Primary prevention of arrhythmic events in arrhythmogenic right ventricular cardiomyopathy (ARVC) remains challenging; Cynthia James explains how she and colleagues developed a model for selecting patients suitable for placement of implantable cardioverter-defibrillators (ICDs), and what further refinements are needed.
Arrhythmogenic right ventricular cardiomyopathy (ARVC) is an inherited cardiomyopathy characterised by frequent ventricular arrhythmias and a substantial risk of sudden cardiac death. Once ARVC is diagnosed, a primary management goal is prevention of sudden cardiac death, for which implantable cardioverter-defibrillators (ICDs) are a common consideration. There is general agreement that most ARVC patients diagnosed following an arrhythmic event benefit from a secondary prevention ICD.
However, appropriate patient selection for primary prevention ICDs is challenging. Studies report annual event rates of 2–10% in primary prevention ARVC populations. These young, active patients also have a considerable risk of ICD complications and inappropriate interventions. A meta-analysis and systematic review conducted by our research team found most studies had insufficient statistical power to assess the independent predictive value of potentially correlated risk factors for arrhythmic risk in ARVC.1
To fill this gap, my colleagues and I aimed to develop a robust model for individualised prediction of incident sustained ventricular arrhythmias by including data from an international cohort of ARVC patients. Our findings were initially published online in March 2019 in the European Heart Journal.2 The model can be accessed at www.arvcrisk.com. The success of this project reflects the work of all members of our large international team.
Developing the model
We assembled a cohort of 528 patients with no history of sustained ventricular arrhythmia at diagnosis per 2010 Task Force Criteria from five international registries (Johns Hopkins, Dutch, Nordic, Swiss, and Canadian). The study population had equal representation from North America (n=259, 49.1%) and Europe (n=269, 50.9%).
Our primary outcome was first sustained ventricular arrhythmia. Sustained ventricular arrhythmia was defined as a composite of sudden cardiac death, sudden cardiac arrest, spontaneous sustained ventricular tachycardia (VT) (VT lasting ≥30 seconds at ≥100 beats per minute, or with haemodynamic compromise requiring cardioversion), ventricular fibrillation/flutter, or appropriate ICD intervention.
Potential predictors were prespecified, based on clinical experience and the ARVC risk stratification literature, including our systematic review and meta-analysis. These predictors were: sex, age, recent cardiac syncope (<six months), non-sustained VT, number of premature ventricular complexes on 24-hour Holter monitoring, extent of T-wave inversion on anterior and inferior leads, and right and left ventricular ejection fractions. Each predictor variable was determined at diagnosis.
We used standard statistical modelling methods. The association between the prespecified predictors and the primary outcome was assessed using Cox regression. The model was internally validated using 200 bootstrap samples. We also compared the performance of our model to that of the consensus-based algorithm for ICD placement published in the International Task Force Consensus Statement for Treatment of ARVC including a decision curve analysis.3
During a median follow-up of slightly less than five years, 146 (27.7%) patients experienced incident sustained ventricular arrhythmias. Most events were appropriate ICD therapy (n=102, 70%) or spontaneous sustained VT (n=35, 23.9%). All prespecified predictors except left ventricular ejection fraction were included in the final model, with younger age, male sex, recent syncope, history of non-sustained VT, greater extent of T-wave inversions, higher premature ventricular contractions (PVC) count, and lower right ventricular ejection fraction associated with arrhythmias. The model accurately distinguished patients with and without events, with an optimism corrected C-index of 0.77 (95% confidence interval [CI] 0.73–0.81) and minimal over-optimism (calibration slope of 0.93, 95% CI 0.92–0.95). By decision curve analysis, our model was superior to the published consensus-based ICD placement algorithm. Our model performed better at any risk threshold, and would achieve a 20.6% reduction of ICD placements with the same proportion of protected patients.
Using the largest cohort of ARVC patients with no history of sustained ventricular arrhythmias at diagnosis, we developed a prediction model that generates individualised estimates of the risk of incident arrhythmias using readily available clinical parameters. Importantly, our study does not aim to prescribe ICD placement for a given patient. Instead, we sought to provide clinicians and patients with information to facilitate shared clinical decision-making.
Potential for improvements
There are opportunities to further refine our model. Participation in high intensity aerobic exercise is an established risk-factor for ventricular arrhythmias in ARVC. Whether exercise history or plans can be used to further refine risk prediction is unresolved. Other predictors may also merit consideration including programmed ventricular stimulation results and genotype.
There are also several caveats to the model that should be kept in mind. Firstly, ARVC is a progressive disease so patients should be periodically re-stratified. Secondly, ascertainment of our study population from tertiary care settings may mean the model over-estimates arrhythmia risk in a community-derived population. This highlights the importance of external validation as a next step. And, finally, by including ICD-treated arrhythmias in our outcome the model almost certainly over-estimates risk of sudden cardiac death. This should be kept in mind when considering the risk estimates generated.
Cynthia James is an assistant professor of medicine (cardiology) and a genetic counsellor in the Arrhythmogenic Right Ventricular Dysplasia/Cardiomyopathy Program at the Johns Hopkins University School of Medicine in Baltimore, Maryland, USA.
1. Bosman LP, Sammani A, James CA, et al. Predicting arrhythmic risk in arrhythmogenic right ventricular cardiomyopathy: A systematic review and meta-analysis. Heart Rhythm 2018; 15(7): 1097–107.
2. Cadrin-Tourigny J, Bosman LP, Nozza A, et al. A new prediction model for ventricular arrhythmias in arrhythmogenic right ventricular cardiomyopathy. Eur Heart J 2019; 40(23): 1850–8.
3. Corrado D, Wichter T, Link MS, et al. Treatment of arrhythmogenic right ventricular cardiomyopathy/dysplasia: an international task force consensus statement. Eur Heart J 2015; 36(46): 3227–37.