Proper Patient Selection Yields Significant and Sustained Reduction in Systolic Blood Pressure Following Renal Artery Stenting in Patients With Uncontrolled Hypertension: Long‐Term Results From the HERCULES Trial

Abstract

The Safety and Effectiveness Study of the Herculink Elite Renal Stent to Treat Renal Artery Stenosis (HERCULES) trial is a prospective, multicenter trial evaluating the safety, effectiveness, and durability of the RX Herculink Elite renal stent system (Abbott Vascular, Abbott Park, IL) in select patients with atherosclerotic renal artery stenosis and uncontrolled hypertension. A total of 202 patients were enrolled between August 2007 and October 2009. The primary endpoint, 9‐month binary restenosis, was 10.5% determined by core laboratory adjudicated duplex ultrasound and/or angiography. Additional analyses included changes in blood pressure, antihypertensive medications, renal function (RF), major adverse events (MAEs) (death, ipsilateral nephrectomy, and embolic events resulting in kidney damage), and clinically driven target lesion revascularization (CD‐TLR) between baseline and 36 months. Freedom from MAE was 98.5% at 30 days. At 36 months, freedom from death, nephrectomy, and CD‐TLR were 90.1%, 100%, and 91.8%, respectively. After 30 days there were no site‐reported embolic events resulting in kidney damage. The mean baseline systolic blood pressure of 162±18 mm Hg significantly decreased postprocedure and through 36 months (mean systolic blood pressure 141 mm Hg [P<.0001] and 146 mm Hg [P<.0001], respectively). No differences were noted in antihypertensive medications or RF compared with baseline. The HERCULES trial demonstrated sustained clinically and statistically significant reduction in SBP in patients with uncontrolled HTN. Coupled with the low core laboratory–adjudicated in‐stent restenosis, acceptable procedural complication rates (1.5%), and <10% CD‐TLR, the study suggests that there may be a role for renal artery stenting using contemporary stent technology.

Renal artery stenting (RAS) has been demonstrated as a safe and effective means of obtaining and maintaining vessel patency for patients with atherosclerotic renal artery stenosis (ARAS) in single‐arm, prospective, multicenter pivotal trials.1, 2, 3 In addition, RAS has been shown to be effective in controlling blood pressure (BP)1, 2, 3, 4, 5, 6 and improving or stabilizing renal function.4, 5, 6, 7, 8, 9, 10 There is still debate on the clinical efficacy of this procedure as a method of lowering BP and preventing deterioration in ischemic renal function resulting in end‐stage kidney disease despite additional data from registries and series confirming the efficacy of improved BP control and renal function.11, 12, 13, 14, 15, 16 However, two randomized trials published in 2009 suggested that stenting offers no benefit in BP control or preservation of renal function when compared with medical therapy, but these studies had multiple study design issues and therefore did not significantly impact multisocietal guidelines.17, 18 In addition, the outcomes of previous renal artery stent registries demonstrated modest reductions in BP suggesting that the pretreatment BP elevations may have been unrelated to renal artery stenosis. The recently reported results of the Benefits of Medical Therapy Plus Stenting for Renal Atherosclerotic Lesions (CORAL) trial raised additional questions about clinical benefit of RAS as initial treatment for hypertension (HTN) and renal artery stenosis.19 Additional randomized studies20, 21 are underway and hopefully will provide results that confirm when appropriate patients are selected, significant and sustained results can be achieved. The (Safety and Effectiveness Study of the Herculink Elite Renal Stent to Treat Renal Artery Stenosis) HERCULES trial demonstrated that when patients are properly selected, BP is significantly reduced and sustained through 9 months post‐procedure.22 The hypothesis leading to the current study was that if the renal stent used in the initial HERCULES trial maintained renal artery patency over time, then the clinical benefit confirmed in the initial trial should be sustained over time. Accordingly, the 3‐year results are reported herein.

Methods

Study Design and Patients

The HERCULES trial is a prospective, multicenter, single‐arm clinical study of patients with significant renal artery stenosis and uncontrolled HTN treated with a stent that is specifically designed to address the challenges unique to the renal artery (RX Herculink Elite Renal Stent System [Abbott Vascular, Santa Clara, CA]). The typical aorto‐ostial renal lesion requires significant radial support and this is often accomplished at the cost of reduced flexibility and increased profile, compromising delivery from the femoral approach. The balance of radial support, low profile, and flexibility were achieved by the use of a contemporary specialty alloy (L‐605 Cobalt Chromium) in combination with a unique monorail 0.014‐inch guidewire compatible delivery system and a stent fabric consisting of a series of zigzagging rings with multiple links per ring designed to enhance ostial radial support. The study was registered at clinicaltrials.gov (NCT00490841), and individual institutional review board approval was obtained at each site. All patients provided signed informed consent prior to enrollment and the conduct, performance, monitoring, auditing, data recording, and analysis all strictly adhered to Good Clinical Practice standards.23

Eligible patients included those with uncontrolled HTN defined as systolic BP (SBP) ≥140 mm Hg or diastolic BP (DBP) ≥90 mm Hg despite maximal doses of at least 2 antihypertensive agents in appropriate combinations in association with renal artery stenosis ≥60% via angiographic visual estimate and suboptimal percutaneous transluminal balloon renal angioplasty (PTRA) result, defined as one of the following: ≥50% residual stenosis, persistent translesional pressure gradient (mean 10 mm Hg, or peak systolic gradient of 20 mm Hg), flow‐limiting dissection, or thrombolysis in myocardial infarction flow <3. Patients with unilateral or bilateral ARAS were eligible, provided that the lesion was within 10 mm of the aorto‐ostial junction, and the reference vessel diameter was between 4 mm and 7 mm. Patients with lesions representing in‐stent restenosis following a prior endovascular stent revascularization were not eligible for enrollment. Patients with serum creatinine >2.5 mg/dL, recent myocardial infarction, stroke or transient cerebral ischemia, impaired left ventricular ejection fraction (≤25%), ARAS to a solitary functioning kidney, or transplant renal artery stenosis were also ineligible for participation.

Procedure

All enrolled patients received anticoagulation per hospital standard during angioplasty of the renal artery. Selective angiography in at least two views was performed to confirm angiographic eligibility (≥60%) of the stenosis. PTRA was performed with a standard balloon dilatation catheter. Immediately following PTRA, the lesion was assessed for continued eligibility, ie, suboptimal PTRA, and treated with the study device. All patients received aspirin 325 mg orally once daily, and clopidogrel either 75 mg orally once daily for 4 days prior to the procedure, or as a single loading dose of 300 mg orally within 24 hours prior to the procedure. Heparin was used as the procedural anticoagulant agent. Following stent placement, aspirin 325 mg orally once daily was continued for a minimum of 12 months and clopidogrel 75 mg orally once daily for at least 4 weeks.

Clinical Follow‐Up and Assessments

Office BP and renal function measurements were performed following the procedure and at 1, 6, 9, 12, 24, and 36 months using accepted and reproducible techniques for BP assessment. All patients underwent renal artery duplex ultrasonography following the procedure and at 1, 9, 24, and 36 months. An independent vascular ultrasound core laboratory with experience in multicenter vascular device trials (VasCore, Massachusetts General Hospital, Boston, MA) analyzed the ultrasound images using prespecified criteria to determine the degree of stenosis. A renal arteriogram was required at 9 months if the duplex ultrasound was determined to be inadequate by the ultrasound core laboratory or at any point in time at the discretion of the investigator. All procedural and follow‐up angiogram results were assessed by an independent Angiographic Core Laboratory (Beth Israel Deaconess Medical Center, Boston, MA) using quantitative angiographic methods.

Statistical Analyses

The primary endpoint was tested under the hypothesis of superiority of the 9‐month binary restenosis rate after renal artery stenting with the study stent to a performance goal (PG), the details of which are documented in the primary HERCULES publication by Jaff and colleagues.22 A composite safety secondary endpoint included all‐cause mortality, embolization resulting in kidney damage, and ipsilateral nephrectomy at 30 days. The 30‐day safety endpoint was summarized by count, percentage, and 95% confidence interval. Other analyses through 36 months postprocedure included changes in BP; antihypertensive medication intake; renal function (measured by serum creatinine and estimated glomerular filtration rate); and all‐cause death, embolization resulting in kidney damage, ipsilateral nephrectomy, and clinically driven target lesion revascularization (CD‐TLR). Freedom from event rates were estimated using Kaplan‐Meier methodology, with standard errors calculated using the Greenwood formula. Demographics and other baseline characteristic variables were summarized using descriptive statistics including means and standard deviations (SDs) for continuous variables and counts and percentages for binary variables. All data summaries and statistical analyses were performed using SAS for Windows, version 9.1 or higher (SAS Institute Inc, Cary, NC).

Results

Patients

The baseline characteristics of the 202 patients enrolled are detailed in Table 1. An expected high incidence of cardiovascular risk factors and concomitant coronary artery disease is present. Baseline mean±SD SBP and DBP were 162±19/78±12 mm Hg, respectively. A total of 99% of patients required ≥2 antihypertensive agents for BP management, and 71% required ≥3 medications (average, 3.4 medications). One third of treated patients had bilateral ARAS, and the mean baseline percent diameter stenosis was 81.3% as determined by the site (66% as determined by the independent angiographic core laboratory) (Table 2).

Table 1.

Baseline Demographic and Clinical Characteristics

Variable	N=202
Age, y	72.1
Female, % (No.)	62.4 (126/202)
Ethnicity, % (No)
Caucasian	83.7 (169/202)
African American	7.9 (16/202)
Latino‐Hispanic	7.4 (15/202)
Diabetes	45.0 (91/202)
History of tobacco use	56.9 (115/202)
Hypercholesterolemia requiring medication	82.2 (166/202)
Coronary artery disease	67.3 (136/202)
Baseline systolic blood pressure, mm Hg	162.3±18.5
Baseline diastolic blood pressure, mm Hg	77.7±11.5
Serum creatinine, mg/dL	1.2±0.4
Estimated glomerular filtration ratea	58.0±20.8
Antihypertensive medication intake, % (No)
1 medication	0.5 (1/202)
2 medications	28.7 (58/202)
3 medications	32.2 (65/202)
≥4 medications	38.6 (78/202)

^aEstimated glomerular filtration rate = 186 × (S_cr)^−1.154 × (age)^−0.203 × 0.742 (if female) × 1.210 (if African American).

Table 2.

Anatomic Characteristics^a

Variable	N=202/241 Lesions
Preprocedure
De novo lesions	100% (241/241)
Unilateral lesions	67.6% (163/241)
Bilateral lesions	32.4% (78/241)
Site‐reported diameter stenosis, %	81.3±10.0
Core laboratory–reported diameter stenosis, %	65.9±11.4
Core laboratory–reported minimum lumen diameter, mm	1.8±0.7
Core laboratory–reported reference vessel diameter, mm	5.4±1.1
Lesion length, mm	8.5±3.1
Heavy calcification	27.8% (67/241)
Suboptimal PTA outcome
Residual stenosis ≥50%	90.7% (214/236)
10 mm Hg mean gradient or 20 mm Hg peak systolic gradient across the lesion	25.0% (59/236)
Flow‐limiting dissection (NHLBI grade D) or TIMI flow <3	4.7% (11/236)
Postprocedure
Core laboratory–reported diameter stenosis,%	3.4±12.5
Core laboratory–reported minimum lumen diameter, mm	5.1±0.8
Study device success	96.0% (237/247)
Procedure success	99.2% (238/240)
Clinical success	98.0% (197/201)

^aAbbreviations: NHLBI, National Heart, Lung, and Blood Institute; PTA, percutaneous transluminal angioplasty; TIMI, thrombolysis in myocardial infarction. Per lesion.

Analyses

The primary endpoint of restenosis at 9 months following renal artery stenting was 10.5% (22 of 209; upper 95% confidence limit, 14.7%), below the predetermined performance goal of 28.6% established for the HERCULES study based on published literature from pivotal renal trials. The primary objective of the study was met (P<.0001) as documented previously.22 At 36 months, the mean (±SD) SBP and DBP were 146±21/75±11 mm Hg, respectively. There was a statistically significant reduction in SBP compared with preintervention (paired t test, P<.0001) and DBP to a lesser extent (paired t test, P<.02) (Table 3). There was no reduction in the number of antihypertensive medications, and renal function, as measured by estimated glomerular filtration rates as calculated by the Cockcroft‐Gault formula, was unchanged (Table 3). The magnitude of absolute reduction in SBP was related to the severity of baseline systolic HTN prior to intervention. For patients who had a reduction in SBP at 36 months, those with preprocedure SBP ≥180 mm Hg experienced a mean SBP reduction of 46 mm Hg, while those patients whose baseline SBP was between ≥140 mm Hg and <160 mm Hg had a 26 mm Hg reduction, and those with a baseline SBP between ≥160 mm Hg and <180 mm Hg had a 27 mm Hg reduction (Table 4). More than 74.3% of patients showed a reduction in SBP at 36 months. There was no statistically significant difference in the magnitude of SBP response among patients treated for bilateral RAS compared with those treated for unilateral RAS (Table 5). The 30‐day procedural safety events included only 1 death, and 2 patients experiencing renal atheromatous embolization had kidney damage. The 30‐day composite safety endpoint rate was 1.5%. Through 36 months, freedom from death was 90.1% (Figure 1) and freedom from nephrectomy was 100%. Beyond 30 days, there were no additional procedure‐related embolic events resulting in kidney damage. None of the 18 total deaths were reported as being related to the study procedure or device. Two of the 18 deaths were of unknown causes. At 12 and 36 months, freedom from CD‐TLR was 94.7% and 91.8%, respectively, demonstrating the durability of the study stent system (Figure 2). The study stent system is currently the only cobalt chromium stent approved in the United States for use in the renal arteries and the unique design characteristics may have contributed to long‐term durability compared with the stainless steel alternatives.24

Table 3.

Blood Pressure and Medications

Variable	Baseline	1 Month	9 Months	2 Years	3 Years	P Value
SBP, mm Hg	162±19	145±21	145±21	144±23	146±21	<.0001a
DBP, mm Hg	78±12	76±11	75±12	74±12	75±11	.02a
eGFR, mL/min per 1.73 m2	58±21	59±21	57±23	58±25	57±23	.53a
≥3 antihypertensive medications, %	71	68	66	70	68	.83b
ACE inhibitor or ARB, %	76	76	76	74	70	.66b
Diuretics, %	66	64	60	60	57	.46b

Abbreviations: ACE, angiotensin‐converting enzyme; ARB, angiotensin receptor blocker; DBP, diastolic blood pressure; eGFR, glomerular filtration rate; SBP, systolic blood pressure.

^a P value: baseline compared with 3 years.

^b P value: baseline compared with 1 month, 9 months, 2 years, and 3 years.

Table 4.

SBP Change by ≥1 Category at 3 Years Compared With Baseline

	Improved by ≥1 Category, % (No.)	Decrease in SBP, mm Hg	3 Years, mm Hg	2 Years, mm Hg	9‐Month SBP, mm Hg
≥180 mm Hga Mean SBP 194±12 (n=39)	78 (25)		148±14	146±20	146±21
≥160  mm Hg, <180 mm Hgb Mean SBP 167±6 (n=49)	77 (27)		140±13	139±13	136±13
≥140  mm Hg, <160 mm Hgc Mean SBP 150±5 (n=113)	46 (37)		124±13	124±9	127±9

One patient had a systolic blood pressure (SBP) value <140 mm Hg at baseline. SBPs missing at 3 years: ^aunknown (n=7), ^bunknown (n=14), ^cunknown (n=32).

Table 5.

Summary of eGFR^a—Bilateral vs Unilateral Artery Treatment

	Bilateral (n=39)	Unilateral (n=163)
Preprocedure, mean±SD (No.)	56.5±20.5 (39)	58.3±20.9 (161)
Postprocedure, mean±SD (No.)	61.6±20.0 (36)	63.8±24.0 (149)
9 months, mean±SD (No.)	56.1±16.1 (30)	57.2±23.8 (135)
24 months, mean±SD (No.)	55.7±18.7 (24)	58.4±26.2 (122)
36 months, mean±SD (No.)	56.1±16.8 (23)	57.4±23.7 (115)

Abbreviations: eGFR, estimated glomerular filtration; SD, standard deviation. ^aeGFR = 186 × (S_cr)^−1.154 × (age)^−0.203 × 0.742 (if female)× 1.210 (if African American).

Figure 1.

Kaplan‐Meier survival curve: freedom from death through 3 years. Note: Patients at risk gives the number of patients at risk for an event at the start of the interval, while patients censored and number of events are the incremental counts of patients censored or with events during the interval. The intervals are denoted as half‐open bracket expression, where the start of interval “is exclusive and the end of the interval” is inclusive.

Figure 2.

Kaplan‐Meier survival curve: freedom from clinically driven target lesion revascularization through 3 years. Note: Lesions at risk gives the number of lesions at risk for an event at the start of the interval, while lesions censored and number of events are the incremental counts of lesions censored or with events during the interval. The intervals are denoted as half‐open bracket expression, where the start of interval “is exclusive and the end of the interval” is inclusive.

Discussion

Approximately 80% of patients with renal artery stenosis have atherosclerosis. This condition is frequently associated with secondary HTN resistant to medical therapy and, in many instances, renal functional impairment.25 This is based largely on the physiology of ARAS‐related HTN. The HTN in unilateral ARAS is caused by the hyperactivity of the renin‐angiotensin‐aldosterone system (RAAS) because the increased BP causes pressure diuresis from the contralateral healthy kidney, leading to plasma volume contraction and further stimulation of RAAS. When the obstruction is released, the activity of RAAS decreases, which leads to lowering of BP alone or, in most cases, with concomitant medical therapy.26 In patients with bilateral ARAS, there is no healthy kidney to respond to pressure diuresis and these patients have HTN that is volume dependent. The release of these obstructions leads to more efficient volume management mainly through the addition or continuation of optimal medical therapy. The key feature in both unilateral and bilateral ARAS is that the stenosis is physiologically significant.

PTRA with stenting has been shown to be effective in reducing BP and preventing or attenuating deterioration of renal function. However, controversy still exists as to whether this approach to the treatment of resistant HTN is better than medical therapy alone.27, 28, 29 The controversy was amplified when two large trials, the Angioplasty and Stenting for Renal Artery Lesions (ASTRAL)18 and the Stent Placement and Blood Pressure and Lipid‐Lowering for the Prevention of Progression of Renal Dysfunction Caused by Atherosclerotic Ostial Stenosis of the Renal Artery (STAR) studies,17 demonstrated that RAS was not superior to medical therapy in reducing BP or preserving renal function. While both trials were important, there were some basic design and execution flaws that are relevant for discussion. The most significant and clinically important contrast to ASTRAL provided by HERCULES is the difference in safety. The rates in the ASTRAL findings were very high and may not reflect the true safety of the procedure when in the hands of experienced operators using contemporary techniques and equipment. The composite 30‐day safety endpoint rate of 1.5% in HERCULES demonstrates that the procedure is safe.

The HERCULES trial included only patients with hemodynamically significant renal artery stenosis as reflected by the mean visual percent stenosis of 81% and core laboratory quantitative measurement of 66%. The difference in operator‐estimated and core laboratory‐adjudicated findings reported herein confirms how visual estimation of stenosis overestimates lesion severity and raises further concerns about suboptimal patient selection in ASTRAL and STAR where no core laboratory was used. The long‐awaited prospective, multicenter, randomized trial of renal artery stenting with optimal medical therapy vs optimal medical therapy alone, the CORAL trial, enrolled 947 patients. The primary endpoint was significantly different from that in HERCULES, representing a composite of cardiovascular and renal death, myocardial infarction, stroke, hospitalization for congestive heart failure, progressive renal insufficiency, or the need for dialysis support. At a median of 43 months, there was no advantage of stent revascularization to medical therapy in meeting the primary endpoint. However, it is important to note that in CORAL, all patients received optimal medical therapy with only one group receiving a stent in addition, not in place of, optimal medical therapy. Interestingly, there was a small but statistically significant advantage in reduction of SBP in the stent‐treated patients, but this did not result in a primary endpoint benefit, even if the baseline SBP was >160 mm Hg.19 The CORAL results represent the strongest data to date on renal stenting and ARAS; therefore, any critical comparison between CORAL and HERCULES would be unwarranted. There are some interesting differences between the studies as they apply to the most severely hypertensive patients in the respective studies. The patients selected for HERCULES were more hypertensive and taking more medication on average than the patients in CORAL. The most relevant difference between CORAL and HERCULES was the inclusion of patients with mild HTN. This limitation is particularly relevant since a recent meta‐analysis of 901 patients enrolled in 5 US Food and Drug Administration investigational device exemption trials found that the only predictor of clinical benefit following renal artery stent revascularization was significant HTN at baseline.30

There are some limitations of HERCULES, including the single‐arm, nonrandomized design, making it difficult to exclude the Hawthorne effect as a contributor to the reduction in BP. In addition, although the definition of resistant HTN requires treatment with ≥3 antihypertensive medications (including a diuretic), we felt that in association with hemodynamically significant renal artery stenosis and confirmation that the average medication usage was 3.4, our modified inclusion criterion was adequate. Another limitation was that almost 40% of the patients were not receiving a diuretic and the possibility that similar BP results would not have been seen without stenting if more stringent medical treatment requirements were followed cannot be excluded. However, there were no differences seen in SBP or DBP at baseline and through all follow‐up time points for patients who received diuretics vs those who did not. The measurement of renal function is a potential weakness of the current study. The Cockcroft‐Gault formula may not be the best method of estimating renal function because of the potential limited utility in women and potential sensitivity to weight changes. Regardless of the potential weaknesses of the randomized renal stent trials, it is increasingly clear that renal stenting should be offered only in patients who meet evolving strict criteria. These study results mirror previous pivotal trial stent patency and BP response results and confirm that contemporary stent designs are safe and perhaps more effective (range of target lesion revascularization, duplex patency, and major adverse events of other trials compared with ours). The reality of ARAS‐related HTN is that there is frequently coexisting essential HTN that will not be “cured” by intervention. There must also be recognition that the presence of diabetes mellitus, hyperlipidemia, and peripheral arterial disease are all associated with poor response to PTRA.26, 27 Therefore, we believe that optimal treatment of HTN in patients with ARAS involves detailed risk‐benefit analysis in the context of current randomized data and results with contemporary stent designs. Stent revascularization should be considered for true resistant HTN with failure of adequate medical therapy, hemodynamically significant ARAS, and flash pulmonary edema or for preservation of renal function with >75% of renal mass effected by significant arterial disease.

Conclusions

The treatment of secondary HTN caused by ARAS remains controversial. There are several studies demonstrating no benefit of stenting. In the HERCULES trial, proper selection of patients with hemodynamically significant stenosis treated with PTRA and stent is both safe and clinically effective. In the future, trials involving PTRA and stenting must not include patients with controlled BP taking low doses of few medications and/or hemodynamically insignificant renal artery stenosis. Based on the HERCULES trial data, there may still be a role for RAS in appropriately selected patients.

Disclosures

George S. Chrysant, MD: Abbott Vascular (MAB, consulting); Medicines Company (Consulting); St. Jude Medical (Advisory Board, Consulting). Michael R. Jaff, DO: Noncompensated advisor, Abbott Vascular; Board member, VIVA Physicians, a 501 c 3 not‐for‐profit education and research organization. Mark C. Bates, MD: Consultant and patent contributor to BioSensors International, Inc.; Consultant and patent contributor to CeloNova Biosciences, Inc.; Founder and shareholder for Nexeon MedSystems; Consultant for Vascular Dynamics, Inc.; Founder and shareholder Vascular Protection Solutions, LLC. William B. Bachinsky, MD: Abbott Vascular (MAB, Consulting). Jeffrey J. Popma, MD: Abbott Vascular (Research Grants). Ms Omran and Ms Peng are all full‐time employees of Abbott Vascular. The other authors have no conflicts of interest or disclosures.

J Clin Hypertens (Greenwich). 2014;16:497–503. ©2014 Wiley Periodicals, Inc.