Attenuation of Splanchnic Autotransfusion Following Noninvasive Ultrasound Renal Denervation: A Novel Marker of Procedural Success

Background Renal denervation has no validated marker of procedural success. We hypothesized that successful renal denervation would reduce renal sympathetic nerve signaling demonstrated by attenuation of α‐1‐adrenoceptor‐mediated autotransfusion during the Valsalva maneuver. Methods and Results In this substudy of the Wave IV Study: Phase II Randomized Sham Controlled Study of Renal Denervation for Subjects With Uncontrolled Hypertension, we enrolled 23 subjects with resistant hypertension. They were randomized either to bilateral renal denervation using therapeutic levels of ultrasound energy (n=12) or sham application of diagnostic ultrasound (n=11). Within‐group changes in autonomic parameters, office and ambulatory blood pressure were compared between baseline and 6 months in a double‐blind manner. There was significant office blood pressure reduction in both treatment (16.1±27.3 mm Hg, P<0.05) and sham groups (27.9±15.0 mm Hg, P<0.01) because of which the study was discontinued prematurely. However, during the late phase II (Iii) of Valsalva maneuver, renal denervation resulted in substantial and significant reduction in mean arterial pressure (21.8±25.2 mm Hg, P<0.05) with no significant changes in the sham group. Moreover, there were significant reductions in heart rate in the actively treated group at rest (6.0±11.5 beats per minute, P<0.05) and during postural changes (supine 7.2±8.4 beats per minute, P<0.05, sit up 12.7±16.7 beats per minute, P<0.05), which were not observed in the sham group. Conclusions Blood pressure reduction per se is not necessarily a marker of successful renal nerve ablation. Reduction in splanchnic autotransfusion following renal denervation has not been previously demonstrated and denotes attenuation of (renal) sympathetic efferent activity and could serve as a marker of procedural success. Clinical Trial Registration URL: https://www.clinicaltrials.gov. Unique identifier: NCT02029885.

continued inclusion in the study. ABPM was performed in each patient with a certified device to determine if the mean 24 h SBP was ≥135 mmHg, as per the inclusion criteria.
Two weeks thereafter, the patient attended the third baseline visit. Medication adherence was further evaluated and routine laboratory tests were performed. A urine sample was collected for assessment of medication adherence but this additional measure was introduced in the protocol after the study was already running. Kidney function was evaluated by measurement of the estimated glomerular filtration rate (eGFR). Duplex renal ultrasound and CT angiography were performed to visualize renal artery flow velocity and to rule out renovascular causes of hypertension. BP measurements were performed to determine if the mean SBP was ≥160 mmHg. Furthermore, if there was a difference of ≥15 mmHg between the second and third visit, the patient was required to attend a further visit 2 weeks later. At this point, if the SBP was still ≥15 mmHg away from the value recorded at least the second visit, the patient was excluded from the study due to instability of baseline BP. Of note, change in office BP was the primary objective following the Symplicity HTN-3 study design, the results of which had not been published at the time when first patient were included. 1 Once all inclusion and exclusion criteria were confirmed, patients were centrally randomized to the intervention. Before patients were sent for intervention, all eligible patients at our centre had a detailed autonomic function test as standard of care in patients assessed for device based therapies of hypertension.

Intervention
Patients were centrally randomized in a 1:1 ratio to receive either the active RDN treatment or the sham control, both administered using the Surround Sound System. The treatment consisted of bilateral RDN using therapeutic levels of ultrasound energy. The sham consisted of bilateral sham treatment using diagnostic levels of ultrasound energy.
Group assignment was carried out by the delivery of an encrypted code directly to the Surround Sound System, which then applied the respective amount of energy to the patient. Both subjects and investigators therefore remained blinded to the randomisation.
In order to mask variations in pain response, all patients were given conscious sedation regardless of randomisation assignment. Blinding was assessed by separately questioning the investigator and the patient to determine if they thought they had received the active treatment or the sham.
The Kona Surround Sound system comprises a generator, a water conditioner, and a treatment module with an imaging probe, all contained in a single mobile patient platform.
The imaging probe is a diagnostic ultrasound array that is used to locate the renal artery and associated structures. Real-time motion tracking allows for accuracy in the location of the applied energy. The treatment module additionally contains a phased array therapeutic ultrasound transducer which delivers the therapy.

Measurement of Cardiac Vagal Tone and Baroreflex Gain
The non-invasive NeuroScope™ method was used (MediFit Instruments Ltd, London), to evaluate autonomic neurophysiology, as has been previously described in the monitoring of brainstem autonomic functions in routine clinical examination of neurodevelopmental disorders such as Rett syndrome and Autistic Spectrum Disorders 2-5 .
A non-invasive continuous index of cardiac vagal tone (CVT) defined as, "pulse synchronized phase shifts in consecutive cardiac cycles" is a form of pulse interval jitter was quantified in real time by the NeuroScope as previously described 6 . CVT is also continuously measured in real time and allows measurement of sudden or rapid changes in response and can be used to monitor the activity of the brainstem parasympathetic system Baroreflex sensitivity (BRS), measured in ms/mmHg, was recorded as a measure of central cardiovascular regulation, by the NeuroScope as previously described 8 . This index is defined as the increase in pulse interval per unit increase in systolic blood pressure and quantifies the negative feedback control of blood pressure beat by beat. The method allows detection of rapid changes in baroreflex gain in real time within a continuous measurement. It therefore facilitates the crucial measurements of sudden changes in autonomic regulatory responses and is also useful for monitoring whether or not this sympathoregulatory function of the brainstem is fully engaged during cardiovascular control at any given moment. Baroreceptor control of the heart has been modelled as a closed loop feedback system with a delay line in the sympathetic nervous system that makes it resonate at 0.1 Hz 9 . Hence the effect of one perturbation is diffused over a period of at least 10 s, which is twelve cardiac cycles if the HR is 72 beats min-1 . For normal heart rates (between 60-100 beats per minute), it is reasonable to use an average of 20 beats to cover a single perturbation.
The baseline autonomic status (CVT and BRS) was recorded in the supine position for at least three minutes with the mean BP varying by less than 5 mmHg. The conditions for autonomic function testing are described in the main manuscript.

Statistics
When looking at within group differences between time points, two-tailed dependent student's t tests were used. A significant difference was defined as having a p value < 0.05. Table S1. Within group changes in resting cardiac vagal tone (CVT) and baroreflex sensitivity (BRS) between baseline and 6 months post procedure.