QRS fragmentation is associated with increased risk of ventricular arrhythmias

Background Fragmentation of the QRS complex (fQRS) on ECG is defined as the presence of additional spikes within the QRS complex. fQRS has been associated with myocardial conduction abnormalities, but whether it predicts ventricular arrhythmias (VA) is uncertain. Purpose To assess the association between the presence of fQRS on standard 12-leads ECG and the risk of VA. Methods In a prospective observational study, we included 243 patients treated with implantable cardioverter-defibrillator (ICD). Baseline ECG was analyzed for fQRS by a trained physician blinded for outcome data. fQRS was defined according to Das (ref) as the presence of an additional R wave (R'), notching in the S wave nadir, or >1 R' in 2 contiguous leads. For wide QRS (≥120ms), fQRS was defined as >2 R waves (R”), >2 notches in the R or S wave in 2 contiguous leads (Figure). Patients were followed at regular ICD controls, and incident ventricular tachycardia (VT), ventricular fibrillation (VF) and treatment with antitachycardia pacing (ATP) or DC-shock were recorded. Results In total, 168 baseline ECG recordings (69%) were interpretable for fQRS, while the remaining were uninterpretable mainly due to low quality and low QRS voltage. The included patients were aged 66±11 years, 14% female, with BMI 27±4 kg/m2 and left ventricular ejection fraction (LVEF) 42±11%. Twenty-two percent had diabetes mellitus (DM), 40% atrial fibrillation, 61% history of myocardial infarction (MI), 81% heart failure and 18% in New York Heart Association Class ≥3. fQRS was present in 16 (10%) patients who had comparable baseline characteristics to those without fQRS, except lower prevalence of DM (p=0.05). Patients with versus without fQRS had comparable QRS duration (p=0.72), QRS axis (p=0.28), corrected QT duration (QTc) (p=0.35) and heart rate (p=0.66). During mean 3.2±0.7 years follow-up 65 (28%) patients had ≥1 VA, including 60 with VT, 21 with VF, and 59 with appropriate ICD-therapy. Presence of fQRS was associated with a 4-fold increased risk of VA (OR 4.15, [95% CI 1.38–12.4], p=0.011). This association persisted after adjusting for age, gender, DM, MI, LVEF and QRS duration (OR 3.99, [95% CI 1.16–13.65], p=0.03). fQRS was strongly associated with incident VT (OR 4.66 [95% CI 1.55–14.0], p=0.006, which persisted after adjustment [p=0.018]), while there was no significant association with incident VF (OR 1.45, [95% CI 0.29–7.09], p=0.64) (Figure). fQRS associated with incident VA irrespective of ICD indication (primary versus secondary, p-for-interaction = 0.80). fQRS was superior to established ECG variables in predicting VA, including QRS-duration, QTc, and presence of Q-waves. Conclusions Interpretation of fQRS in standard ECG is feasible in 70%. fQRS is associated with increased risk of VA, independent of established risk factors, and is an easily available tool that may be useful in identifying patients at increased risk of VA. Funding Acknowledgement Type of funding sources: Public hospital(s). Main funding source(s): Akershus University Hospital fQRS and ventricular arrhythmias