DISCUSSION
This is the first case series reporting the acute and long-term results of a novel anatomical approach for SNM to treat IST targeting the AR under ICE guidance. The main findings of our study are as follows: 1. ICE allowed real-time visualization of the lateral and septal aspects of the AR at the level of the SVC-RA junction; 2. RF applications at the AR earliest activation site during maximally tolerated isoproterenol infusion (up to 10 mcg/min), produced a suboptimal reduction of the HR in all patients; 3. Extending the ablation lesion set from the earliest site at the AR towards its septal region led to acute and long-term modulation of the SN in all patients without the need to extend the ablation lesion set towards the lateral/posteroinferior extension, thus avoiding risk for PN damage or SN dysfunction.
The importance of the recognition of the AR for SNM by ICE was previously described by Asirvatham et al. who reported a case of a 48-year-old woman with multiple failed ablation procedures for IST performed at the high posterolateral RA earliest site of activation7. The patients underwent a repeat procedure showing early activation at the lateral aspect of the SCV-RA. However, detailed 3D mapping under ICE visualization demonstrated an even earlier activation site in the AR area, where RFA was effective and led to definitive control of the patient’s arrhythmia and symptoms. In our study, ICE also facilitated the identification of the AR early activation sites, initially masked in all the patients during contact activation mapping, by extrapolation of early activation sites between the SCV and the RA. Our approach included correction of this extrapolation based on tagged points on the surface of the AR, taken with the ablation catheter under ICE guidance, and including those activation points on the corrected anatomy, which included the AR. Based on our case experience and anatomical understanding of this region, we believe AR recognition by ICE increases procedure efficacy and safety, by focusing the mapping and ablation efforts on the AR and preventing unnoticed ablation inside the right atrial appendage. Notably, mapping and ablation of the earliest activation sites alone did not produce the desired HR reduction of around 25%. However, an anatomic extension of the ablation lesion set towards the septal aspect of the AR, performed independently of the local activation times, produced a transient HR increment followed by a significant HR reduction >25% and a change in the P wave axis. These findings suggest that part of the SN complex was located in the septal region, but it was not identified by the activation mapping under the isoproterenol effect. Notably, the reduction of the HR was sustained during follow-up, measured by cardiac monitoring and stress ECG, and associated with symptomatic relief. Despite the encouraging results of this novel anatomic approach to modulate the SN, our investigation does not provide insight into the mechanistic role of the septal aspect of the AR on IST. Furthermore, empirically extending the ablation target to the septal and not the lateral region of the AR, may not achieve the desired HR modulatory effect in some patients, owing to different characteristics in their SN anatomy and extensions. Anatomical studies traditionally describe the SN complex as a crescent-shape structure with a “comma” disposition, in which the “head” lies subepicardial within the terminal groove, formed by the lateral junction of the SCV with RA, and the “tail” extends toward the posteroinferior aspect of the CT reaching the inferior vena cava and RA junction9,10 (Fig. 1a). This usual “comma” depiction of the SN complex on anatomic drawings, has directed efforts of mapping and ablation for SNM to the lateral SVC-RA junction where the PN resides. Although the “comma” anatomical disposition is the most prevalent, a “horseshoe” arrangement of the SN has also been found in around 12% of hearts9. In the “horseshoe” arrangement, the SN complex is situated both medial (“head”) and lateral (“tail”) to the mid-line of the SCV-RA junction at the AR10 (Fig. 4 a, b). This anatomic variation of the SN showing an extension towards the septal aspect of the SCV-RA junction seems to be validated by the first “in situ” 3-D visualization of the human cardiac conduction system performed by means of advanced imaging tools of virtual CT dissection on human cadavers11(Fig. 4 c). Based on this anatomic variation, it is plausible that the extension of RF ablation from the earliest activation site at the AR towards its septal region could explain the results of our proposed approach described in this study. As mentioned previously, the usual target for SNM is defined as the earliest activation site on the AR under the effect of Isoproterenol. A concept based on physiological studies demonstrating shifting of earliest activation from the “tail” to the “head “of the SN complex under progressive HR increments4. However, if patients do not tolerate high doses of Isoproterenol during the procedure, it is possible that the most anterosuperior aspect and septal extension of the SN are not activated, and therefore not targeted, leading to suboptimal results, unless extending the lesion empirically towards the septal AR is performed (electro-anatomical approach). Coincidentally, an electroanatomic analysis of the sinus impulse propagation in normal human atria has shown that the area of earliest activation in the RA has a “spindle” shape in 5 of 7 patients and a “horseshoe” shape in the remaining 2 located between the SCV-RA junction, which could represent the functional characterization of this anatomic SN variation12. Finally, it is possible that the extension of RF delivery toward the septal aspect of the SCV-RA junction at the AR in proximity to the interatrial groove could produce a modification of the neural input of the SN with an additive neuromodulatory effect. Interrupting autonomic nervous input to the sinus node by isolating right pulmonary veins, superior vena cava, and ablating fat tissue surrounding the sinus node is an integral part of the surgical treatment of IST13. A recently published SN sparing hybrid ablation technique including RF bipolar clamp isolation of the superior and inferior vena cava and lateral ablation line across the Crista terminalis while sparing the SN region (plus adjunct RF endocardial touch-up to complete lesion sets in 46% of cases); was compared to a conventional SNM approach (endocardial and/or epicardial mapping and RF ablation at the site of earliest atrial activation) in 100 consecutive patients (50 in each group)14. The SN sparing hybrid ablation group showed a significant acute and long-term improvement in mean daily heart rate and peak 6-minute walk heart rate15. These results are consistent with our small study population, but with a much more simplistic approach. As mentioned previously, perhaps the anatomic disposition of the SN complex with a “horseshoe” type and a regional neuromodulatory effect explains why our conservative approach led to excellent acute and long-term results.
Beyond the anatomical explanation of plausibility, it seems that the neuromodulation could play an important role in our results. First, in the body of the sinus node, there is a higher concentration of sympathetic neurons compared to parasympathetic postganglionic neurons16. Since the target of our approach is the body of the sinus node, the ratio of sympathetic neurons removed compared to parasympathetic neurons would be greater, with a result of a predominance of the parasympathetic ones. This is a simplistic explanation; however. As we mentioned in our approach, ablation lesions were extended towards the septal region of the AR where the dorsal right atrial ganglionated subplexus (DRAsGP) is located. DRAsGP spread widely into the dorsal and lateral right atrium, including the sinoatrial nodal region and the superior surface of the right atrial appendage17. The ablation effect on sympathetic and parasympathetic neurons may explain an improved regulatory effect over the SN hyperactivity seen in IST patients, but this autonomic interaction is more complex than a simple calculation of the number of fibres or GPS ablated.
One question yet to be resolved is why, despite an inadvertent parasympathetic ganglion modulation, a decrease in heart rate was documented in our results. The answer to this question can be addressed by a feature of the sinus node known as functional inhomogeneity, which refers to the difference in the concentration of autonomic receptors along the sinus node, and how 3 areas in the sinus node can be discriminated that have a different response to neurohumoral influences. When performing our ablation approach, the target would be the upper part of the sinus node, leaving cells with a diminished response to neurotransmitters as a guide for the pacemaker. On the contrary, if the conventional ablation method is chosen, it is likely that the target will be cells from the upper zone and transition, leaving cells with a high sensitivity to adrenaline as a subsidiary pacemaker and it could explain why patients could persist with tachycardia18.
Finally, after explaining our findings from an anatomical and functional point of view, another possible explanation could be the injury to the sinus node artery. Scanavacca et al19 published a case series, where they documented injury to the sinus node artery after performing cardiac denervation procedures with a frequency of 4.76%. This was presented and confirmed by CT after performing ablation in the area of ​​the cavo-atrial junction and the area between the SVC and the aorta, the latter is a target for ablation in our approach for modulation of the sinus node, however in none of our patients presented acute signs of sinus dysfunction.