Fine tuning renal denervation

A watershed moment has been reached in the by now long-running saga of the treatment of hypertension with catheter-based renal nerve ablation, with the recent Lancet publication of three positive, sham-controlled studies [1–3]. Ambulatory blood pressure (BP) was similarly, and materially lowered in the SPYRAL HTN-OFF MED trial [1], the SPYRAL HTN-ON MED trial [2] and RADIANCE-HTN SOLO trial [3]. BP response heterogeneity, however, remains as the primary clinical problem with this device treatment of hypertension. This also is a scientific problem: to what extent does this unpredictability in BP response represent ‘noise’ (attributable to random pressure variability)? To what extent is it a consequence of the diverse pathophysiology of essential hypertension, which is commonly but by no means invariably neurogenic [4]? To what degree is it due to vagaries and inconsistencies in the performed denervation procedure, which are legendary [5]? The article by Wolf et al.[6] addresses this last problem: what might be the optimal procedural performance for renal denervation, and less centrally, what might be the optimal denervation catheter. The authors have tested the renal sympathetic denervation capacity of two radiofrequency ablation catheters, the Medtronic Symplicity Spyral catheter and the Terumo IberisBloom catheter in domestic pigs. These are both multielectrode catheter systems, of rather similar design, which as it turned out also performed similarly. What was not done here, but would have been instructive, was to compare radiofrequency catheter denervation using Spyral and IberisBloom catheters with intraarterial ultrasound denervation with the Recor Paradise catheter, the catheter used in the RADIANCE-HTN SOLO study [3]. Whether radiofrequency and ultrasound catheter denervation systems are comparable in terms of efficacy and safety is not known. The comparable outcomes in the SPYRAL HTN-OFF MED and RADIANCE-HTN SOLO studies [1,3] do suggest they may be, so that there perhaps is a class effect for BP lowering with renal denervation, however this might be produced. The background to the testing done by Wolf et al.[6] was provided by the recent demonstration [7–9] that the renal sympathetic nerves are not equidistant at all points from the lumen of the renal artery, but converge on the distal renal artery and the renal artery divisions, potentially making denervation easier in the distal renal arterial tree. Actually this is old knowledge [10], derived previously in an era when renal denervation was performed surgically for intractable renal pain, but regrettably was not acted on in the early days of catheter-based renal denervation, when energy was often preferentially directed into the proximal renal arteries. Actually, it is demonstrably easier to ablate the renal sympathetic nerves when radiofrequency energy is directed into the distal renal artery and renal artery divisions [7,11]. With this starting point, Wolf et al.[6] have set out to establish the optimal procedural parameters for renal sympathetic denervation. They have measured the depth of radiofrequency (RF) energy tissue penetration in the main renal arteries and the renal artery branches, have quantified renal denervation achieved when RF energy is delivered both proximally and distally, and determined what is the optimal spacing of energy applications along the renal artery, for maximum denervation. The Spyral catheter and IberisBloom catheter systems performed similarly and the results can be pooled. Mean RF energy penetration was approximately 6 mm for the main renal arteries, and less, approximately 3 mm for the renal artery branches, which in both sites is adequate to reach most of the targeted sympathetic nerves [7–9]. RF energy can be deflected by tendons and lymph nodes, or disappear into veins which can act as a sink [12]. Whether the differing energy penetration in the main renal arteries and branches was explicable in terms of these processes is not known. Renal sympathetic nerve denervation can be quantified in experimental animals by measurement of the reduction in the renal cortical content of the sympathetic transmitter, noradrenaline. The authors documented greater reduction in renal noradrenaline content with RF energy application in the renal artery branches than in the main renal artery, and with closer spacing of energy application in the arteries [6]. For effective denervation, closer spacing of energy application was needed in the main renal arteries, median 0.26 ablations per mm of arterial length, compared with 0.16 ablations per mm in the branches. The density of RF application demonstrated here to be needed for near-total sympathetic denervation was much greater than that applied in the negative but misleading Symplicity HTN-3 study [13]. Recently, and on good evidence [7–9,11] it has become customary to apply RF energy at high density, crowded into the available arterial space [1,2], much as was done here by Wolf et al.[6]. This is necessary because of the mentioned unpredictability of energy flow toward the neural target, with deflection by some tissues, and disappearance in venous sinks [12]. The renal afferent (sensory) nerves were not studied by Wolf et al.[6]. Nociceptive renal afferent nerves which project to the brain are believed to cause central sympathetic excitation; their ablation lowers sympathetic outflow and, most likely, contributes to BP lowering with renal denervation [14,15]. It is not clear whether the path of the renal afferent nerves is identical to that of the sympathetic efferent nerves, closest to the artery distally. Perhaps this is the case, as there is evidence from experimental studies that both sympathetic efferent and afferent nerves can occupy the same neural bundle [16]. If this is so, application of RF energy in the distal renal artery would be optimal for ablation of afferent nerves also. This matter is unsettled. Catheter-based renal denervation, as typically performed, pays little attention to the renal afferent nerves. Contrary to almost all previous research on renal denervation, the study by Wolf et al.[6] for the first time ‘raises a red flag’ concerning possible collateral damage. In the past, efficacy has been questioned, but not safety. The initial concern noted was that renal artery lumen diameter at renal angiography, in the main artery and its divisions, and with both RF devices, was reduced by 3–10% 5 months after the renal denervation. The biological character of this minimal luminal narrowing (<1 mm, at the limits of angiographic detection) is not known and its significance if any is uncertain. There was no endothelial damage of focal stenoses. Certainly, clinical significant renal artery stenosis caused by renal denervation is a rarity [17]. The authors also noted some necrosis in contiguous structures, specifically in psoas muscle, lymph nodes and the ureter. They tended to discount these changes, describing them as minimal. With renal artery branch RF energy delivery, some kidney damage was detected. The precise nature of this renal injury was not evident from the authors’ report. Could it be that prior general advice to administer energy proximally in the renal artery, to ‘keep the RF shadow off the kidneys’ was actually well placed? Observations from this small study in ten pigs, however, must be balanced against the much wider experience in humans of RF application in the renal artery branches, where safety so far has been unequivocal [1,2]. ACKNOWLEDGEMENTS M.D.E. receives salary support from a Senior Principal Research Fellowship of the National Health and Medical Research Council of Australia. His research is supported in part by the Victorian Government Operational Infrastructure Support Program. Conflicts of interest There are no conflicts of interest.