 3 years of follow-up following RDN.
Continuous variables were presented as mean ± SD. Between-group differences in continuous variables were tested using analysis of covariance (ANCOVA), adjusting for baseline BP. Within-group differences in continuous variables from baseline to follow-up were tested using paired t-tests. To account for multiple comparisons over time points, the significance level was determined by the Bonferroni method as indicated in the table and figure footnotes. Categorical variables were presented as counts and percentages and were compared between groups using the Fisher exact test for binary variables and chi-square test for multilevel categorical variables. Some BP measurements at follow-up were adjusted for baseline BP using ANCOVA analyses. These measurements were not adjusted for differences in other baseline covariates. Analyses were performed according to the intention-to-treat principle. BP reduction and adverse events were compared for patients with baseline ASCVD risk scores <10%, ≥10% and <20%, and ≥20%. Changes in BP over time and adverse event rates were compared for the following subgroups: AF versus no AF, age ≥65 years versus age <65 years, DM versus no DM, and ISH versus non-ISH patients. A 2-tailed p value <0.05 was considered statistically significant, unless otherwise indicated. Analyses were performed using SAS version 9.4 (SAS Institute, Cary, North Carolina).
As of March 2019, there were 2,652 patients enrolled at 196 centers in 45 countries, with 2,466 patients reaching 3-year follow-up at the time of this report. A total of 99.8% had a history of hypertension. Median follow-up was close to 3 years. Reductions in office and 24-h SBP for the entire population were sustained from 6 to 36 months post-procedure (Table 1). Baseline ASCVD risk scores were calculated for 1,485 patients (56% of total) primarily due to missing cholesterol measurements. The median ASCVD risk score was 19.8% (Q1, Q3: 9%, 37%). BP reduction in patients with measurements at 6, 12, 24, and 36 months showed similar reductions in office and 24-h SBP for differing baseline ASCVD risk scores (Table 2).
BP reduction in several high-risk cohorts was consistent and sustained across the groups from 6 months to 3 years post-RDN (Table 3). In addition, SBP reduction was compared between subgroups and similar decreases in office (Figure 1) and 24-h SBP (Figure 2) were observed.
Patients with highest baseline ASCVD risk scores (≥20%) had higher 3-year rates of death (8.4%), CV death (4.5%), and hospitalization for new-onset heart failure (5.3%) or AF (6.3%) compared with patients with lower risk scores (Table 4). Adverse events at 3 years were similar between subgroups (Table 5), although several exceptions to this are noted. For patients with versus without DM, there was a higher rate of myocardial infarction (4.0% vs. 1.6%; p = 0.002), end-stage renal disease (2.8% vs. 1.0%; p = 0.005), and elevated creatinine levels (2.4% vs. 0.8%; p = 0.007). Death and CV death rates at 3 years were higher for patients age ≥65 years and patients with DM. Patients with versus without AF had higher rates of death (9.2% vs. 3.1%; p < 0.001). ISH patients had a higher rate of CV death than non-ISH patients (4.0% vs. 2.2%; p = 0.04).
Identification of patient cohorts that may derive specific benefit from RDN is important in the context of wider implementation of this therapy in clinical practice and requires balancing of the potential procedural risks against the expected benefits from BP lowering and other RDN-induced effects. Patients at high CV risk represent a potential target cohort. This post hoc analysis from the largest real-world registry of RDN demonstrated clinically and statistically significant reductions in office and 24-h BP in the entire cohort that were maintained out to at least 3 years following the procedure with low adverse event rates (Central Illustration). The number of prescribed antihypertensive medications for all patients differed slightly from baseline to 3 years (4.56 ± 1.36; n = 2,621 vs. 4.39 ± 1.45; n = 2,449; p < 0.001), but changes were not clinically significant, as opposed to what was reported in a smaller RDN registry (7). It is unlikely that changes in medication contributed to the observed decrease in BP, although adherence measurements were not performed routinely. The baseline adjusted BP-lowering effects were similar across subgroups. BP reductions were not dependent on baseline CV risk, although, expectedly, more adverse clinical events occurred in higher CV-risk subgroups. Overall, these results indicate a safe and durable BP response to RDN in a wide range of patients with uncontrolled BP by conventional drug therapy, including those at high CV risk.
The GSR includes a large population of patients over the age of 65 years (n = 1,059) compared with the recent sham controlled trials of RDN (1–3). We observed office and 24-h SBP reductions in patients age ≥65 years that were sustained to 3 years. Interestingly, the BP reductions were similar for patients above compared with below age 65 years. These results are consistent with a previous report (7) also showing significant BP reductions at 6 months in patients of age ≥75 years.
A post hoc analysis from previous trials and a single-center study both reported less BP reduction for ISH patients after RDN compared with patients with combined systolic-diastolic hypertension (8,9). However, decreases in SBP from our analysis were not different between patients with ISH compared with those without. The present analysis encompasses more patients with longer follow-up and also adjusts for differences in baseline SBP. This is supported by a recent subanalysis of the prospective, randomized, uncontrolled RADIOSOUND (Randomized Comparison of Ultrasound Versus Radiofrequency Denervation in Patients With Therapy Resistant Hypertension) trial, which showed no difference in the adjusted drop in 24-h mean SBP in ISH patients compared with combined hypertension patients (10). Further randomized controlled trials in patients with ISH are warranted to clarify the response to RDN in this high-risk cohort (8).
Increased sympathetic activity is implicated in both the initiation and maintenance of atrial tachyarrhythmias including AF (11). Most previous trials of RDN, including SPYRAL HTN-OFF MED and SPYRAL HTN-ON MED, included relatively small proportions of patients with AF history (1,3). In the present analysis, office and 24-h SBP reductions were similar for patients with versus without AF. Interestingly, recent reports including prospective randomized trials in patients with hypertensive heart disease and in patients with both AF and uncontrolled hypertension have shown significant reductions in AF burden following RDN alone (12) and in combination with pulmonary vein isolation (PVI) (13). Ongoing prospective randomized trials are further evaluating the role of RDN in patients indicated for PVI with a history of hypertension (NCT02064764 and NCT02115100).
Patients with higher baseline CV risk, regardless of comorbidity, may derive particular benefit from BP reduction. Indeed, the current American Heart Association/American College of Cardiology guidelines for BP control stratify recommended drug therapy for patients with stage 1 hypertension based on ASCVD risk >10% (14). The GSR subgroups stratified by baseline CV risk score were not associated with a different BP response, including patients with risk above or below 10% (Table 2). A previous meta-analysis also found that the estimated number of avoidable CV events increased with higher baseline CV risk and more pronounced BP reduction (15). Therefore, patients with high CV risk might particularly benefit from RDN therapy, which may also extend to cost effectiveness.
Renal nerves have been shown to anatomically recover following ablation in some anatomical models (16). However, functional recovery of nerves following radiofrequency (RF) RDN in animal and humans is uncertain (4). A recent preclinical study in sheep with hypertensive chronic kidney disease showed regrowth of renal nerves and return of function at 30 months following RF-RDN, but levels were only partially restored to levels of intact, suggesting that RDN lowers BP in the long-term and is renoprotective and cardioprotective as a result of lesser nerve regrowth in chronic kidney disease (4). Therefore, clinical evidence of the long-term durability of RDN is vital to determine its potential impact on hypertension control rates (17). Long-term renal function and durability of BP reduction following RDN was previously reported for GSR, the SYMPLICITY HTN-1 and HTN-2 trials, and a smaller registry from Sweden (18–20). To date, GSR is the largest RDN registry with 3-year results corroborating prior evidence of the durability of the RDN procedure.
Long-term safety following the RDN procedure remains of utmost importance for multiple stakeholders, including patients, clinicians, regulators, and payers. Short-term reports of adverse events following RDN with both RF and ultrasound-based devices have been encouraging (1–3). Likewise, a previous report from GSR that focused on patients treated with the first-generation single-electrode RF device showed favorable short- and long-term safety with adverse events proportionate to the basal risk of the population (20). Current results report a low 3-year rate of 0.3% for new renal artery stenosis >70%, and previous work suggests the 3-year rate of natural renal artery disease progression is up to 18% in renal arteries categorized as normal at baseline (21). Moreover, the present results show that clinical event rates following RDN are expectedly related to the baseline comorbidity profile. Results from a meta-analysis suggest that a 10-mm Hg reduction in office SBP leads to a 13% reduction in all-cause mortality (22), and mean office SBP reduction at 3 years in GSR was 16.5 mm Hg. Therefore, because the risks associated with RDN appear to be reasonably low, the potential to reduce clinical events by RDN treatment to lower BP independent of BP drug therapy remains high.
The current report includes a post hoc analysis from a large, prospective, single-arm, open-label, real-world registry. As often observed in registries, not all patients were available for 3-year follow-up, and no control group was available for comparison. However, the overall large number of patients available and the extended duration of follow-up clearly demonstrated a persistent BP reduction over 3 years. Baseline BP has been reported to be a predictor of BP drop, and several of the comparison subgroups had different baseline SBP (23). However, these differences were appropriately adjusted using ANCOVA (24), and the benefit in BP reduction was consistent. CV risk scores could not be recalculated at follow-up, because serum cholesterol measurement was not mandatory. However, because SBP is a key determinant of risk, one might assume that calculated risk improved in this group.
BP reduction after RDN was similar in patients with and without baseline conditions associated with increased sympathetic activity and irrespective of ASCVD risk. The reduction in BP was sustained to 3 years, demonstrating the durability of BP reduction by RDN across various subgroups, particularly in patients with high CV risk. Rates of new-onset end-stage renal disease and elevation in serum creatinine levels were very low in patients at high and low CV risk. Clinical events increased with increasing ASCVD risk score, and elevated rates were also seen in patients with AF and diabetes, identifying these subgroups who might derive even greater clinical benefit from improved BP control using RDN.
COMPETENCY IN PATIENT CARE AND PROCEDURAL SKILLS: In a large global registry, renal denervation resulted in similar reductions of blood pressure in hypertensive patients across a wide range of cardiovascular risk.
TRANSLATIONAL OUTLOOK: Future studies should assess the efficacy of renal denervation to prevent major adverse cardiovascular events in patients with isolated systolic and other specified forms of hypertension.
The Global SYMPLICITY Registry is funded by Medtronic. Prof. Mahfoud is supported by Deutsche Gesellschaft für Kardiologie and Deutsche Forschungsgemeinschaft; and has received speaker honoraria from Medtronic and ReCor. Prof. Mancia has received speaker fees from Boehringer Ingelheim, Ferrer, Gedeon Richter, Medtronic Vascular, Menarini, Merck Healthcare KGaA, Neopharmed-Gentili, Novartis, Recordati, Sanofi, and Servier. Prof. Schmieder has received speaker and consulting honoraria from Medtronic, ReCor, Ablative Solutions, and Rox Medical. Prof. Narkiewicz has received speaker honoraria from Adamed, Berlin-Chemie/Menarini, Egis, Gedeon Richter, Krka, Polpharma, Sandoz, and Servier; and has received honoraria or consultation fees from Medtronic, Servier, Krka, Berlin-Chemie/Menarini, Egis, Sandoz, Idorsia, Polpharma, and Gedeon Richter. Prof. Ruilope has served as an advisor/speaker for Medtronic. Prof. Schlaich is supported by an NHMRC Senior Research Fellowship; and has received consulting fees and/or travel and research support from Medtronic, Abbott, Novartis, Servier, Pfizer, and Boehringer Ingelheim. Prof. Zeller has received honoraria from Abbott Vascular, B. Braun, Biotronik, Boston Scientific, Cook Medical, Gore & Associates, Medtronic, Philips-Spectranetics, TriReme, Veryan, and Shockwave; has received consulting fees from Boston Scientific, Cook Medical, Gore & Associates, Medtronic, Spectranetics, Veryan, Intact Vascular, MedAlliance, and Vesper Medical; and owns stock in QT Medical. Prof. Stawowy has received consultancy and lecture honoraria from Amgen, Novartis, San, Bristol-Myers Squibb/Pfizer, Daiichi-Sankyo, Bayer, Boehringer Ingelheim, BerlinChemie, B. Braun, Springer Nature, Medtronic, and AstraZeneca. Dr. Cohen is an employee of and owns stock in Medtronic. Mr. Fahy is an employee of Medtronic. Prof. Böhm is supported by the Deutsche Forschungsgemeinschaft (DFG, TRR-219, S-01, M-03, M-05); and has received personal fees from Amgen, AstraZeneca, Bayer, Boehringer Ingelheim, Bristol-Myers Squibb, Servier, Medtronic, Vifor, Novartis, and Abbott. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
The authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’ institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information, visit the JACC author instructions page.
- Abbreviations and Acronyms
- atrial fibrillation
- atherosclerotic cardiovascular disease
- blood pressure
- diabetes mellitus
- isolated systolic hypertension
- renal denervation
- systolic blood pressure
- Received February 21, 2020.
- Revision received April 1, 2020.
- Accepted April 13, 2020.
- 2020 The Authors
- Townsend R.R.,
- Mahfoud F.,
- Kandzari D.E.,
- et al.
- Azizi M.,
- Schmieder R.E.,
- Mahfoud F.,
- et al.
- Singh R.R.,
- McArdle Z.M.,
- Iudica M.,
- et al.
- Stone N.J.,
- Robinson J.G.,
- Lichtenstein A.H.,
- et al.
- Ziegler A.K.,
- Bertog S.,
- Kaltenbach B.,
- et al.
- Mahfoud F.,
- Azizi M.,
- Ewen S.,
- et al.
- Fengler K.,
- Rommel K.-P.,
- Lapusca R.,
- Blaze S.,
- et al.
- Schirmer S.H.,
- Sayed M.M.,
- Reil J.C.,
- et al.
- Ukena C.,
- Becker N.,
- Pavlicek V.,
- et al.
- Steinberg J.S.,
- Shabanov V.,
- Ponomarev D.,
- et al.
- Whelton P.K.,
- Carey R.M.,
- Aronow W.S.,
- et al.
- Beaney T.,
- Schutte A.E.,
- Tomaszewski M.,
- et al.
- Mahfoud F.,
- Böhm M.,
- Schmieder R.,
- et al.
- Caps M.T.,
- Perissinotto C.,
- Zierler R.E.,
- et al.
- Mahfoud F.,
- Böhm M.,
- Edelman E.R.
- Pocock S.J.,
- Bakris G.,
- Bhatt D.L.,
- Brar S.,
- Fahy M.,
- Gersh B.J.