Long QT syndrome (LQTS) is caused by mutations in genes encoding cardiac potassium and sodium ion channel subunits or cellular structural proteins. Patients often present with symptoms at a young age and are at high risk of nonfatal (syncope) and fatal (sudden cardiac death [SCD]) cardiac events (1). The incidence of syncope in LQTS patients is approximately 5% per year (1), depending on the mutation causing the syndrome (2- 3), whereas the incidence of SCD is much lower, approximately 1.9% per year (1). However, nonfatal events remain the strongest predictor of fatal events in LQTS patients (1,4), and the overall risk of subsequent SCD in an LQTS patient who has experienced a previous episode of syncope is approximately 5% per year (1). Thus, an LQTS patient who presents for clinical assessment after a nonfatal syncopal episode is already at high risk of a subsequent LQTS-related fatal event. Recent studies from the International LQTS Registry identified risk factors for cardiac events in LQTS patients (1,3,5- 9). However, there are no data regarding specific risk factors for SCD within the high-risk subgroup of symptomatic LQTS patients who have experienced a previous syncopal episode. Thus, the subsequent treatment of these patients depends largely on the clinical judgment of the physician based on risk assessment. Specifically, a paucity of data exists regarding the management of LQTS patients who experience syncope while on beta-blocker therapy. The aim of this study was to determine clinical predictors of subsequent SCD in LQTS patients presenting to the clinician with first syncopal episode and to evaluate the efficacy of beta-blocker therapy for the prevention of sudden death in this high-risk population.

The study population was drawn from the International LQTS Registry (10) and included affected or genotype-positive individuals born after 1959 to maximize the number of patients who were treated with beta-blockers. Patients were followed through age 41 years. Affected individuals were defined as any subject with a corrected QT (QTc) interval of ≥450 ms, as corrected by Bazett's formula (11), who experienced a syncopal episode. The final study group comprised 1,059 LQTS subjects from 764 families. The LQTS genotype was determined with standard mutational analytic techniques involving 5 established genetic laboratories associated with the International LQTS Registry. Genotype data were available for 445 patients (LQT1 = 212, LQT2 = 163, LQT3 = 35, LQT5 = 4, LQT6 = 3, LQT7 = 2, LQT8 = 1; genotype-negative affected = 36). Symptomatic genotype-negative subjects according to the above criteria were included if incomplete genetic studies had been performed.

Beta-blocker therapy was initiated at the discretion of each patient's attending physician. During the initial patient contact, information was collected on whether beta-blocker treatment had been started, the specific beta-blocker initiated, the date started, the prescribed dose, and the patient's weight. At subsequent yearly contacts, information was recorded on whether the patient continued taking beta-blockers and, if so, the daily dose; if patients discontinued therapy, the date that the medication was stopped was recorded. Among patients who died, we retrospectively determined whether the patient had been taking a prescribed beta-blocker before and on the day of death.

Episodes of loss of consciousness were categorized as syncope if the episode was abrupt in onset and offset. Patients were classified into 3 prespecified categories based on the clinical nature of the syncopal events: 1) a first syncopal event in patients not receiving beta-blocker therapy; 2) repeated syncopal events in patients not receiving beta-blocker therapy; and 3) any syncopal event occurring in patients receiving beta-blocker therapy. Patients in the last category could have had any number of syncopal episodes while off beta-blocker therapy before the final episode while on beta-blocker therapy. Once a patient experienced a syncopal event while receiving beta-blocker therapy, the patient remained in this group independently of future syncopal events and treatment.

The primary end point was a life-threatening cardiac event. Twenty percent (n = 212) of the study population had an implantable cardioverter-defibrillator (ICD) implanted. We therefore used the end point of severe arrhythmic events (SAEs) defined as LQTS-related SCD, aborted cardiac arrest, or appropriate ICD therapy for an LQTS-related ventricular tachyarrhythmia, whichever occurred first. Adjudication of the ICD treatment as appropriate or inappropriate was performed by the treating electrophysiologist at the time of ICD interrogation.

Variables were tested for normality using visual inspection. Student t test and Pearson's chi-square test were used in the univariate comparison analyses where appropriate. The cumulative probability of a first cardiac event was assessed by the Kaplan-Meier method with significance testing by the log-rank statistic. The Cox proportional hazards survivorship model was used to evaluate the independent contribution of clinical and genetic factors to the first occurrence of time-dependent cardiac events from birth through age 40 years. Pre-specified covariates included in the multivariate model were QTc duration, sex, history of syncope, and time-dependent beta-blocker therapy. Beta-blocker treatment, syncope, and the interaction between recurrent syncope and beta-blocker therapy were treated as time-dependent covariates in a Cox model, and all reported hazard ratios (HRs) and p values stem from these models.

To illustrate the risk associated with syncopal events occurring while on and off beta-blocker therapy, Kaplan-Meier survival curves for patients, all experiencing 1 syncopal event while off beta-blocker therapy, were created for the following treatment and syncopal groups: 1) patients not starting beta-blocker treatment after first syncope; 2) patients starting beta-blocker treatment after the first syncopal event and experiencing no subsequent syncope during beta-blocker treatment; and 3) patients starting beta-blocker treatment after the first episode of syncope and experiencing subsequent syncope episodes during beta-blocker treatment. All patients were initially in group 1, and the time of the syncope occurring while off beta-blocker therapy was used as time origin. If patients started beta-blocker therapy, they moved into group 2, now using time of initiation of beta-blocker treatment as the time origin for outcome. If patients in group 2 experienced syncope during beta-blocker treatment, they moved into group 3, now with the time of the syncope occurring while receiving beta-blocker treatment as the time origin.

Similarly, the figure showing the risk of syncope occurring during beta-blocker treatment was constructed with patients in the corresponding age and sex groups at the time of beta-blocker treatment initiation. If the patients started beta-blocker treatment before age 14 years and were followed past age 14 years, the patient was censored at age 14 years and restarted at time 0 in the appropriate sex group with the 14th birthday as the origin of the curve. The methodology shown in (Figure 1) and (Figure 2) was used for illustrative purposes only, and no hypothesis testing was done using this approach.

All statistical analyses were performed using SAS version 9.1.3 (SAS Institute, Cary, North Carolina).

Baseline characteristics of the study population by the occurrence of SAEs during follow-up are shown in (Table 1). The group with SAEs had a lower frequency of beta-blocker use, but initiated beta-blocker therapy at a younger age than those without SAEs. A larger proportion of patients with SAEs was treated with device therapy or surgery, indicating the severe clinical presentation of the syndrome in this group. The type and dose of beta-blocker treatment were balanced in the 2 groups.

A total of 210 SAEs occurred, of which 82 (39%) occurred during beta-blocker treatment. There were no differences in the proportion of the SAEs that occurred during beta-blocker treatment among patients with LQT1 (45%), LQT2 (33%), and LQT3 (40%) (p = 0.48).

The most important risk factor for SAEs was whether a syncopal episode occurred during beta-blocker treatment. This is illustrated in (Figure 1). Patients who began beta-blocker therapy after their first and only syncopal episode and did not experience further episodes were at low risk of SAEs. However, patients experiencing syncopal episodes during beta-blocker therapy were at the same high risk of SAEs as patients who never started beta-blocker therapy. Accordingly, in a multivariate analysis (Table 2), syncope occurring during beta-blocker treatment was the most powerful predictor of subsequent SAEs (HR: 3.6, p < 0.001). Patients who experienced multiple versus single syncopal episodes while off beta-blocker treatment had twice the risk of an SAE. Beta-blockers were generally protective against SAEs, and there were no significant sex or age group interactions. The risk of SAEs after a syncopal event was also significantly increased among patients with severe QTc interval prolongation (QTc interval >500 ms) and female patients in the 14 to 40 years age group. Females and males have a similar risk of SAEs after the first syncopal episode during the preteen years, but after age 14 years, female patients had almost twice the risk of SAEs compared with male patients in the same age group (HR: 1.86, p < 0.001).

Fifty-two patients were treated with left cervicothoracic sympathectomy (LCTS) during the course of the study. All patients started beta-blocker therapy before LCTS, and most patients (43 [83%]) remained on beta-blocker therapy throughout the study. Six SAEs occurred in this group despite concomitant treatment with beta-blockers. The patients receiving LCTS had longer QTc intervals, both among subjects with SAEs (QTc interval = 519 ± 5 ms) and without SAEs (QTc interval = 520 ± 5 ms), but were in other aspects similar to the study population. The few individuals with sympathectomy did not allow evaluation of this treatment in the Cox models.

To determine the risk for recurrent syncope while receiving beta-blocker therapy, 746 patients in whom beta-blocker therapy was initiated after experiencing syncope were included in a subset analysis. In this analysis, follow-up time was assessed from the date beta-blocker therapy was initiated.

As illustrated in (Figure 2), the risk of syncope during beta-blocker treatment did not show any association with the QTc interval, but the risk was markedly influenced by the age and sex of the patients. (Figure 2) shows a high but similar risk of syncope during beta-blocker treatment before puberty in both sexes. However, after puberty, female patients remain at high risk, whereas the risk in male patients decreases markedly. (Table 3) shows the results from the multivariate analysis. There were no significant differences between male and female patients ages 0 to 13 years (HR: 1.04, p = 0.85) as well as between female patients ages 0 to 13 years and female patients ages 14 to 40 years (HR: 1.39, p = 0.10).

This study highlights the association of syncopal episodes with the subsequent risk of potentially fatal arrhythmic events in LQTS patients. New important findings in this study are as follows: 1) in LQTS patients presenting with the first syncopal episode, fatal arrhythmic events are effectively prevented with beta-blocker treatment in those without recurrent syncope; 2) patients experiencing syncope while receiving beta-blocker therapy are at high risk of subsequent SAEs, a risk similar to that observed in patients who are not treated with beta-blockers; and 3) there is an important sex difference in the risk of experiencing syncope while being treated with beta-blockers. Before puberty, the efficacy of beta-blockers in preventing subsequent syncopal episodes seems to be equal in both sexes, whereas after age 14 years, this risk is drastically lowered in male patients, but not among female patients. The risk of syncope while being treated with beta-blockers among patients with previous syncopal events does not seem to be related to the QTc interval.

Why some patients keep having symptoms despite treatment with beta-blockers is unknown. A possible explanation may be the known patient variability in beta-blocker efficacy in blocking sympathetic activation (12- 13) that may have genetic underpinnings. In a previous study, failure of beta-blocker therapy was related to the genotype, because LQT1 genotype-positive subjects showed the highest proportion of beta-blocker therapy failures, and to the type of beta-blocker used (14). This finding contrasts with our study in which the type of beta-blocker did not significantly influence the results, and beta-blocker effects were consistent across genotypes. Instead we found that sex and age had an influence on the risk of syncope while receiving beta-blocker treatment. Those experiencing syncope on beta-blocker therapy were at high risk of SAEs.

A recent study evaluated occurrences of aborted cardiac arrest/SCD in LQTS subjects receiving beta-blocker therapy and found that a significant number of these events were due to noncompliance or concomitant treatment with QT-prolonging drugs (15). We were not able to investigate this, but noncompliance is an important confounder in this population consisting of mainly young individuals prone to side effects. However, we cannot explain why noncompliance should be much higher in female patients older than 14 years than in male patients older than 14 years, and we believe that other factors such as sex hormones are likely to play a role in this difference.

When to treat a symptomatic LQTS patient with an ICD is an important clinical question, and the benefits and risk of ICD therapy in high-risk LQTS patients have yet to be defined. The risk of SAEs in LQTS patients presenting with syncope is low if treated with beta-blockers. However, experiencing syncope while being treated with beta-blocking agents is a high-risk situation, and this study shows that the risk of fatal arrhythmias in such patients can be considered equal to the risk in patients not treated with beta-blockers. Even though some of the syncope episodes in this study could have occurred because of noncompliance or undertreatment, it is unlikely that these nontherapy factors explain our findings. ICD therapy is very effective in preventing SCD in LQTS patients (16- 18), and ICD therapy should be considered in patients experiencing syncope during beta-blocker treatment.

Beta-blocker treatment was not allocated at random, and unmeasured factors could have influenced the effects of therapy. Also, the efficacy of beta-blockers has been linked to the genotype of the patients. Only a subset of the study subjects in this study was genotyped, and the small number of end points in these patients did not allow us to address differences between the different genotypes. We did separate models for LQT1 and LQT2 patients and found identical patterns for the beta-blocker treatment. We believe that the bias caused by unknown phenotype is small and that the reported results are applicable to most genotypes. Family membership of the study subjects is likely to be influenced by other genotypic traits in the family. In this study, only a few study subjects were related, and we did not find any difference in the results when using the covariance estimator sandwich (19) to adjust for family membership. The impact of LCTS surgery on the study results could not be fully evaluated due to limited power in the Cox analysis.

In general, LQTS patients presenting with syncope are effectively treated with beta-blockers. However, patients experiencing ≥1 syncopal events during beta-blocker therapy are at the same risk of fatal events as patients who were not treated with beta-blockers. Thus, ICD treatment should be considered in these high-risk patients. The risk of syncopal events during beta-blocker treatment is highest before puberty. After puberty, the risk remains high in female patients.

For a list of the investigators from the International Long QT Syndrome Registry who contributed patients to the study, please see the online version of this article.

Risk of Fatal Arrhythmic Events in Long QT Syndrome Patients After Syncope