SCN5A mutation type and topology are associated with the risk of ventricular arrhythmia by sodium channel blockers

BACKGROUND
Ventricular fibrillation in patients with Brugada syndrome (BrS) is often initiated by premature ventricular contractions (PVCs). Presence of SCN5A mutation increases the risk of PVCs upon exposure to sodium channel blockers (SCB) in patients with baseline type-1 ECG. In patients without baseline type-1 ECG, however, the effect of SCN5A mutation on the risk of SCB-induced arrhythmia is unknown. We aimed to establish whether presence/absence, type, and topology of SCN5A mutation correlates with PVC occurrence during ajmaline infusion.


METHODS AND RESULTS
We investigated 416 patients without baseline type-1 ECG who underwent ajmaline testing and SCN5A mutation analysis. A SCN5A mutation was identified in 88 patients (S+). Ajmaline-induced PVCs occurred more often in patients with non-missense mutations (Snon-missense) or missense mutations in transmembrane or pore regions of SCN5A-encoded channel protein (Smissense-TP) than patients with missense mutations in intra-/extracellular channel regions (Smissense-IE) and patients without SCN5A mutation (S-) (29%, 24%, 9%, and 3%, respectively; P<0.001). The proportion of patients with ajmaline-induced BrS was similar in different mutation groups but lower in S- (71% Snon-missense, 63% Smissense-TP, 70% Smissense-IE, and 34% S-; P<0.001). Logistic regression indicated Snon-missense and Smissense-TP as predictors of ajmaline-induced PVCs.


CONCLUSIONS
SCN5A mutation is associated with an increased risk of drug-induced ventricular arrhythmia in patients without baseline type-1 ECG. In particular, Snon-missense and Smissense-TP are at high risk.


Introduction
Ventricular fibrillation in patients with Brugada syndrome (BrS) is often initiated by premature ventricular contractions (PVCs) [1]. Mutations in SCN5A, the gene encoding the cardiac sodium channel protein Na v 1.5, are an important cause of BrS, and BrS patients with baseline type-1 ECG who carry such mutations have increased risk of PVCs after exposure to sodium channel blockers (SCB) [2]. However, the impact of SCN5A mutations on the risk of drug-induced ventricular arrhythmia in BrS patients without baseline type-1 ECG (BrS in these patients is diagnosed through SCB testing) is unknown. As a result, no guidelines or consensus recommendations exist regarding the use of SCB in SCN5A mutation carriers without baseline type-1 ECG.
Because conduction slowing is a key pathomechanism in BrS and Na v 1.5 is critical for impulse propagation [3], clinical severity should be greatest in patients who carry SCN5A mutations that disrupt Na v 1.5 function the most. Accordingly, we previously showed that nonmissense mutations leading to premature truncation of Na v 1.5 increased the sensitivity of the cardiac conduction system to SCB more than missense mutations, as reflected by more PR and QRS prolongation during SCB testings [4]. Similarly, missense mutations that cause severe loss of Na v 1.5 current (I Na ) caused more conduction slowing than mutations that reduced I Na less. In our study, we derived the magnitude of I Na from published biophysical studies [4]. However, such studies are laborintensive and not available for many mutations. While magnitude of I Na reduction and clinical severity may be easy to predict for non-missense mutations (severe), we hypothesized that this can also be estimated for missense mutations based on their topology. In support of this hypothesis, recent evidence suggests that SCN5A missense mutations affecting the transmembrane or pore regions of Na v 1.5 (severe I Na reduction) are more likely to be pathogenic than mutations in intracellular or International Journal of Cardiology 266 (2018) 128-132 extracellular regions (i.e., N-terminus, C-terminus, interdomain or intersegment linkers) (limited I Na reduction) [5].
In this study, we aimed to establish whether SCN5A mutation presence/absence, type and topology determine the risk of PVC occurrence during SCB (ajmaline) testing in patients without baseline type-1 ECG. Such knowledge may drive clinical management strategies.

Patient inclusion
In this study, we included 416 consecutive subjects (N15-years-old) who had undergone ajmaline testing and SCN5A mutation analysis. No subject displayed type-1 ECG at baseline. Indications for the test were aborted cardiac arrest (ACA), ventricular arrhythmia, syncope, palpitations, family history of BrS and/or sudden cardiac death (FH-SCD), or an ECG suspicious but not diagnostic for BrS.

Mutation analysis, ajmaline testing and ECG analysis
Informed consent was obtained. The study conformed to the ethical guidelines of the 1975 Declaration of Helsinki. Genomic DNA extraction from peripheral blood lymphocytes and SCN5A mutation analysis was performed as described previously [4]. Ajmaline testing was performed using the protocol of the BrS consensus conference [6]. Ajmaline infusion was stopped when type-1 ECG appeared or immediately after occurrence of PVCs. Twelvelead ECGs were analyzed at baseline and peak ajmaline dose (ajmaline peak ; i.e., at maximum dose of~1 mg/kg or at the first dose when type-1 ECG or PVCs occurred). Ajmaline testing was considered positive if type-1 ST elevation ≥2-mm appeared in ≥1 right-precordial lead [6].

Statistical analysis
Differences between groups were compared using Fisher exact test or χ [2] test (categorical variables), or Student t-tests or analysis of variance (continuous variables). Homogeneous subsets of groups were determined with the Standardized Residual Methods (categorical variables) and Student-Newman-Keuls post hoc multiple comparison of groups (continuous variables). For ECG parameters, since multiple tests were performed, the significance level was set at 0.001. In the Tables, homogeneous subsets (no statistical difference) are indicated by an equals (=) sign. A logistic regression analysis was performed to identify predictors for PVCs during ajmaline infusion, and variables with P b 0.05 were selected for multivariable analysis. A correction for the relatedness among individuals was applied and the linearity assumption for the numerical predictors was checked. Results of the logistic regression are expressed as odds ratio (OR) with confidence interval (CI). Data are expressed as number (percentage) or mean ± standard deviation (SD), where appropriate.
At baseline, S + had longer PR than S − , while baseline heart rate (HR), QRS and QTc did not differ. Ajmaline peak was lower in S + than S − . At ajmaline peak , S + had slower HR and longer PR and QRS than S − . Ajmaline induced type-1 ECG more often in S + than S − (59 [67.0%] vs. 112 [34.1%], P b 0.001).

Comparisons between patients with different SCN5A mutations and non-carriers
To further study the role of the SCN5A mutation on the occurrence of ajmaline-induced PVCs, we compared patients with non-missense mutations (S non-missense ; n = 14), patients with missense mutations in transmembrane/pore regions (S missense-TP ; n = 41), patients with missense mutations in intra−/extracellular regions (S missense-IE ; n = 33), and S − ( Table 2). Except for FH-SCD, other clinical characteristics did not differ between groups.
At baseline, PR was longer in S non-missense and S missense-TP than S missense-IE and S − . Other baseline ECG parameters did not differ between the groups. Ajmaline peak was lower in S non-missense and S missense-TP than S missense-IE and S − . At ajmaline peak , S non-missense had slower HR and longer QRS than other S + and S − . The proportion of patients with ajmaline-induced type-1 ECG did not differ between different mutation groups.
Ajmaline induced PVCs more often in S non-missense and S missense-TP than S missense-IE and S − . Expect for baseline PR (212 ± 33 ms. in S non-missense , 223 ± 21 in S missense-TP , 175 ± 16 ms. in S missense-IE , and 147 ± 26 ms. in S − , P b 0.001), other ECG parameters at baseline or ajmaline peak and clinical characteristics did not differ between patients with PVCs in different groups (i.e., S non-missense and S missense-TP , S missense-IE , and S − ).

Discussion
In the last 15 years, several studies have attempted to identify predictors for drug-induced ventricular arrhythmia in patients undergoing SCB testing. However, these studies did not investigate SCN5A mutation status and/or included patients with baseline type-1 ECG (who are wellrecognized to be at high risk for adverse events in the presence of SCB) [2,7]. As a result, it is still unknown whether carriers of a SCN5A mutation without baseline type-1 ECG are at higher risk of druginduced ventricular arrhythmia.
In this study, we first systematically studied the impact of SCN5A mutation on the occurrence of PVCs during ajmaline testing in patients without baseline type-1 ECG. We found that presence of SCN5A mutation increases the likelihood that ajmaline exposes BrS and evokes PVCs, and that the risk of PVCs in S + is regardless of the occurrence of type-1 BrS ECG during ajmaline testing (i.e., the occurrence of PVCs was similar between Ajmaline positive -S + and Ajmaline negative -S + and higher than Ajmaline positive -S − [ Table 1]). In line with earlier data [8], S − often had PVCs with LBBB morphology and inferior axis, suggesting a right-ventricular outflow tract origin, while S + had PVCs originating from both left and right ventricle.
In addition, we found that the risk for drug-induced PVCs depends on the type and topology of the SCN5A mutation. Multivariable analysis identified S non-missense and S missense-TP mutations as strong predictors for ajmaline-induced PVCs. Moreover, although the effect was weaker, baseline heart rate and QRS after ajmaline were also identified as predictors for PVCs. Ajmaline-induced type-1 ECG was not associated with PVC occurrence and did not predict PVCs, and the proportions of patients who developed type-1 ECG did not differ between mutation groups. Based on these findings, one may speculate that the mechanism underlying ajmaline-induced PVCs is large reduction of I Na and the depolarization reserve in the heart [3,4,9]. I Na reduction by S non-missense or S missense-TP mutations may be significant, but depolarization reserve is large enough to ensure that QRS in S non-missense and S missense-TP at baseline is not different from S missense-IE or S − . However, the added presence of ajmaline results in further I Na reduction and decline in the depolarization reserve to a level that results in QRS prolongation [4] and ultimately PVCs [9]. Since ajmaline requires repetitive opening and closing of Na v 1.5 to act as a blocker ('use-dependent block'), Na v 1.5 is blocked more potently at higher heart rates. Together these processes may explain why SCN5A mutation type and topology, higher baseline heart rates and longer QRS intervals after ajmaline are associated with the risk for ajmaline-induced PVC occurrence.
The concept that the magnitude of I Na reduction due to a SCN5A mutation plays a crucial role in the occurrence of ajmaline-induced PVCs is in line with our finding that the peak ajmaline dose was lower in S non-missense and S missense-TP than S missense-IE or S − , while the proportion of patients with PVCs was much larger in S non-missense and S missense-TP . This suggests that S non-missense and S missense-TP mutations cause more I Na reduction and therefore require less Na v 1.5 block by ajmaline for PVCs to occur. Interestingly, the same concept may also apply for the occurrence of ajmaline-induced type-1 ECG. While the proportion of ajmaline-induced type-1 ECG did not differ between S non-missense , S missense-TP , and S missense-IE , the required peak ajmaline dose was higher in the latter, suggesting that the degree of I Na reduction may also play an important role in the pathophysiology of BrS, and that S non-missense and S missense-TP are at higher risk of developing BrS in the presence of SCB than S missense-IE (or S − ). The limitations of our study include its retrospective design and the small size of various SCN5A mutation groups. In addition, although type-1 ECG was absent in all patients on at least two time points, we cannot exclude the presence of transient type-1 ECG, e.g., during fever. Moreover, the study population was not screened for large genomic rearrangements or mutations in other genes that have been anecdotally linked to BrS. However, in this regard it is important to note that in an earlier genetic screening study in 38 BrS patients from our center, we have excluded large genomic rearrangements and mutations in other BrS-linked candidate genes [10].

Conclusions
The presence of a SCN5A mutation increases the likelihood that ajmaline exposes BrS and evokes ventricular arrhythmia in patients without baseline type-1 ECG. Moreover, the risk for drug-induced arrhythmia depends on type and topology of the SCN5A mutation, and patients with S non-missense or S missense-TP mutations may be at highest risk. We recommend SCN5A mutation analysis in individuals who experience ventricular arrhythmia while using SCB, particularly if arrhythmia is preceded by QRS prolongation, also in patients in whom type-1 ECG has not been achieved (yet). Moreover, we discourage prescription of SCB in patients with SCN5A mutation, including patients without BrS. S + , SCN5A mutation carriers; S − , patients without SCN5A mutation; non-missense, non-missense mutations; missense-TP, missense mutations in transmembrane segments or pore region of Na v 1.5; missense-IE, missense mutation in intra-or extracellular regions of Na v 1.5. HR, heart rate; PVC, premature ventricular contractions; QTc, heart rate-corrected QT interval; SCD, sudden cardiac death. Data are expressed as number (percentage) or mean ± SD. N indicates number of patients. Δ indicates differences between ECG values at peak ajmaline dosevalues at baseline. P values indicate results of statistical comparisons between the four groups (columns). In case of an overall statistical significant difference, homogeneous groups (with no statistical difference) are indicated by an equals (=) sign.