Year in Review in Cardiac Electrophysiology

The overall incidence of arrhythmias in the community is not as well understood as is the incidence of individual arrhythmias that often receives more focus, such as AF and ventricular arrhythmias. Khurshid et al³ used the UK Biobank—a national prospective cohort—to determine the incidence of rhythm abnormalities and to characterize associated risk. Of >500 000 adults, 2.35% were found to have some baseline rhythm abnormality, which increased over time to an overall prevalence of 4.84% during follow-up. In total, AF, bradyarrhythmias, and conduction system disease were the most common incident arrhythmias seen. Older age, hypertension, male sex, heart failure, and chronic kidney disease were all associated with development of new rhythm abnormalities. Overall incidence of new arrhythmias was 4.72 per 1000 person-years. It should be noted, however, that this likely reflects the minimum population incidence because, depending on the amount of screening performed, there may have been many individuals with otherwise undiagnosed, asymptomatic arrhythmias. This highlights the potential importance of considering how novel technologies may facilitate population-wide screening to better understand overall incidence and prevalence of arrhythmic disease, particularly those that confer modifiable risk of other morbidities (eg, AF).

One long-studied but still limited element of invasive electrophysiology, irrespective of the arrhythmia being treated, is determining ultimate lesion size resulting from radiofrequency ablation. Several reasons exist for wanting to know the final lesion size, irrespective of arrhythmia type—large lesions in thin structures may increase risk of injuring adjacent structures, and insufficient lesion depth in thick tissue may not penetrate to arrhythmogenic sites. Although several studies on predicting lesion size and depth have been published in 2018, here we focus on two that utilize imaging techniques to attempt to characterize lesion size in real time.

Ghafoori et al⁴ used magnetic resonance imaging (MRI) performed immediately post-ablation to determine areas of edema and final lesion extension. In 10 canines, ablation was performed with serial MRIs done immediately after and then at 1, 2, 4, and 8 weeks. They noted that immediate MRI post-ablation demonstrated a region of microvascular obstruction that was >3× larger than chronic lesion volume. It was shown that microvascular obstruction region obtained 26.1 minutes after contrast injection best predicted chronic lesion volume. As consideration of integrating imaging techniques, such as MRI, in invasive ablation procedures receives increased attention, in particular as it relates to utility for identifying acute lesion distribution, it is critical to recognize how such acute findings correlate with the final lesion to avoid overestimating the ablation effect.

Wright et al⁵ used real-time near-field ultrasound imaging to similarly determine whether it was possible to characterize lesion formation. In contrast with Ghafoori et al, they sought to determine whether it is possible to use novel tools during the ablation procedure itself. They used a near-field ultrasound probe integrated into the ablation catheter tip and created lesions in canine atria and ventricles. The ultrasound data was then evaluated to determine correlation with postmortem histology. They demonstrated that ultrasound could characterize dynamics of lesion growth and recognize impending steam pops. These data suggest a potential novel safety element that may be obtained via real-time ultrasound during cardiac ablation in characterizing both elements of radiofrequency effect and safety.

The left anterior oblique view is most commonly used during right ventricular (RV) lead implantation to ensure lead placement on the RV septum (as opposed to the free wall), but the ideal angle is variable from patient to patient. Squara et al⁶ described an elegant technique to identify the left anterior oblique angle that provides an accurate profile of the septum using the superimposition of the lead temporarily placed at the RV apex with a guidewire placed in the superior and inferior vena cava. Lead placement using this view resulted in better discrimination between septal and free wall location compared with a standard left anterior oblique 40° view. Understanding of these essential fluoroscopic concepts may facilitate improved approaches to lead positioning while avoiding complications.

The association between RV apical pacing and higher incidence of heart failure has led to a search for alternate pacing sites or approaches. Permanent HBP can potentially mimic physiological activation of the ventricles through the Purkinje network and prevent pacing-induced dyssynchrony. Building on recent advances in techniques and tools for HBP, several important studies on the feasibility of HBP and its utility in select patient populations were published in 2018. Abdelrahman et al⁷ reported successful HBP in 92% of 332 consecutive patients in a multicenter observational study. Patients with HBP pacing had a narrower paced QRS duration compared with RV apical pacing (128±27 versus 166±22 ms; P<0.01), which was achieved at the cost of higher capture threshold and increased incidence of lead revision (4.2% versus 0.5%) in the HBP group. HBP was associated with reduced incidence of heart failure hospitalization compared with RV apical pacing (12.4% versus 17.6%; P=0.02), with patients pacing >20% deriving the most benefit. HBP has also been reported to recruit the right bundle branch and lead to narrowing of the QRS and improvement in left ventricular (LV) ejection fraction (LVEF) and New York Heart Association functional class in patients with heart failure and right bundle branch block—a population that often does not respond to CRT.⁸ Further data on the feasibility of widespread application of HBP and its long-term impact on outcomes are needed.

Research on optimal patient selection and device programming has provided additional tools to maximize CRT response with well-established technologies. Zweerink et al⁹ reported that QRS duration normalized to LV end diastolic volume is a better predictor of CRT response than QRS duration alone in a retrospective cohort. Plesinger et al¹⁰ further utilized high-frequency ECG applied to 12-lead ECG Holters obtained in the MADIT-CRT trial (Multicenter Automatic Defibrillator Implantation–Cardiac Resynchronization Therapy) to assess the temporal and spatial patterns of depolarization. Greater time difference between amplitude maximum of high-frequency QRS signals in the septal (V1 and V2) and lateral (V5 and V6) leads, that is, ventricular electrical delay, was predictive of beneficial CRT response. Finally, Gold et al¹¹ used a simple measure of interventricular delay, that is, the time between RV and LV electrograms at the sites of final lead implantation, and found significant correlation with CRT response in the SMARTDELAY Determined AV Optimization trial. Patients with prolonged RV to LV time were also more likely to benefit from atrioventricular delay optimization. Thus, combining LV pacing at a site with maximal activation delay with subsequent atrioventricular optimization may have additional benefit.

Implantable cardioverter defibrillator (ICD) implantation is not indicated in the first few months after a myocardial infarction, which, however, remains a high-risk period for sudden death. The VEST (Vest Prevention of Early Sudden Death Trial) randomized 2348 postacute myocardial infarction patients with LVEF ≤35% to the wearable defibrillator versus no device, with both arms receiving guideline-directed medical therapy.¹² At 3 months, there was no difference in arrhythmic death between the device and control groups (1.6% versus 2.4%; P=0.18). Misclassification in adjudicating arrhythmic death, low adherence to the device (mean, 14.0±0.9 hours per day), and likely underpowering for the primary end point may have affected the outcomes. However, the lack of efficacy in reducing the risk of sudden death immediately after myocardial infarction, demonstrated for both ICDs and the wearable defibrillator, is likely because sudden death in this population is frequently because of terminal ventricular arrhythmias or nonventricular arrhythmic events, such as pulseless electrical activity. Hence guideline-directed medical therapy remains the standard of care to prevent sudden death in the immediate post–myocardial infarction period.

Several studies also advanced our understanding of the complications associated with transvenous lead extraction. Analysis of a national cohort of transvenous lead extraction from the National Cardiovascular Data Registry reported major complications in 2.3% of patients, including urgent cardiac surgery in 0.36% and in-hospital mortality in 0.9%.¹³ Predictors of complications included female sex, device infection, extraction of ≥3 leads, and longer duration of implantation. ICD leads with smaller lead diameter, flat wire coil compared with round wire coil, and greater surface area of the proximal coil were also associated with more complications. Lee et al¹⁴ reported the occurrence of stroke in 1% of patients after lead extraction, the majority of which occurred in patients with a patent foramen ovale and thrombi on transvenous leads. Potential for stroke because of presumed paradoxical embolism should be considered before lead extraction in patients with patent foramen ovale. Feasibility and efficacy of shunt closure or use of distal embolization protection devices merit further study in this population.

In addition to improved understanding of approaches to diagnosis, stroke risk management, and nonablative options to manage AF, the past year has seen presentation and publication of numerous studies regarding the approaches to and relative value of ablation of AF. Although all cannot be summarized here, we sought to discuss some of the major potentially guideline-changing or practice-evolving research.

In substudy analysis of the randomized VANISH trial (Ventricular Tachycardia Ablation Versus Escalated Antiarrhythmic Drug Therapy in Ischemic Heart Disease), Parkash et al²⁸ demonstrated that ischemic heart disease patients with drug-refractory ventricular tachycardia (VT) who were amiodarone refractory at baseline benefitted the most from catheter ablation versus escalated antiarrhythmic drugs (AAD), with a 47% risk reduction in any recurrent ventricular arrhythmia (P=0.02). No significant differences were observed in treatment-based outcomes among sotalol-refractory patients. Amiodarone- compared with sotalol-refractory patients had a trend toward higher mortality (HR, 2.4; 95% CI, 0.93–6.22; P=0.07) in the escalated AAD therapy arm. This analysis provides incremental insights in highlighting the relative benefit of catheter ablation and futility of escalating AAD treatment among patients with ischemic heart disease who have already failed amiodarone treatment.

The International VT Center Collaborative group—a registry comprised of >2000 patients with structural heart disease and ventricular arrhythmias who underwent ablation at among 12 international tertiary referral centers—continued to provide important insights in 2018. From this registry data, the first available risk prediction tool (I-VT Score [International Ventricular Tachycardia Ablation Center Collaborative Group Score], available online at www.vtscore.org) was derived and validated to predict 1-year VT recurrence and survival, both before and after VT ablation, from survival-tree analysis of 16 baseline and procedural variables.²⁹ The utility of the risk prediction tool in preprocedural and periprocedural planning of VT ablation procedures for structural heart disease patients remains to be seen but has great promise given the online availability and ease of use. Another notable International VT Center Collaborative substudy found that there were marked postablation differences in survival and VT recurrence based on nonischemic cardiomyopathy pathogenesis, with myocarditis and arrhythmogenic RV cardiomyopathy patients faring best in terms of VT recurrence rates, as well as in 1-year freedom from the combined end point of recurrent VT, transplant, or death; in contrast, those with sarcoidosis, valvular cardiomyopathy, and hypertrophic cardiomyopathy were 1.7 to almost 2× more likely to have recurrent VT, and those with valvular cardiomyopathy and hypertrophic cardiomyopathy had 1.7- to 1.8-fold increased risk of the combined end point in adjusted analyses.³⁰ The study emphasizes the prognostic and potentially therapeutic importance of identifying the underlying pathogenesis in nonischemic cardiomyopathy patients with ventricular arrhythmias.

Advances in techniques to perform high-density intracardiac mapping, as well as in integrated cardiac imaging and simulation, have allowed greater understanding about the potential complexity and 3 dimensionality of VT circuits.

Martin et al³¹ among 6 European centers were able to acquire and characterize complete, high-density activation maps of 36 VT circuits among 31 patients with scar-related VT (87% post-infarction) with use of the Rhythmia electroanatomic mapping system. Notable findings were that reentrant circuits were more anatomically complex than conceived previously, with 1 of 3 of those mapped having multiple entrances, exits, or blind loops and 31% with tortuous isthmuses. Among those with multiple induced VTs (n=5), all circuits were found to originate from the same scar-based substrate, most (60%) with shared isthmus or isthmus components but with varied wavefronts of propagation. The study not only provides quantitative evidence for the complexity of scar-based VTs but also provides justification for why comprehensive, substrate-based ablation is important in preventing VT recurrence.

Pashakhanloo et al³² derived 3-dimensional models from chronically infarcted porcine hearts using ex vivo high-resolution late gadolinium enhancement MRI and demonstrated the complexity and distribution of surviving myocardium within infarcted areas. Virtual programed ventricular stimulation was performed using these models, with simulated VTs observed to course 3 dimensionally throughout surviving tissue channels within these regions, with most observed circuits involving the subendocardium and most involving thin surviving tissue (90th percentile, ≤2.2 mm) amidst scar.

Taking the concept of integrating high-resolution cardiac MRI and simulation data to refine understanding of VT circuits further and incorporating the concept of personalized VT ablation guided by predominantly noninvasively obtained data, Prakosa et al³³ investigated the utility of such a virtual-heart model in noninvasive identification of VT circuits to guide effective ablation. The virtual-heart arrhythmia ablation targeting method applied in 5 chronically infarcted swines, and retrospectively in 21 patients with postinfarction VT undergoing ablation, demonstrated proof of concept, and was then prospectively applied in 5 patients. The virtual-heart arrhythmia ablation targeting predicted sites of VT elimination correlated well in sites of ablation success in both the animal and retrospective human analyses; applied prospectively in 5 human ablation cases at 2 different institutions, virtual-heart arrhythmia ablation targeting–targeted ablation was helpful acutely in 3 of the 5 patients; long-term follow-up for most is still pending.

In the arena of nonischemic cardiomyopathy and VT, Glashan et al³⁴ demonstrated how much more we have to learn about accurate identification of substrate. Electroanatomic ventricular voltage maps were compared with whole-heart human histology obtained from 8 patients before death (7) or cardiac transplant (1) and demonstrated variability in both unipolar and bipolar voltage based on wall thickness, illustrating that commonly used cutoffs for normal bipolar and unipolar voltage are flawed, especially in nonischemic cardiomyopathy. Notably, however, a linear relationship was noted between voltage (unipolar and bipolar voltage) and wall thickness, such that the amount of fibrosis could be accurately calculated if either unipolar or bipolar voltage and wall thickness are known, in a thickness range of 10 to 20 mm. This information could be useful in identifying midmyocardial or epicardial substrate during ablation, especially if preprocedural or periprocedural imaging is not available.

Given limitations in standard radiofrequency energy in targeting either intramural substrate or in achieving durable effect, adjunctive techniques for ablation have been developed, several of which were reported in 2018.

Use of decreased ionic concentration fluid (half-normal saline) during open-irrigated radiofrequency ablation has been demonstrated in preclinical and anecdotal clinical experiences in recent years to create larger radiofrequency lesions and assist in more effectively refractory ventricular arrhythmias. In a multicenter prospective clinical study, half-normal saline irrigation during radiofrequency ablation was utilized among 94 patients at 12 centers with refractory, intramural ventricular arrhythmias previously failing standard radiofrequency ablation using normal saline open irrigant (71% with ≥1 prior procedure).³⁵ Overall 1-year premature ventricular contraction/VT–free survival after a single half-normal saline radiofrequency ablation was 89.4%. There were no half-normal saline-associated complications, although the rate of nonclinically significant steam pops (12.8%) highlights the potential for increased risk and need for judicious use.

The unique potential of irreversible electroporation to selectively destroy relatively large regions of tissue but leave adjacent anatomic structures unharmed has reawakened interest in this technology for a variety of catheter ablation procedures. Livia et al³⁶ demonstrated the feasibility of eliminating all Purkinje potentials in an ex vivo Langendorff canine heart model using unipolar direct-current applications, resulting in significant reduction in the ventricular fibrillation window of vulnerability (17.8±2.9 J pre-irreversible to 5.7±2.9 J postirreversible electroporation; P<0.001) and without any evidence of surrounding myocardial damage. Whether or not irreversible electroporation has potential clinical application in improving the efficacy and safety of ventricular arrhythmia ablations, however, is under active investigation.

Finally, Robinson et al³⁷ expanded on their initial experience of noninvasive cardiac radioablation using stereotactic body radiation, reporting outcomes among 19 patients with drug-refractory VT or premature ventricular contraction cardiomyopathy, in ENCORE-VT (Electrophysiology-Guided Noninvasive Cardiac Radioablation for Ventricular Tachycardia Trial). In this prospective single-arm, single-center phase I/II trial, which was presented as a late-breaking trial at the 2018 American Heart Association Scientific Sessions, median treatment time was 15 minutes, 6-month ventricular arrhythmia burden was reduced by 75% in 89%, AAD use was reduced, and 1-year survival was 72%; serious adverse event rate was 10.5%. Larger, prospective study of this technique is warranted to better understand long-term benefits and risks, but its utility as a bailout therapy continues to show great promise.

Given the potential lifetime risk of device infection and the increasing complication risk of lead removal with longer indwelling leads, epicardial pacing, particularly among patients with complex congenital substrate and limited venous access to eh heart, is frequently used. In a single-center retrospective evaluation, Mah et al⁴¹ reported on 145 patients with epicardial pacing leads who underwent diagnostic coronary angiography or computed tomography and alarmingly describe a 5.5% incidence of coronary artery compression secondary to the leads. The authors report that computed tomography scan and chest radiographs may be important tools to evaluate for coronary artery compression in this population. Further larger scale, multicenter evaluations will be need to further assess the true prevalance, risk, workup, and possible interventions for patients with epicardial pacing leads. However, these results are provocative in that they highlight a potential unforeseen risk of such leads.

Wolff-Parkinson-White syndrome is known to carry a small risk of SCD, and prior studies have indicated that the risk may be higher in children than adults. Etheridge et al⁴² retrospectively analyzed 912 children with Wolff-Parkinson-White, including 96 patients with life-threatening events and found that 65% of patients presented with a life-threatening event as the initial presenting symptom. They also noted that 36% of patients with such an event would not have been characterized as high risk at invasive electrophysiology study, with a shortest preexcited RR interval in AF at electrophysiology study >250 ms. These data demonstrate that asymptomatic children can present with a life-threatening event as a presenting symptom, and that standardly applied electrophysiological criteria to stratify risk may not be applicable in some populations.

Jayaraman et al⁴³ evaluated SCA events in young people 5 to 34 years of age in a large metropolitan area to identify risk factors and the circumstances associated with death. Events were noted in 186 young people, and only a small percentage had any prior warning signs (29%) or occurred with sporting activity (14%). More than half of the patients (58%), though, had ≥1 cardiovascular risk factor, including obesity, diabetes mellitus, hypertension, hyperlipidemia, or smoking. These data highlight that efforts to prevent such events in the young ought not to be limited to those engaged in competitive sports but include larger scale population considerations focused on at-risk youth. Closer alliance with general pediatricians may facilitate early identification, stratification of risk, and potentially preventive therapeutic intervention before index SCA event.

Finally, Lynge et al⁴⁴ evaluated SCD episodes in Denmark from 2000 through 2009 and compared events in patients with and without complex congenital heart disease (CHD). The authors found a ≈10-fold greater risk of sudden death in patients with CHD albeit a low event rate with activity. After implementation of a nationwide fetal screening program to detect CHD, there was a decline in the rate of sudden death, suggesting that screening could reduce the incidence of sudden death in patients with CHD. Further, these findings suggest that limiting exercise prescriptions among patients with CHD may not be the optimal approach to management when trying to prevent sudden death events. Further research is needed to clarify what optimal management strategies ought to be to alleviate the higher burden of sudden death risk among patients with CHD.

Hosseini et al⁴⁵ conducted a gene curation study of 21 genes reported to be assessed with BrS to further clarify whether they are truly associated with the disease. They assembled 3 gene curation teams to individually assess the clinical validity of these genes by using a semiquantitative scoring system of genetic and experimental evidence for gene-disease causality or association. They found that only 1 gene, SCN5A, had definitive evidence for gene-disease association. In fact, the other 20 genes were subsequently classified as disputed in regard to any claim of disease causality in the case of BrS. The findings of this study are provocative in that it calls into question the validity of classifying any rare variant gene or frankly any gene other than SCN5A as associated with the Brugada phenotype. The authors conclude that SCN5A really is the only gene that has a rigorous enough level of genetic and experimental evidence to offer any confidence regarding disease causality. The gene curation teams did work and came to these conclusions independently. This study will have implications as to how genetic testing results can be interpreted by physicians and patients moving forward.

Pappone et al⁴⁶ sought to explore and systematically evaluate the pathological substrate responsible for BrS and whether it could be modified via ablation. In a large series of patients who either manifested BrS-related symptoms (n=88) or no symptoms (n=103) and underwent invasive electrophysiological study and mapping, they sought to clarify how the nature of provocable substrate abnormality correlated with overall risk. They found that the substrate size was the only independent predictor of inducibility and that a substrate region >4 cm² best identified patients at risk of inducible ventricular arrhythmias. This study, although small, does reinforce the notion that risk of ventricular arrhythmias is likely associated with the extent of myopathic involvement, even in BrS, which is typically thought of as principally a channel abnormality. This study also supported that ablative modification of the substrate could prevent arrhythmias among patients with BrS. These data augment prior work by Nademanee et al and suggest that earlier invasive strategies in patients with BrS may attenuate arrhythmic risk.

Vink et al⁴⁷ performed a comprehensive analysis of a large cohort of genotype-positive LQT cases to assess similarities and differences between tangent versus threshold methods of QT measurement on the surface ECG and how they correlate with arrhythmia risk. Only gene-positive LQT1 (KCNQ1), LQT2 (KCNH2), or LQT3 (SCN5A) patients were included in this study, resulting in a cohort of 1484 individuals from 265 families being analyzed. For all genotypes, QT tangent computed a shorter QT interval compared with QT threshold, but the difference was less pronounced for LQT2 patients. Both methods had high interobserver and intraobserver validity and diagnostic accuracy, but despite similar specificity, given a current guideline cutoff of a QTc of 480 ms for diagnosis, the 2 methods yielded different sensitivities. Thus, based on the method of QT measurement even from the same ECG, it was possible to misclassify patients as having risk of LQT syndrome. As a result, the authors suggest an improved classification schema for LQT that does not just rely on age and sex but also the method used to measure the QT. This study supports using several different methods of QT interval measurement in the same patient given appropriate clinical scenarios and considering the presence of LQT syndrome even if only 1 measurement falls in the LQT range.

Treatment of neurally mediated syncope is a challenge in current clinical practice. This is because of the limited therapeutic options and difficulties related to phenotyping of the disease. In light of this, several investigators have sought to use ablation strategies to modulate not just rapid rhythms but syncope and bradyarrhythmias that may be partly mediated through autonomic ganglia in and around the heart. Debruyne et al⁴⁸ sought to evaluate a tailored approach to ablation for treating functional sinus node dysfunction and neurally mediated syncope. Ablation approach was partly planned form computed tomographic scan on 20 patients (12 with syncope, 8 with sinus node dysfunction). They noted that among sinus node dysfunction patients, the number of beats <50 beats per minute on Holter monitoring was reduced by a median 100% at 6-month follow-up. In turn, syncope burden was reduced by 95% at 6-month follow-up. These data are provocative in that they suggest potential utility in cardioneuromodulation for treatment of otherwise difficult-to-manage clinical syndromes. Furthermore, avoidance of a pacemaker may avoid several long-term complications related to chronic indwelling leads. However, patient selection will be critical and require further research to help guide the practicing clinician in who to refer for this therapeutic modality.

The arrhythmic risk in hypertrophic cardiomyopathy includes both atrial and ventricular arrhythmias. In many patients, ICD is recommended to reduce risk of sudden death. Maron et al⁴⁹ sought to understand the long-term outcomes associated with ICDs in high-risk hypertrophic cardiomyopathy patients. Among 486 patients, 19% experienced appropriate intervention for VT, at a rate of 3.7% per year. After intervention for VT, freedom from mortality was 100%, 97%, and 92% at 1, 5, and 10 years. There was noted heightened anxiety of recurrent shocks among those receiving ICD therapy, but overall well-being and quality of life remained intact. These data highlight that hypertrophic primary prevention patients may have better morbidity, mortality, and quality-of-life outcomes than similar primary prevention ischemic cardiomyopathy patients described in prior studies. Tailored support for these patients in understanding their long-term risk is important in differentiating suggested long-term outcomes from patients with other cardiomyopathies.

Cardiac sarcoidosis has received increased attention in electrophysiology because it is likely underdiagnosed. However, once diagnosed, long-term outcomes are poorly understood. Specifically, patients with cardiac sarcoid may present with atrioventricular block, heart failure, ventricular arrhythmias, or any combination of the three. However whether the presence of any one suggests risk of others is unclear. As a result, when patients present even with isolated atrioventricular block, ICD is often recommended. Nordenswan et al⁵⁰ sought to evaluate this recommendation further. Three hundred twenty-five cases of sarcoidosis were identified, with 139 presenting only with high-grade atrioventricular block requiring pacing. Forty-nine patients had concomitant ventricular arrhythmias identified or an ejection fraction <35%, whereas 90 had only atrioventricular block. Risk of SCD or VT was 56% in those with atrioventricular block plus either ventricular arrhythmias or low ejection fraction versus 24% in those who only had atrioventricular block. Based on this, it appears well founded that patients with lone atrioventricular block and diagnosis of sarcoid continue to have a high risk of incident VT and sudden death events during follow-up, indicating utility of an ICD as first-line therapy.