Iatrogenic risks of left atrial appendage isolation: Targeting the bull's eye or circling the drain?

The opinion on whether the left atrial appendage (LAA) is a friend or foe depends on whether it is considered a vestigial structure whose sole purpose is to serve as a potential source of arrhythmias or thrombus, or whether its important electromechanical and neurohumoral role is considered. The LAA is the most compliant part of the left atrium and therefore buffers against a sudden rise in pressure with increases in left atrial (LA) volume. It may contribute between 20% and 40% of left atrial systolic function and contains high amounts of atrial natriuretic peptide granules, released in response to stretch, that help to regulate volume homeostasis.

Automatic or reentrant atrial tachycardias may arise from the LAA and may also serve as triggers for atrial fibrillation (AF).3, 4, 5, 6 Isolating the LAA with either radiofrequency catheter or cryoballoon ablation appears to decrease the risk of recurrent AF, particularly for persistent fibrillation.7, 8 This is not always successful, owing to the thickness of the neck and internal pectinate muscles, with incomplete isolation leading to a higher risk of recurrent arrhythmias. This may require further ablation or even mechanical isolation. Because of the proximity of the left phrenic nerve and circumflex artery, care must be taken to avoid unintended collateral damage. The LAA is also a very thin-walled structure, increasing the risk of hemopericardium during catheter manipulation. Delay in activation or electrical isolation of the LAA may lead to undesired electromechanical consequences, including dyspnea or stroke. Either lifelong oral anticoagulation or endocardial LAA occlusion are therefore recommended if endocardial LAA electrical isolation is performed.10, 11 We still do not know whether mechanical or electrical LAA isolation may decrease the risk of AF by eliminating triggers, or whether the reduction of effective contiguous LA surface area decreases the likelihood of supporting multiple-wavelet reentry.

In this issue of Heart Rhythm Case Reports, Li and colleagues describe a patient who had undergone bioprosthetic mitral valve replacement and LAA ligation 10 years earlier and presented with an atrial tachycardia (AT) of variable cycle length (298–325 ms). Using high-density endocardial mapping with a grid catheter, the origin was mapped to just within the 10-mm-wide LAA stump. Unfortunately, during catheter manipulation prior to ablation, the arrhythmia was terminated and no longer inducible. Ablation was performed using an irrigated catheter at the assumed site of origin; however, the arrhythmia recurred after 2 days. Repeat mapping confirmed the same site of origin and a cryoballoon was used to isolate the LAA stump, terminating the tachycardia. Subsequent pacing confirmed acute isolation of the LAA stump. Two years later a new AT occurred despite confirmation that the previous LAA stump isolation was intact. The tachycardia proved to be reentrant, rotating clockwise around the base of the LAA stump, and was ultimately terminated by a linear lesion from the anterior margin of the circular LAA isolation to the mitral annulus.

The case illustrates several important considerations for arrhythmias arising from near the LAA. The first is the time interval between cardiac surgery and arrhythmia of 10 years. This suggests that electrical tissue may change very slowly over time until critical substrate is present to initiate and perpetuate arrhythmias. This has important implications on gauging the success rate of AT/AF ablation procedures, since new substrate or spontaneous modification of previous substrate may occur. This should be kept in mind when judging long-term success rates of these procedures.

The second is that occurrence of a new, distinct arrhythmia 2 years after LAA isolation demonstrates how the efforts to eradicate one arrhythmia may inadvertently create substrate for a second. Given the proximity of the LAA to the mitral annulus, ablation in this region may create a small isthmus of slow conduction. This patient was even more susceptible, having previously undergone a mitral valve replacement, which would result in perivalvular scarring near the base of the LAA, thus ultimately providing the perfect milieu for reentry. Similar proarrhythmia may occur from wide-area circumferential ablation during a pulmonary vein isolation if the circular lesions come into proximity near the high posterior wall, creating a narrow isthmus of slow conduction that may facilitate sustained roof-mediated tachycardias. Ultimately slow conduction, due to either spontaneous or iatrogenic substrate in the setting of a minimum path length, will allow reentry. Although one may observe termination during ablation of ATs, this is not sufficient to prevent recurrence. Further testing with pacing maneuvers, high-output pacing on the ablation line, high-density mapping, and programmed stimulation with or without beta-adrenergic agents are critical to establish complete block and noninduciblity, as is usually done for cavotricuspid isthmus ablation.

A third teaching point of this case is the difficulty in judging whether an ablation procedure succeeded or failed. The second procedure and first recurrence were due to inability to fully map and target the index arrhythmia owing to loss of inducibility from the catheter “bump” of the substrate. The third ablation was for a completely new arrhythmia mechanism, albeit the substrate was close to the previous ablation. Should this be considered a failure? We have the same issue when we define success or failure after ablation for AF. We define recurrence as any atrial tachyarrhythmia more than 30 seconds after the initial 3-month blanking period. If the patient’s index arrhythmia was persistent AF but the only recurrence is an atrial tachycardia of more than 30 seconds, should that be considered a failure of the previous AF ablation? If a recurrent AT is nonsustained and asymptomatic, but more than 30 seconds in a patient with previous persistent AF, is that considered to be a failed ablation?

Finally, the patient in this case report had first undergone mechanical LAA ligation but had a residual stump and a new arrhythmia. This emphasizes the importance of complete electrical and mechanical isolation at the neck of the LAA; otherwise, risks of both thromboembolism and arrhythmias continue. Mechanical isolation of the LAA can be performed with the LARIAT device, via epicardial atrial clipping using the Atricure clamp during open heart surgery or thoracoscopically, or with direct surgical excision and oversewing. Although mechanical isolation may have the benefit of simultaneous electrical isolation, the results reported from the aMAZE trial (presented by Dr David J. Wilber at the American Heart Association Virtual Annual Scientific Sessions [AHA 2021], November 14, 2021) were disappointing and showed that that LAA ligation with the LARIAT plus pulmonary vein antral isolation in persistent AF patients was not superior to pulmonary vein antral isolation alone to prevent recurrent AF.

The LAA is a remarkable anatomic and physiologic structure and there are times it is more a foe than a friend. It is up to us to be judicious in deciding when to declare war on it and which of our therapeutic weapons we should use.