Revisiting Atrial Pacing in the Long QT Genotype Era

Successful cardiac pacing to treat ventricular tachycardia was first described in the 1960s.¹ Within two years, the technique was applied to patients at risk for arrhythmia secondary to QT prolongation.² By the 1980s, clinicians demonstrated that pacing decreased breakthrough cardiac events (BCE) in selected children and adults with Long QT syndrome (LQTS).^3–5 In a seminal study published in 1991, Arthur Moss and colleagues described 30 patients from their registry who had permanent pacemakers implanted to manage BCEs.⁶ The median number of cardiac events decreased in patients using atrial pacing, ventricular pacing, and dual chamber pacing, all of which appeared to have a salutary effect on the incidence of BCEs.

A steady drumbeat of small studies continued to support the first publications. Case reports and case series subsequently published data on high-risk patients, including neonates,⁷ those with Andersen-Tawil syndrome,⁸ and those with Jervell and Lange-Nielsen syndrome.⁹ With improved permanent pacing systems, these studies coalesced around atrial pacing to achieve arrhythmia reduction. Few patients with active ventricular pacing continued to be reported.

In the current issue of the Journal of Cardiovascular Electrophysiology, Dr. Kowlgi and colleagues described their retrospective experience with “Intentional Permanent Atrial Pacing”, or IPAP as the authors abbreviate the concept.[CITATION REQUIRED] This thoughtful and well-done retrospective study is helpful for three reasons. First, it is a systematic reinforcement of the early conclusions that the rate of BCE decreases after IPAP is initiated. Second, it advances the hypothesis that the corrected QT interval (QTc) decreases in response to atrial pacing. Third, it offers novel genotype information, which begs the question of how patients should be selected for IPAP moving forward.

Both the Moss study and the current Kowlgi study noted a statistically significant decrease in BCE after the institution of IPAP. The magnitude of decrease is similar. Both studies described an event rate between 0.5 and 0.9 events per year before IPAP and less than 0.2 events per year afterwards. Several plausible biological mechanisms explain why artificially increased heart rates in the absence of adrenergic stimulation (i.e. permanent atrial pacing) might decrease BCE. Pauses have been implicated in approximately three-quarters of torsade de pointes initiation events.¹⁰ While these are “pause-related” arrhythmias, rather than strictly “pause dependent” ones, the marked T-wave changes that occur after a sudden increase in RR interval beyond the base cycle length reflect increased dispersion of repolarization and enhanced early after-depolarizations (EADs). Both mechanisms are central to the current theory of arrhythmia initiation in patients with prolonged QTc intervals, especially those with KCNH2 and SCN5A genotypes.¹¹ Moving from cellular electrophysiology to practical clinical medicine, pacing may also permit patients to take higher doses of beta blockers. Beta blockers remain the cornerstone of LQTS pharmacotherapy, especially in patients with KCNQ1 and KCNH2 genotypes. Atrial pacing provides a heart rate floor. In cases where symptomatic bradycardia prevents patients from taking enough beta-blocker, providing a heart rate floor may decrease or eliminate side effects and allow them to take a higher dose. Thus, even before considering whether atrial pacing shortens the QTc, there are mechanisms by which pacing may help decrease the probability of BCE: permitting optimal pharmacotherapy, decreasing the frequency of triggering events (EADs), and decreasing the dispersion of repolarization.

However, one intriguing result presented in the current study is the decrease in mean QTc interval (from 533 ms to 488 ms). The magnitude of the decrease is important because it was not only statistically significant, but because the mean QTc crossed a major clinical boundary at 500 ms. It is not a foregone conclusion that the QTc should shorten with atrial pacing. In the largest previous study, Dr. Moss and colleagues showed no statistical association between pacing and the median QTc.⁶ In addition, a decrease in the QTc interval was not seen when pacing healthy dogs,¹² children,¹³ or in several other studies of patients with long QT syndrome who received atrial pacing.^{4, 5, 14} However, the current study offers data that were not previously considered, including genotype. The decrease in QTc was largest and most consistent in patients with LQT2. Earlier reports that failed to find an association between pacing and QTc did not report genotype and may have contained a different distribution of genotypes than the current study, perhaps including a higher proportion of LQT1, given the era of those publications. If so, it offers a possible explanation for the difference. In the current report, the change in QTc was most subtle in LQT1 patients.

By contrast, the patient with the most dramatic decrease in QTc was diagnosed with LQT3. This is potentially consistent with existing in vitro data from SCN5A variants. In the in vitro model, when faster heart rates caused the expected elevation in intracellular calcium concentration, the abnormal late sodium current conducted through mutant sodium channels decreased. However, the same heart rate-related calcium elevation did not reduce current through normal wild-type channels, potentially protecting cells from delayed repolarization and offering a plausible mechanism for a decrease in QTc when heart rates are artificially elevated by atrial pacing in LQT3.¹⁵ However, genotype is likely to be only a partial answer, as suggested by Figure 3 in the manuscript published this month. Dr. Kowlgi and colleagues demonstrated that not all patients who received pacing had a decrease in QTc and not all who had a decrease in QTc remained free from subsequent BCE, regardless of genotype. Some of this may be due to understandable biases in retrospective data collection of QTc intervals, but these data also underscore that BCE in LQTS remains a probabilistic phenomenon, not a stoichiometric one.

The literature has now established a 40-year track record that pacing is beneficial in some patients with LQTS. The question remains: “Which ones?” Like all the LQTS pacing studies before it, this report studied a highly selected group of candidates. An ICD had been implanted in 92% of 52 patients studied and 35% had undergone cardiac sympathectomy. Many were young children and infants at diagnosis; sudden cardiac arrest had occurred in 17%; the mean baseline QTc was over 490 ms. In addition, it is remarkable that in 1991, Dr. Moss reported pacing in 30 of 1,016 patients. The current study reported that IPAP was used in 52 of 1,065 patients. Even after 30 years of effort to improve diagnosis and management for LQTS and to identify ideal candidates for each therapeutic modality, the frequency at which smart clinicians applied pacing therapies remained similar and was mostly confined to the highest risk patients.

Finally, the specter of complications looms over all studies of pediatric devices. Surgical revision for device complications was required in 14% of patients in the current study, a number that is congruent with nearly all contemporary reports in children and young adults. The report from Dr. Kowlgi and colleagues supports the longstanding conclusion that pacing can decrease BCE in some patients and offers new insight into the driving genotype of LQT2, but devices do not come for free and a bright line indicating who will benefit from pacing remains elusive.