QRS-microfragmentation and non-planarity as predictors of clinical outcome and left ventricular ejection fraction improvement following cardiac resychronization therapy

Abstract Background Predicting response to cardiac resynchronization therapy (CRT) remains challenging. QRS-microfragmentation (QRSµf) and QRS non-planarity (QRSnp), novel ECG-derived parameters which quantify subtle abnormalities in ventricular depolarization, have emerged as potent predictors of adverse outcomes in cardiac patients. Purpose To determine whether QRSµf and QRSnp can predict clinical outcome and left ventricular ejection fraction (LVEF) improvement in CRT recipients. Methods In a single-center retrospective study, including all patients who received CRT in a tertiary care hospital, mean QRSµf and QRSnp were calculated on the last ECG before CRT implantation. The primary endpoint was a composite of all-cause mortality, heart transplantation (HTX), or implantation of a ventricular assist device (VAD). Secondary endpoints were first hospitalization for heart failure, and a ≥5% increase in LVEF after CRT. QRSµf and QRSnp were analyzed after stratification into tertiles. Time-to-event endpoints were analysed using Kaplan-Meier analysis with log-rank testing and adjusted Cox proportional hazard regression. LVEF was analyzed using adjusted logistic regression. Variables included in the models were selected based on clinical relevance using a forward stepwise approach. Harrel's C-indices were calculated for the first two endpoints and AUC for the third, for the baseline model and after adding QRSµf and QRSnp. Results A total of 1,164 CRT recipients were enrolled, median age of 70.5 years (IQR 62.8–76.7) and median LVEF of 28% (IQR 21–34). Of these, 306 (26.3%) were female, 516 (44.3%) had ischemic cardiomyopathy, 701 (60.2%) received a CRT-D, and 655 (56.3%) had left bundle branch block. The median follow-up was 4.4 years (IQR 2.2–7.4). The ranges of the tertiles for QRSµf and QRSnp, along with the results, are presented in Table 1. The highest tertile of QRSµf (HR 1.38, 95% CI 1.09-1.76, p = 0.008) and QRSnp (HR 1.34, 95% CI 1.08-1.68, p = 0.008) were associated with an increased incidence of the primary endpoint, although the increase in C-index was limited. The highest tertile of QRSµf (HR 1.52, 95% CI 1.13-2.05, p = 0.006) and QRSnp (HR 1.61, 95% CI 1.22-2.13, p = 0.001) were associated with heart failure hospitalizations, with a ΔC-index of 0.012 when both parameters were combined. Kaplan-Meier and cumulative incidence curves are shown in Figure 1. The highest tertile of QRSµf (OR 0.41, 95% CI 0.28-0.59, p < 0.001) and QRSnp (OR 0.50, 95% CI 0.35-0.71, p < 0.001) were associated with an absence of LVEF improvement, with ΔAUC 0.029 when both parameters were combined. Conclusion This study highlights the potential of QRSµf and QRSnp in predicting clinical outcomes, as well as LVEF improvement following CRT. Prospective data are needed to assess their practical applications in clinical settings.Table 1. Figure 1.