Author + information
- ↵⁎Reprint requests and correspondence:
Dr. Kalyanam Shivkumar, UCLA Cardiac Arrhythmia Center, UCLA Health System, David Geffen School of Medicine at UCLA, Suite 660, Westwood Boulevard, Los Angeles, California 90095-1679
The assessment of risk of sudden death (SCD) in patients with long QT syndrome (LQTS), especially those with mild-to-moderate corrected QT interval (QTc) prolongation, is the most challenging aspect of caring for such patients. The incidence of SCD in patients with a normal QTc interval during the first 40 years of life is about 4% (1–3), underscoring the need for risk stratification techniques to identify patients at high risk. Current risk assessment in LQTS integrates clinical and genetic features known to be associated with SCD. The unmet need in the clinical setting is the ability to more accurately assess risk. This could lead to more appropriate use of therapies, potentially saving lives and avoiding unnecessary treatment.
Features associated with high risk in LQTS include history of aborted cardiac arrest, history of syncope, marked QT prolongation, and combinations of age, sex, and LQTS genetic subtype. Boys are at higher risk than girls, but during adolescence, the sex–risk relationship changes, and women have a higher risk than men. Women with LQTS type LQT2 have an especially high risk, which is even more pronounced during the postpartum period. Syncope is less common in patients with LQT3, who not infrequently present with SCD as their first symptomatic episode (3).
There is a reported 4-fold increase in risk of aborted cardiac arrest or SCD among patients with a QTc interval ≥500 ms (3,4). However, QTc intervals can vary temporally and among individuals with the same mutation. The risk associated with QTc prolongation is more pronounced among women than men. In women, risk increased with QTc intervals >500 ms as compared with men, where risk increased with QTc intervals >550 ms. Mutation characteristics have recently been shown to determine cardiac risk in patients with confirmed genotypes for LQTS but normal range QTc intervals (5). This suggests a strong genetic component to cardiac risk that is not currently understood.
Costa et al. (5) recently reported that in patients with LQT1, with mutations localized to the membrane-spanning domains and the cytoplasmic loops of the KCNQ1 protein, have a greater QTc prolongation during exercise, thereby increasing risk of cardiac events even during mild exercise. These findings suggest that a genotype-specific approach, incorporating clinical and mutation location/functional data, might further improve the risk assessment and management of patients with the most common genetic subtype of LQTS.
Heterogeneity of repolarization is increasingly thought to be a significant determinant of risk of SCD in LQTS. Markers of heterogeneity of repolarization identified are microvolt T-wave alternans, increased T-wave peak to T-wave end interval, and more recently, dispersion of mechanical contraction time (6,7). It is yet to be determined how such markers can be incorporated into models of risk prediction in LQTS.
In a study published in this issue of the Journal, Hoefen et al. (8) have demonstrated that computational models can help estimate risk in patients with LQTS. They developed an elegant “in silico” 1-dimensional electrocardiographic (ECG) computer model into which cellular electrophysiology functions from wild type and LQT1 mutants was introduced. The study consisted of a single population of older LQTS patients and did not include high-risk patients who died at a young age. The transmural repolarization (TRP) for each mutation was evaluated in the computer model and related to adverse cardiac events among patients with these mutations. They reported that a TRP increase of 10 ms was associated with a 35% increment of adverse cardiac events. More significantly, they showed that among patients with mild-to-moderate QTc duration (<500 ms), the risk associated with TRP was maintained, whereas the patient's individual QTc interval was not associated with a significant risk increase after adjustment for TRP. They also demonstrated that simulated TRP could identify high-risk mutations.
In this study, it is unclear whether the model-computed TRP and the measured QTc data are independent predictors; in addition, the relationship between the transmural ECG and the body surface ECGs needs to be better defined. Computer modeling using higher-dimensional tissue structure is necessary to study arrhythmia mechanisms in detail, and therefore, the 1-dimensional model used in this study can only be considered a first step. This study illustrates the role of computational modeling in risk assessment in patients with mild-to-moderate QTc prolongation where the greatest challenge lies in risk prediction.
Modeling of electrophysiological parameters started 50 years ago with the Hodgkin-Huxley model of a giant squid axon (9). Since that time, the models have undergone extensive modifications and have reached a high degree of physiological detail, and so can be used to predict consequences of complex changes in channel function to the overall heart rhythm. Although computational models allow experimentation with complex biological systems, ultimately, they are only as good as the biological/experimental data that strengthen them and bring them closer to reality.
Data from voltage clamp experiments are the foundation for developing a computational model. However, it is well recognized that there is variability in the same species across different levels of organization, such as differential expression of ion channels, in the action potential, and at the organ level in the range of heart rate, ECG measurements, and heart rate and response to medications (10–12). In addition, different combinations of electrophysiological parameters can generate voltage traces that look virtually indistinguishable (13). Most models simplify the input data (which is necessary for computing), and thereby discrepancies can be great when such modeling experiments are performed to generate predictions. Models that include multiple experimental results would be expected to generate more robust predictions (14). The ability of the computational model to truly allow robust, accurate, and reproducible throughput testing of ion channel behavior in the clinical situation is therefore unclear. A variable such as the activation of the autonomic nervous system can profoundly alter cellular- and organ-level behavior of electrical propagation and thus impact arrhythmogenesis (15).
We live in a world where mathematical modeling already makes a great impact on our daily lives. For instance, weather prediction, a necessity of life, is surprisingly accurate and is a testament to mathematical modeling (16). It is also an example of a situation where models have been improved by constant feedback from real-life data, thereby improving predictions. The fact that sometimes such predictions fall short (despite years of improvement in that field) highlights the need for learning about new variables that need to be discovered in nature. This type of a progression toward meaningful modeling is bound to happen in biology as clinical and scientific data are used to improve models, and modeling in turn prompts new research strategies. Creating a tool for the clinician is a clear goal of such efforts. The present paper is a very useful step in this direction.
The results of the study by Hoefen et al. (8) are probably the first report of computer modeling to validate risk demonstrated in a real patient population. Using a relatively simple model to validate clinical data illustrates the potential of computer modeling to predict risk in rare conditions where prospective trials are difficult to conduct or may never be done.
Supported by the National Heart, Lung, and Blood Institute (grant R01HL084261 to Dr. Shivkumar).
Dr. Madapati has reported that he has no relationships relevant to the contents of this paper to disclose.
↵⁎ Editorials published in the Journal of the American College of Cardiology reflect the views of the authors and do not necessarily represent the views of JACC or the American College of Cardiology.
- American College of Cardiology Foundation
- Goldenberg I.,
- Horr S.,
- Moss A.J.,
- et al.
- Moss A.J.,
- Schwartz P.J.,
- Crampton R.S.,
- et al.
- Haugaa K.H.,
- Smedsrud M.K.,
- Steen T.,
- et al.
- Hoefen R.,
- Reumann M.,
- Goldenberg I.,
- et al.
- Antzelevitch C.,
- Sicouri S.,
- Litovsky S.H.,
- et al.
- Kannankeril P.,
- Roden D.M.,
- Darbar D.
- Lynch P.