Variant-specific therapy for long QT syndrome type 3

Congenital long QT syndrome (LQTS) is an inherited cardiac arrhythmia syndrome characterized by a prolonged QT interval on the electrocardiogram.¹ Long QT syndrome type 3 (LQT3) is caused by gain-of-function variants in the SCN5A-encoded α-subunit of the voltage-gated cardiac sodium (Na⁺) channel Nav1.5. LQT3-linked SCN5A variants typically interfere with fast inactivation of the channel, which causes increased sustained inward Na+ current (I_Na) and prolongation of the cardiac action potential.² In addition, some variants such as p.D1790G can prolong the ventricular action potential in the absence of an LQT3 variant-induced sustained Na⁺ current.³ Regardless of the biophysical mechanism, the abnormal Na⁺ influx can lead to early afterdepolarizations, which in turn cause triggered activity and the development of potentially fatal “torsades de pointes (TdP)-like” polymorphic ventricular tachycardia.

The current standard of treatment for LQT3 includes standard preventative measures and beta-blockers for patients with syncope or seizures.⁴ Options for treatment intensification include other medications such as the class 1B antiarrhythmic drug mexiletine, left cardiac sympathetic denervation, a pacemaker, and/or an implantable cardioverter-defibrillator (ICD). LQT3 can be challenging to treat because the QT interval shortens with exercise, rendering beta-blockers less effective. As such, there currently is an unmet need to determine genotype-specific medications for treatment of LQT3.

In this issue of Heart Rhythm Journal, Stutzman et al⁵ characterize a novel variant of LQT3, p.F1760C-SCN5A, which was identified in an infant who presented with TdP despite treatment with beta-blocker, mexiletine, bilateral stellate ganglionectomy, and an ICD. Whole-cell patch clamping of TSA-201 cells expressing p.F1760C-SCN5A revealed a positive shift in half inactivation voltage (−72.2 ± 0.7 mV) compared with cells expressing wild-type SCN5A (−86.3 ± 0.9 mV; P <.0001), thereby increasing the window current by 2-fold. Optical mapping of induced pluripotent stem cell–derived cardiomyocytes (iPSC-CMs) from the LQT3 patient showed a marked increase in baseline action potential duration (APD₉₀ 601 ± 4 ms) compared to isogenic controls (423 ± 15 ms; P <.0001). Whereas mexiletine treatment failed to shorten APD₉₀, phenytoin rescued both the electrophysiological phenotype and APD of the cells expressing the p.F1760C SCN5A variant.

Interestingly, although p.F1760C is a novel disease-linked SCN5A variant in humans, it is not the first-time p.F1760C-SCN5A has been studied. In 2005, Carboni et al⁶ studied the p.F1760C and p.F1760A mutations in HEK29 cells to determine their reaction to covalent modification by thiols. Amino acids F1760 and Y1767 are part of the local anesthetic binding site on Nav1.5 channels. Using whole-cell patch clamping of HEK293 cells expressing p.F1760C-SCN5A, they showed a positive shift in half inactivation voltage compared with wild type. In addition to altered gating, the p.F1760C mutant Na⁺ channels exhibited baseline use-dependent current reduction and reduced sensitivity to lidocaine in comparison to wild-type channels. These findings are consistent with the conclusion of Stutzman et al⁵ that lidocaine is ineffective in treating p.F1760C.

The study by Stutzman et al⁵ nicely contributes to multi-decade efforts to find LQT3 variant-specific treatments through preclinical models. An early example is a study on SCN5A variant p.D1790G expressed in HEK293 cells, which revealed that these channels were not inhibited by class 1B antiarrhythmic drugs (ie, lidocaine), whereas the class 1C antiarrhythmic drug flecainide did normalize arrhythmic channel activity.⁷ These findings were confirmed in a cohort of human p.D1790G variant carriers, in whom flecainide but not lidocaine shortened the QTc interval.⁸ These studies highlight the importance of studying the biophysical alterations in Na⁺ channel function induced by LQT3-associated variants.

The findings by Stutzman et al⁵ showing that the class 1B antiarrhythmic drug phenytoin is effective in treating p.F1760C-SCN5A is interesting and, at the same time, somewhat surprising. These findings are consistent with previous studies showing correction of gating defects of the p.Q1475P-SCN5A variant in HEK293 cells⁹ and normalization of the QT interval in vivo in a mouse model of Rett syndrome.¹⁰ However, phenytoin was among the medications attempted in the treatment of the infant patient with the p.F1760C variant; despite this, the clinicians found mexiletine and nadolol to be most effective.⁵ Moreover, because of the location of the p.F1760C variant in the local anesthetic binding site,⁶ one might expect that lidocaine and mexiletine would be less effective in inhibiting the mutant channel.

This study highlights not only the need for genotype- and variant-specific medications but also the power of iPSC-CMs in modeling arrhythmias for therapeutic testing. Although the experimental findings reported by Stutzman et al⁵ are generally consistent with other studies of this and other SCN5A variants, it remains why the patient responded better to mexiletine compared to phenytoin. Additional studies in iPSC-derived organoids or animal models are needed to gain deeper insight into the mechanisms underlying the differential effects of various class 1 antiarrhythmic drugs on mutant SCN5A channels found in LQTS patients.