Discussion
Right and left sided OTVA are often considered to be the same entity. They exhibit similar rates of inducibility during an electrophysiological study, have similar findings on magnetic resonance imaging and respond equally to both adenosine and verapamil, suggesting a common underling electrophysiological mechanism (cyclic adenosine monophosphate-mediated delayed afterdepolarisations) (8). Embryologically, the outflow tracts are also both formed from a common primitive heart tube, rather than having separate origins (8,9).
Adrenergic tone has been shown to be an important influence on all OTVA (10). Furthermore, whilst the autonomic nervous system (ANS) influences VE activity, the presence of modest burden ectopy has also been shown to alter the activity of cardiac neurons and VE-induced cardiomyopathy can be characterised by sympathetic hyperinnervation, which may exacerbate arrhythmogenesis (11,12). This suggests a bidirectional relationship between the ANS and VE.
Although these studies show mechanistic similarities between RVOT and LVOT, we have demonstrated that the behaviour of OTVA across a 24-hour period is dependent on SOO, which must imply a difference in the underlying mechanisms, or perhaps a difference in the autonomic influence on VE activity. We postulate that the differential balance between the parasympathetic and sympathetic innervation may be different in the two outflow tracts. This hypothesis is supported by canine models, where a higher density of sympathetic fibres, compared to parasympathetic fibres, has been identified in the RVOT (13). The density of these fibres is also particularly high at sites where VE and VT can be induced using high frequency electrical stimulation (14).
In human subjects, right-sided OTVA are known to be induced by periods of wakefulness and activity and are less pronounced during periods of sleep, further supporting a role for sympathetic hyperinnervation in the RVOT (13–15). Aortic root ganglionic plexi, in the region of the LVOT, have also been shown to have a higher relative density of parasympathetic (cholinergic) neurons compared to sympathetic (adrenergic) neurons (16).
Our data show that whilst the overall number of total VE as well as bigeminy/trigeminy episodes was higher in the LVOT group, the variability in both of these parameters was greater in the RVOT group, with wide fluctuations in VE activity throughout a 24-hour period and a greater probability of quiescent hours. This suggests that right-sided OTVA appears to be responsive to changes in sympathetic tone, whereas left-sided OTVA is less responsive to autonomic influences and has far more consistent activity throughout the day and night. This may also partially explain the predilection to LV systolic dysfunction that is more commonly observed in left-sided OTVA since there are no periods when the ventricle is given the opportunity to recover from the mechanically deleterious effects of frequent ectopy (4,17). The greater VE variability seen particularly in the morning hours (06:00-12:00) in the RVOT cohort occurs during a period of transition from sleep to wakefulness and a corresponding physiological surge in catecholamines that is seen during these hours (18) and contrasts with the more stable VE burden seen throughout the day in the LVOT cohort. This gives further evidence for a key role of the ANS in the behaviour of these arrhythmias.
Numerous morphological ECG algorithms have been developed to differentiate OTVA with right and left sided SOO. However, these are limited by the complex 3-dimensional anatomy of the outflow tracts (3,5,6), where subtle variations in ECG electrode positioning, cardiac rotation, or body habitus can all impact the QRS morphology of both sinus and ectopic beats (5,19,20). Parameters developed in an attempt to correct for cardiac rotation, such as TZI, correct for some but not all such limitations.
In this study we have devised and validated two novel non-morphological parameters for predicting SOO based on variations in ectopy activity over a 24-hour period. When applied prospectively both parameters are highly effective, with diagnostic accuracy comparing favourably to current ECG prediction parameters. Our parameters would not be impacted by cardiac rotation, chest wall shape or ECG electrode position. An example case is displayed in Figure 4. In this case a 60 year old female has OTVA with a LBBB morphology in which the ECG parameters offer diverging opinions on the likely SOO; TZI predicting a left sided origin and V2S/V3R a right sided origin. In this example our novel parameters both predict that the SOO is right sided, which was confirmed on mapping. Since a 24-hour ECG monitor is nearly universal in the assessment of OTVAs, we find these two parameters to be particularly useful and easy to implement in clinical practice with a high degree of objectivity and reproducibility.