Author + information
- Received August 11, 2015
- Revision received November 22, 2015
- Accepted November 23, 2015
- Published online March 28, 2016.
- Tim A. Fischell, MDa,b,∗ (, )
- Adrian Ebner, MDc,
- Santiago Gallo, MDc,
- Fumiaki Ikeno, MDd,
- Laura Minarsch, RTe,
- Félix Vega, VMDf,
- Nicole Haratani, RN, BSNb and
- Vartan E. Ghazarossian, PhDb
- aBorgess Heart Institute, Kalamazoo, Michigan
- bAblative Solutions, Inc., Kalamazoo, Michigan
- cSanatorio Italiano, Asunción, Paraguay
- dStanford University, Stanford, California
- eMMC Medical, Laguna Beach, California
- fPreclinical Consultation, San Francisco, California
- ↵∗Reprint requests and correspondence:
Dr. Tim A. Fischell, Michigan State University, Borgess Heart Institute, 1521 Gull Road, Kalamazoo, Michigan 49048.
Objectives This study evaluated the first clinical use of a new endovascular approach to renal denervation, using chemical neurolysis, via periadventitial infusion of dehydrated alcohol (ethanol) to perform “perivascular” renal artery sympathetic denervation.
Background Renal denervation remains a promising technology for the treatment of hypertension and other disorders.
Methods A novel 3-needle delivery device (Peregrine System Infusion Catheter, Ablative Solutions, Inc., Kalamazoo, Michigan) was introduced into the renal arteries of 18 subjects with refractory hypertension. Microdoses of alcohol were infused bilaterally via the 3 needles into to the adventitial space (0.30 ml/artery, 37 arteries). Renal artery angiography was performed at the time of the procedure and at 6 months (n = 16). The primary safety endpoints were complications associated with the catheter insertion and delivery of the neurolytic agent or any major vascular access complications. The secondary performance endpoint was a reduction in office-based systolic blood pressure at 6 months compared with baseline.
Results Procedural success was achieved in 100% of subjects (N = 18) and arteries (N = 37). There were no study-related adverse clinical events at follow-up. One death of a subject was recorded but determined by the investigator and an independent medical monitor to be non–study related. There were no angiographic observations of renal artery stenosis, aneurysms, or other renal artery abnormalities at 6 months (32 renal arteries). Sixteen of the 18 subjects had a 6-month follow-up. The mean office systolic blood pressure decreased from 175 ± 17 mm Hg to 151 ± 26 mm Hg (−24 mm Hg). There was an average reduction of antihypertensive medications from 3.4 (baseline) to 2.0 per subject at 6 months.
Conclusions Chemical renal denervation using the infusion of very low doses of alcohol directly into the adventitial space appears to be feasible and safe. This approach may be a promising alternative approach to perform catheter-based renal denervation. These results need to be confirmed in larger scale clinical studies.
From the 1930s through the 1950s, surgical sympathectomy was used to treat severe hypertension (1–3). Despite successful lowering of blood pressure (BP) observed with surgical denervation, this technique was abandoned due to relatively high morbidity and mortality and as a result of the development of effective oral antihypertensive medications.
More recently, catheter-based renal sympathetic denervation has been performed using a point-by-point radiofrequency (RF) ablation catheters and ultrasound ablation from within the renal artery (4–11). This technique has been shown to disrupt renal sympathetic nerve activity (4–7), resulting in significant and sustained reduction in BP in patients with severe and medically refractory hypertension (4–14).
Despite encouraging early results, the randomized, sham-controlled, SYMPLICITY HTN-3 (Renal Denervation in Patients With Uncontrolled Hypertension) study failed to demonstrate significant BP lowering compared with the sham (15,16). Although these results might have been unexpected based on the early and reproducible success with this technique, many have now pointed to a lack of adequate sympathetic denervation, a lack of operator experience, as well as complexities related to medication adjustments and compliance in both the treatment and sham groups as possible causes of the trial’s failure (17,18).
Recent studies have also suggested that it may be challenging to create adequate renal sympathetic denervation with RF due to inconsistent circumferential denervation and a lack of adequate depth of nerve injury (19), particularly because the sympathetic nerves may traverse a course at distances of 10 to 12 mm deep to the intimal surface (20,21). There may also be potential safety concerns associated with the use of transmural thermal injury traversing the intimal and medial layers of the renal artery, particularly if a more aggressive ablation strategy is used (22–26).
In an attempt to overcome some of the perceived limitations of “energy-based” approaches, we developed and reported encouraging pre-clinical results using very small volumes of alcohol, a known potent neurolytic agent, delivered precisely and locally to the adventitial space of the renal artery as a means to perform perivascular chemical renal sympathetic denervation (27,28).
The purpose of the current study was to evaluate the safety and feasibility of performing chemical neurolysis via adventitial infusion of microdoses of dehydrated alcohol as a means to perform renal artery sympathetic denervation, in a first-in-human experience, in patients with refractory hypertension.
This was a first-in-human, single-center, prospective feasibility study intended to test the safety of a novel 3-needle-based delivery device (Peregrine System Infusion Catheter, Ablative Solutions, Inc., Kalamazoo, Michigan) to perform chemical renal sympathetic denervation in patients with refractory hypertension, with the use of microdoses (0.3 ml per renal artery) of dehydrated alcohol as the neurolytic agent.
The protocol for this study was reviewed and approved by the Bioethics Committee of the Office of the Superintendent of Health, Paraguay. Nineteen subjects were enrolled after signing the Ethics Committee-approved informed consent form. The inclusion and exclusion criteria are outlined in Tables 1 and 2 and are similar to those used in most RF-based renal artery sympathetic denervation clinical trials at the time of this study (3–6,12). All subjects had to have “resistant hypertension,” defined as an office-based systolic BP reading of >160 mm Hg (>150 mm Hg if the subject had type 2 diabetes) while taking at least 3 antihypertensive medications, including a diuretic agent. There were no inclusion/exclusion criteria in the study related to renal artery length.
Clinical and laboratory assessments were obtained in-hospital and at 1, 3, and 6 months after bilateral renal denervation. The primary safety data included procedure- and device-related adverse events and any post-procedure adverse cardiovascular or vascular complications, deterioration in renal function, or renal artery abnormality at angiographic follow-up at 6 months.
Baseline office-based BP (OBP) was measured at the time of enrollment, pre- and post-procedure, and then at follow-up visits. OBP at each time point was reported as the mean of 3 consecutive BP readings, taken 5 min apart and performed by study participating site nurse or subinvestigator. Ambulatory BP monitoring (ABPM) measurement was not available at the trial site and was not included in this study protocol. Subjects were instructed to maintain their medications at their pre-intervention doses throughout the 6-month follow-up period.
There were challenges in validating medication compliance, particularly after the renal denervation procedure. Medication compliance data were evaluable in 12 of the 16 evaluable subjects at 6-month follow-up. Laboratory measurements, including routine chemistry, blood urea nitrogen, creatinine, estimated glomerular filtration rate (eGFR), liver function tests, complete blood count, partial thromboplastin time, international normalized ratio, and urine analysis, were obtained at baseline and at each follow-up visit. A 12-lead electrocardiogram was obtained at baseline, discharge, and 1-month follow-up. A central university-based (Stanford University) independent qualified angiography reviewer read the angiographic renal artery studies from the time of treatment for all 18 subjects (37 renal arteries) as well as at the 6-month endpoint. A flow diagram for the study is shown in Figure 1.
The renal artery was engaged using a 7-F guiding catheter, introduced via the femoral artery under fluoroscopic guidance. Intravenous unfractionated heparin was given to attain an activated clotting time ≥250 s. Modest sedation per the institution’s standard practice was used in all cases with relatively low doses of fentanyl and midazolam (50 μg and 5 mg, respectively, intravenously). No general anesthesia was used in any subject.
Angiography was performed of the renal arteries before intervention. After the initial angiography, the Peregrine System Infusion Catheter was advanced through the guiding catheter and into the distal half of the renal artery, typically at the midpoint.
The Peregrine System Infusion Catheter is used under fluoroscopic guidance by a single operator with standard endovascular techniques and allows safe and reproducible fluid infusion into the adventitial and periadventitial spaces of a target vessel (Central Illustration, Figures 2 and 3). The three 0.008-inch (220-μm) needles of the Peregrine catheter are contained in a pre-deployment state within the guide tubes. Once deployed, the 3 guide tubes, spaced 120° apart, are moved outward and serve to reproducibly “center” and stabilize the device within the renal artery (Central Illustration, Figures 2 and 3). These atraumatic tubes have radiopaque distal tip markers to clearly define the position of the tubes (Figure 3C). The predictable 360° spread of neurolytic agent (alcohol) has been confirmed in pre-clinical testing (27).
Once the infusion (distal) section of the device was positioned within the target site, the 3 guide tubes were simultaneously deployed against the intimal surface using the advancement mechanism in the control handle (Central Illustration, Figure 3C). The specialized handle was then used to simultaneously advance the three 0.008-inch needles. These microneedles are radiopaque so that they can be easily seen under fluoroscopy (Figure 3D). The needles were advanced to their pre-set depth of 3.5 ± 0.25 mm deep to the intima (i.e., beyond the tip of the guide tube) and into the adventitial space. This allows an infusion depth that approximates the border between the renal artery adventitia and periadventitia and that corresponds to a depth of the middle of the renal sympathetic nerve field, as observed in pressure-fixed human histopathological studies (20,21). Using this technique, it is essentially impossible to infuse the alcohol into the renal artery lumen.
The neurolytic agent (0.3 ml of dehydrated alcohol injection, USP (Akorn, Inc., Lake Forest, Illinois) was then delivered with a 60-s infusion using a 1-ml Luer-lock syringe, connected at infusion port located at the proximal end of the handle. Once the alcohol was infused, the dead space within the catheter was flushed with 0.1 ml of normal saline. Angiography was performed after the withdrawal of the Peregrine catheter.
Independent of the formal protocol, the principal investigator queried the subjects about the level of pain at the time of the infusion and again when the infusion was completed. The outcome of this independent analysis is described in the Results section.
As a safety measure per protocol, in the first 5 subjects, only 1 renal artery was treated (unilateral denervation) during the initial procedure. These first 5 subjects returned at ∼4 weeks after the initial treatment for angiography of the treated renal vessel and then treatment of the contralateral renal artery to complete their bilateral renal denervation (Figure 1). The subsequently treated 13 subjects underwent bilateral denervation performed in 1 procedure. All 18 subjects (23 procedures) were discharged on the morning after the procedure.
For statistical purposes, this feasibility study was viewed as descriptive. The study was not powered to evaluate efficacy endpoints. Data are shown in graphs as mean ± SD.
Demographic data are shown in Table 3. Entry and follow-up BP data are shown in Table 4. Device and procedural success, defined as successful infusion of the alcohol without serious procedural adverse events, was achieved in 100% of patients (N = 18) and vessels (37 of 37 arteries) treated. Procedure time, measured from the advancement of the device into the renal artery, followed by deployment of needles, infusion, flushing, and withdrawal from the guiding catheter, averaged 10 min (range, 2 to 21 min). The mean activated clotting time was 328 ± 38 s (range, 263 to 523 s). One patient was excluded after the initial renal angiography due to a very large diameter renal artery (>8 mm) and was not treated. One subject presented with dual renal arteries on the left with diameters of 4.8 mm and 5.3 mm, both of which were treated. No subject was excluded by virtue of any anatomic variation (e.g., short artery, sharp take off), and no subject had a significant renal artery stenosis before intervention.
There were no cases of perforation, dissection, or significant spasm (>20% diameter stenosis) by visual assessment. In 14 of 37 arteries, the renal angiogram obtained immediately after device withdrawal showed a very small amount of contrast within the microneedle tracks. Angiography performed 1 to 3 min later showed complete resolution of this event in all cases. There was no evidence of this finding or any associated abnormalities observed in these cases at the 6-month angiographic follow-up. There were no device- or intervention-related complications or adverse events. One subject was lost to follow-up at 6 months. This subject had a BP lowering from 174/131 at baseline to 151/101 at the 1-month follow-up.
One subject died 9 weeks post-procedure due to mechanical small bowel obstruction that was treated surgically. Five days later, peritonitis developed, followed by a septic shock and organ failure. The subject went into cardiopulmonary arrest and died. This serious adverse event was determined by the principal investigator and adjudicated by an independent medical monitor to be unrelated to the device or the procedure.
Angiographic follow-up of 32 treated vessels in the 16 subjects available for follow-up at 6 months showed no evidence of renal artery stenosis or any other angiographic abnormalities. Specifically, there were no aneurysms, thrombi, pseudoaneurysms, dissections, and the like and no meaningful change from the baseline angiography in any vessel (Figures 1C and 2F).
There were no adverse nephrotoxic or systemic effects seen at 1-, 3-, or 6-month follow-up by laboratory testing (Figure 4). Overall, the subjects’ serum creatinine, blood urea nitrogen, and electrolytes remained stable over the study period. The mean eGFR (n = 16) at baseline was 66 ± 16 ml/min and was 75 ± 13 ml/min at 6-month follow-up (p = 0.15) (Figure 4).
Finally, at the 6-month follow-up, there was a mean OBP lowering of 24 ± 22 mm Hg (systolic) and 12 ± 9 mm Hg (diastolic) in the 16 subjects who completed the study. The detailed results showing the 16 patients’ data are shown in Table 4. In this study, a reduction in BP was noted despite a reduction or discontinuation of antihypertensive medications in 9 of 12 of the subjects (75%) for whom medication data were evaluable (Tables 4 and 5, Figure 5). No patient for whom medication data were recorded at baseline and 6 months had an increase in antihypertensive medications at follow-up.
In the subjects for whom accurate medication data were available, the subjects started with an average of 3.4 ± 0.7 medications and were taking 2.0 ± 0.9 medications at the 6-month follow-up. The BP reduction observed for patient groups based on the number of discontinued medications is shown in Table 5. Overall, despite a reduction of medications in this subset of patients (n = 12), the OBP was reduced by an average of 24 ± 16 (systolic) and 12 ± 9 (diastolic) at 6 months. Two subjects had large systolic BP decreases (>40 mm Hg) when discontinuing 2 or more antihypertensive medications. Four other patients had a ≥30-mm Hg decrease when discontinuing at least 1 antihypertensive medication.
The independent pain analysis conducted by the principal investigator showed no or minimal pain reported by the subjects during infusion of the alcohol. In the patients who noted some discomfort, this resolved within 1 to 2 min without any intervention.
In this first-in-human study we have demonstrated the safety and feasibility of “chemical” renal sympathetic denervation using microdoses of alcohol infused via microneedles into the renal artery perivascular space (perivascular renal denervation).
Renal denervation may prove to be a valuable intervention to treat uncontrolled and severe hypertension as well as a number of other conditions that may be driven by sympathetic imbalance or “overdrive.” In early registry studies and in a randomized clinical trial (6,7), denervation of the renal artery was shown to be effective in the treatment of refractory and drug-resistant hypertension (3–6) and may also provide clinical benefit in the management of congestive heart failure (29), central obstructive sleep apnea (30,31), left ventricular hypertrophy (32), metabolic syndrome (33,34), and chronic kidney disease (35,36) and in patients with atrial and/or ventricular arrhythmias (37,38). Given the large potential of this therapy, it will be important to develop and evaluate the safest, most predictable and effective methods for performing renal denervation.
The targeted drug delivery system used in this study is a novel endovascular delivery device that contains 3 distal needles housed inside individual guide tubes contained within the catheter. The catheter is used under fluoroscopic guidance by a single operator using standard endovascular techniques to access the vessel of choice and perform infusion into the adventitial and periadventitial spaces of that vessel. The radiopaque microneedles are deployed with minimal trauma to the normal renal arterial wall to target delivery directly into the adventitial and periadventitial spaces. The device was very simple to use, with catheter positioning and denervation being performed in as little as 2 to 5 min per artery. This approach appears to be reliable and safe. There were no device-related complications or any late renal artery narrowing, as determined by university-based, independent qualified angiography reviewers.
The known neurolytic agent dehydrated alcohol injection (ethanol 98%) was chosen for the evaluation of the drug delivery catheter used in this study (39–43). Alcohol produces injury to tissue cells by dehydration and by precipitation of protoplasm. At very low doses, ethanol at this concentration is known to produce neuritis and nerve degeneration (neurolysis). Deliberate injury to nerves by the targeted injection of ethanol results in more or less an enduring block of sensory, motor, and autonomic function (39–41). The ability to target locally and deliver precisely a neurolytic agent to a deep periadventitial space may allow more complete renal denervation than might be easily or safely achieved using energy-based systems from within the renal artery (27,28).
As seen in the pre-clinical evaluation (27) and suggested by the data from this study, the dose of alcohol that is effective to create substantial renal sympathetic neurolysis is a dose that is so small that it would be unlikely to produce any adverse systemic effects. The amount of alcohol infused locally around the renal artery in this trial is ∼1/40th the amount of alcohol contained in a single alcoholic drink.
Kandzari et al. (17) published a report demonstrating the potential importance of adequacy of renal nerve denervation. They observed a “dose response” with an association between the number and adequacy of the RF ablations and the decrease in ABPM in a post-hoc analysis of the SYMPLICITY HTN-3 results. Thus, it is likely that a lack of adequate depth of nerve injury and a lack of predictable circumferential nerve ablation may lead to inadequate denervation in some patients when using RF systems (17,18,20,21).
The histopathology from the preclinical evaluation of alcohol infusion demonstrated predictable and circumferential renal sympathetic nerve injury, with nearly complete sparing of injury to the intima and media of the renal artery, using doses of 0.15 to 0.6 ml of alcohol per artery (27). The ability to predictably damage the renal sympathetic nerves in a dose-dependent fashion using very low volumes of alcohol delivered in the adventitial space and with minimal or no injury to the normal renal arterial wall structures is appealing. However, these pre-clinical data also suggest that while the dose used in this first-in-humans study (0.3 ml/artery) represents a safe dose with a signal of efficacy, a dose of 0.6 ml/artery has shown similar safety and superior efficacy (27). Further clinical testing using this higher dose will be of interest.
Although there are limitations using thermal ablation for renal denervation, it remains to be determined how effectively chemical denervation with alcohol will overcome these drawbacks. However, the pre-clinical data and these early clinical data from the current study are encouraging and suggest that chemical denervation with microdoses of alcohol is safe and essentially painless and may effectively reduce BP in patients with resistant hypertension, even in the face of a reduction of antihypertensive medications.
This was intended as a safety and feasibility study. As a result, there are limitations in assessing efficacy and OBP lowering due to the small sample size of this study. There were also no ABPM measurements and no separate control group. The dose chosen for the study (0.3 ml/artery), although showing good safety and a signal for efficacy, may be improved on by using a higher dose. Finally, compliance with antihypertensive medication treatment was inconsistent and may have created a negative bias due to the self-discontinuation of medications by many of the patients after their procedure(s).
We report the first clinical use of adventitial, targeted local delivery of very low doses of a neurolytic agent, dehydrated alcohol, to perform renal sympathetic denervation. These data suggest that chemical “perivascular” renal denervation with alcohol is a feasible and safe alternative to energy-based renal denervation. Additional blinded, randomized clinical trials, with the addition of control patients and ABPM measurements, will be necessary to further evaluate this renal denervation therapy.
WHAT IS KNOWN? Renal denervation remains a promising technology for the treatment of hypertension and other disorders. Recent results from radiofrequency-based renal denervation have cast questions regarding the efficacy of renal denervation.
WHAT IS NEW? We evaluated the first clinical use of a new endovascular approach to renal denervation, using chemical neurolysis, via peri-adventitial infusion of dehydrated alcohol (ethanol) to perform “perivascular” renal artery sympathetic denervation. Chemical renal denervation may provide a safe alternative to radiofrequency (thermal) renal denervation and may allow more consistent renal denervation.
WHAT IS NEXT? Additional clinical evaluation of renal denervation, in general, and chemical renal denervation, will be required using ABPM and randomized and sham controlled trials to further evaluate the safety and efficacy of renal denervation.
The authors acknowledge Yoshiaki Mitsutake, MD, PhD, Stanford University; Wook Bum Pyun, MD, PhD, Ewha Womans University, Seoul, South Korea; Agility Medical; and the expert technical contributions of Darren Kent, Jeff A. Burke, and Phil C. Burke, REV-1 Engineering.
Ablative Solutions, Inc. provided funding for this study. Drs. Fischell and Ghazarossian are principals in and cofounders of Ablative Solutions and hold equity in the company. Dr. Vega and L. Minarsch are consultants to Ablative Solutions, Inc. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- blood pressure
- estimated glomerular filtration rate
- office-based blood pressure
- Received August 11, 2015.
- Revision received November 22, 2015.
- Accepted November 23, 2015.
- American College of Cardiology Foundation
- Krum H.,
- Schlaich M.,
- Sobotka P.,
- et al.
- Schlaich M.P.,
- Krum H.,
- Sobotka P.A.,
- Esler M.D.
- Bakris G.,
- Townsend R.,
- Liu M.,
- et al.,
- SYMPLICITY HTN-3 Investigators
- Kandzari D.E.,
- Bhatt D.L.,
- Brar S.,
- et al.
- Vink E.E.,
- Goldschmeding R.,
- Vink A.,
- Weggemans C.,
- Belijs R.L.,
- Blankestijn P.J.
- Tzafriri A.,
- Mahfoud F.,
- Keating J.H.,
- et al.
- Sakakura K.,
- Ladich E.,
- Cheng Q.,
- et al.
- Karanasos A.,
- Van Mieghem N.V.,
- Bergmann M.W.,
- et al.
- Kaltenbach B.,
- Id D.,
- Franke J.C.,
- et al.
- Persu A.,
- Sapoval M.,
- Azizi M.,
- et al.
- Brandt M.C.,
- Mahfoud F.,
- Reda S.,
- et al.
- Mahfoud F.,
- Schlaich M.,
- Kindermann I.,
- et al.
- Hering D.,
- Mahfoud F.,
- Walton A.S.,
- et al.