An acute rise in aortic pressure causes ventricular arrhythmias which are suppressed by β adrenergic antagonism.1 ,2 The fact that these arrhythmias occur in isolated heart preparations,2 and after stellate ganglion excision,1 excludes the possibility that the effectiveness of β blockade could be caused by the blocking of a reflex with sympathetic efferent limb. In order to find the reason for this, we utilised the well established chronically denervated canine preparation3 in which the ventricles are almost completely depleted of catecholamines. The corresponding clinical situation is cardiac transplantation, but our study could not be performed ethically in such patients.

Beagle dogs, weighing between 13.7–17.0 kg of either sex, were denervated under 1.5% halothane in N₂O/O₂, removing the parasympathetic preganglionic neurones and the sympathetic postganglionic neurones. Adequate postoperative pain killing drugs were administered routinely. The study was made after 3–4 weeks, when the myocardial catecholamines had depleted. Each dog (four denervated and four controls), breathing air spontaneously, was then anaesthetised with intravenous sodium methohexitone (1%) followed by intravenous chloralose (100 mg/kg). A balloon catheter was introduced from a carotid artery and positioned under fluoroscopic control in the ascending aorta. Left ventricular pressure was recorded via a Gaeltec catheter that had been introduced via a femoral artery. Mean arterial pressure was monitored from an arterial cannula via a fluid filled line connected to a Statham P23Db strain gauge transducer. A standard three lead ECG was displayed continuously and, together with the left ventricular pressure (proximal to the balloon) and mean arterial pressure (distal to the balloon), was recorded on an eight pen chart recorder (Devices M16). All animals were prepared and studied in this way, whether in the innervated or cardiac denervated state. The aortic balloon catheter was inflated using a predetermined effective volume (10–25 ml) and maintained for a standard duration of four seconds. A period of six minutes was then allowed to elapse before repeated inflations. Between the repeated inflations we checked that there were no significant changes in the left ventricular and mean arterial pressure from the initial values obtained at the beginning of the experiment. The concentrations of myocardial noradrenaline (norepinephrine) and adrenaline (epinephrine) were assessed with the use of high pressure liquid chromatography with an electrochemical detection method.

In innervated dogs, obstruction to the ascending aorta by balloon inflation increased left ventricular systolic pressure, and produced an average of 3.6 ectopic beats per balloon inflation, with varying QRS duration. In the dogs with chronically denervated hearts, similar inflation produced no, or many fewer, premature ventricular beats (average of 0.9 per inflation). The between group F variance ratio (non-parametric analysis of variance (ANOVA)) was 76.06, which yields a probability of no difference between innervated and cardiac denervated in number of ectopic beats of 0.0001. The mean (SD) increase in systolic left ventricular pressure was not significantly different between the two groups: denervated 48.9 (19.47) mm Hgv control 51.5 (16.03) mm Hg, p = 0.47.

The total data for premature ventricular beats in the control and cardiac denervated dogs was compared with that for myocardial adrenaline and noradrenaline content (fig 1). In all cardiac denervated dogs these catecholamines were severely depleted, which matches previous findings in our laboratory.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

The reduction in the total number of ectopic beats (A) in cardiac denervated animals (grey columns) and innervated hearts (open columns) was comparable to the reduction in myocardial tissue noradrenaline (B) and adrenaline (C). All columns indicate means (1 SD). All differences were highly significant (p < 0.005).

This study shows that depletion of myocardial catecholamines by chronic cardiac denervation causes the left ventricle to lose most of its sensitivity to acute pressure, load induced arrhythmias. If we study the previously published records of Elzinga and colleagues4 we can observe analogous results in the chronically denervated heart during aortic obstruction (although this is not mentioned in that paper which was devoted to the Anrep effect). The results of the present study are compatible with the preliminary results with β blockade2 in isolated hearts.

Our present results pinpoint the involvement of tissue catecholamines. We attribute the great reduction in ectopic frequency in denervated myocardium to myocardial catecholamine depletion rather than loss of sympathetic and parasympathetic afferent or efferent neuronal pathways because pressure load induced arrhythmias occur after acute denervation.1 ,2 Acute pressure loading of myocardium induces release of proarrhythmic noradrenaline from the myocardium, presumably from sympathetic neurones.5