PhD Scientific Days 2026

Cardiovascular Medicine and Research 1.

Identifying Patients at Risk of Slow Recurrent Ventricular Tachycardia: A Multicenter Clinical Prediction Model for Optimizing ICD Programming

Name of the presenter

Szakál, Imre

Institute/workplace of the presenter

Semmelweis University Heart and Vascular Center

Authors

Imre Szakál¹, Ferenc Komlósi¹, Mihály Szőnyi², Dávid Pileczky², Bence Arnóth¹, Zsófia Varga¹, Gyula Bohus¹, Béla Merkely¹, László Gellér¹, Attila Kardos², Vivien Klaudia Nagy¹
1: Semmelweis University Heart and Vascular Center
2: Gottsegen National Cardiovascular Center

Text of the abstract

Introduction: ICD programming in secondary prevention of VT remains largely empirical, particularly regarding the lowest detection threshold and therapy zones. Although current guidelines recommend setting the VT zone at least 10 bpm below the documented VT rate, a subset of patients experience recurrent VTs with a slower rate than the detection threshold, resulting in undetected arrhythmias.

Aims: To develop a clinical prediction model specifically aimed at identifying secondary prevention patients at risk of recurrent ventricular tachycardia below conventional ICD detection thresholds.

Methods: Data was obtained from VT databases of two national cardiology centers and retrospectively analyzed. Patients with structural heart disease hospitalized for sustained monomorphic VT were followed for 12 months and categorized according to recurrence characteristics. The 185 bpm threshold reflects commonly applied conventional VT detection programming in contemporary secondary prevention ICD practice. Group 1 included patients who experienced a recurrent VT with a rate below 185 bpm and a frequency decrease of at least 10 bpm, while Group 2 included patients a) without recurrence or b) with a recurrent VT above 185 bpm or c) with less than 10 bpm rate change compared to the index VT. Multivariable logistic regression was used to identify independent predictors. Backward conditional stepwise approach was used to determine the optimal predictor combination to create a multivariate logistic regression model.

Results: A total of 302 patients were included (Group 1: n=46; Group 2: n=256). Electrical storm or incessant VT (OR 3.91, p<0.001), heart failure with reduced ejection fraction (OR 3.59, p=0.04), and moderate-to-severe mitral regurgitation (OR 2.19, p=0.04) independently predicted slow recurrent VT. Left ventricular end-systolic diameter >50 mm demonstrated a clinically relevant trend (OR 2.02, p=0.11). The model showed good discriminative performance (AUC = 0.81), with balanced sensitivity and specificity (71.7% and 71.5%).

Conclusion: We successfully identified structural and arrhythmic factors associated with a markedly slower recurrent VT in patients with structural heart disease. The identified predictors consistently reflected advanced structural remodeling and arrhythmogenic substrate progression.