Autonomic modulation for chronic heart failure: a new kid on the block?

Abstract

This editorial refers to ‘Left cardiac sympathetic denervation for treatment of symptomatic systolic heart failure patients: a pilot study’, by G.E. Conceição-Souza et al., published in this issue on pages 1366–1373. The recognition that autonomic dysfunction not only occurs but that it also plays a major role in the development and maintenance of chronic heart failure (CHF), and in the demise of many affected patients, is far from new.1,2 What is new, instead, is what resembles the Alaskan ‘gold rush’: a veritable stampede of investigators and companies alike trying to find the most effective way to modulate the autonomic nervous system in order to correct the dangerous ‘autonomic imbalance’ characterized by increased sympathetic activity and by reduced vagal activity.2 The former is epitomized by the evidence of increased plasma norepinephrine levels and of its excessive spillover from sympathetic nerve fibres in CHF patients;3 the latter by the impairment in muscarinic modulation in the failing heart.4 The dawn of the clinical application of autonomic modulation in CHF patients was marked by European studies employing chronic vagal stimulation,5,6 studies which have already reported their encouraging results. They are now followed by large7 and medium sized8 clinical trials. These in turn are accompanied by other studies using either spinal cord stimulation9–11 or baroreflex stimulation.12–14 Contrary to what may appear at a first and superficial glance, these seemingly different approaches have a lot in common. To be more precise, they all share the same fundamental target: namely, to correct the detrimental autonomic imbalance present in heart failure and strive simultaneously to increase cardiac vagal efferent activity and decrease cardiac sympathetic activity, as proposed with no uncertain words2 on the basis of a strong experimental rationale.15 The first clinical study of vagal stimulation in patients with CHF was published in this journal,5 and it seems fitting that it is here that the first study to employ left cardiac sympathetic denervation (LCSD) in similar patients also appears.16 Conceição-Souza et al. report their results with LCSD in patients with CHF, New York Heart Association (NYHA) class II and III, with a left ventricular ejection fraction > 40%. Their conclusion is that LCSD is feasible and seems safe in patients with systolic CHF. They regard the effects of LCSD as beneficial and suggest performing a larger randomized trial. The report by Conceição-Souza et al. is quite interesting, and from several very different aspects. The study is not without weaknesses, as these are actually fairly abundant, but it possesses the one critically important saving feature in medical science: to be the very first to have explored the novel approach of surgical left cardiac sympathectomy in an attempt to improve the condition of CHF patients. The limitations are numerous and they cannot be ignored. The study population is very small: 10 patients were randomly allocated to LCSD on top of optimal medical treatment and five to the same medical treatment. These numbers and the 6-month follow-up (even though for some patients data are presented at 2 years) call for great caution in the interpretation of any result. Within the 6 months, two patients died of cardiac causes in each group, leading to a final comparison of eight vs three patients. In the LCSD group, there was a third death which, however, was due to H1N1 viral infection and occurred 3 years after surgery, and one patient underwent cardiac transplant. In the medically treated group, two patients died within 6 months, one underwent cardiac transplant, and the remaining two were in NYHA class IV. Whereas it is rather hazardous to comment on mortality in these two very small groups, the most encouraging data seem to come from the evolution in NYHA class as shown in their figure 3. This shows the opposite pattern between the two groups: most of the LCSD patients improved their functional class, with three of them being in class I at 2 years after surgery—which is honestly impressive; in contrast, the two medically treated patients still alive at 2 years follow-up were both in class IV. The 6-month data are likely to include a component of the placebo effect know to accompany surgical procedures, but the presence of five LCSD patients in class I–II at 2 years after surgery deserves attention. Finally, the unconventional surgical approach (clipping) used by the authors provides no advantages, and might result in incomplete denervation. We strongly recommend that future similar studies in CHF patients use either the traditional retropleural approach17 that has provided excellent and reproducible results in long QT syndrome18,19 and in catecholaminergic polymorphic ventricular tachycardia,20 or the thoracoscopic approach.21 The reasons for restricting sympathectomy to the left side have been widely reviewed19 and have been reported by the authors. However, there is another very important conceptual point, also alluded to by the authors in the Discussion, which merits further discussion. Left cardiac sympathetic denervation is almost always carried out with the logical objective of preventing neural release of norepinephrine in the heart. As a consequence, the expected results are those that might follow section of the efferent sympathetic fibres. However, LCSD implies the simultaneous section of the cardiac sympathetic afferent fibres transmitting centrally sensory information from the heart22 and generating both sympatho-sympathetic excitatory23 and sympatho-vagal inhibitory24 reflexes. The interruption of these afferent sympathetic fibres has an important excitatory effect on the vagal efferent activity directed to the heart, which increases considerably as a consequence of the removal of the tonic inhibition mediated by these fibres.25 This is clearly shown in Figure 1. In these experiments, we were recording single vagal efferent fibres running in the cardiac branch of the right vagus at their entrance close to the sinus node.25 The upper part of the figure shows the discharge, with a clear cardiac rhythm, of one such single fibre reflexly activated by a blood pressure increase produced by phenylephrine. In the control condition, there is one action potential per cardiac cycle. The lower part of the figure shows that after section of the left stellate ganglion, at the same level of blood pressure, this fibre now fires four times per cardiac cycle. This corresponds to a four times greater release of acetylcholine in the heart and demonstrates that left cardiac sympathectomy removes the tonic restraint exerted by cardiac sympathetic afferents. Thus, LCSD performed by Conceição-Souza et al. not only produces the interruption of sympathetic stimulation on the heart from the left side but also increases cardiac vagal efferent activity and represents an intervention potentially capable of correcting, at least in part, the autonomic imbalance present in CHF and described earlier. The interest in autonomic modulation in the attempt to improve the clinical condition of CHF patients seems to increase without pause and represents a good example of translational medicine.26 While not oblivious to the limitations of the study by Conceição-Souza et al., we believe that the Brazilian investigators should be congratulated for having proposed one additional and novel approach to the tireless effort of the medical community to counteract the consequences of the pathophysiological changes underlying CHF. Conflict of interest: none declared.