The Postimplantation Electrocardiogram Predicts Clinical Response to Cardiac Resynchronization Therapy

• Published in: Pacing and clinical electrophysiology Vol. 38; no. 5; pp. 572 - 580
• Main Authors: COVERSTONE, EDWARD, SHEEHY, JUSTIN, KLEIGER, ROBERT E., SMITH, TIMOTHY W.
• Format: Journal Article
• Language: English
• Published: United States Blackwell Publishing Ltd 01-05-2015
• Subjects: Aged; biventricular pacing; Cardiac Resynchronization Therapy; Cardiomyopathies - physiopathology; Cardiomyopathies - therapy; Comorbidity; congestive heart failure; electrocardiogram; Electrocardiography; Female; Humans; Male; Predictive Value of Tests; Survival Analysis; Treatment Outcome; biventricular pacing; congestive heart failure; electrocardiogram; cardiac resynchronization therapy
• ISSN: 0147-8389; 1540-8159
• DOI: 10.1111/pace.12609

Background
Biventricular (BiV) pacing for cardiac resynchronization therapy (CRT) is intended to improve left ventricular function by coordinating systolic activity of the septum and free walls. Optimal resynchronization should be manifested by 12‐lead electrocardiogram (ECG) patterns consistent with resynchronized activation, a tall (≥4 mm) R wave in V1, and predominant negative deflection in lead I (RV1SI). We investigated whether the presence or absence of RV1SI predicts heart failure outcomes within 1 year of CRT implant.
Methods
Two independent physicians reviewed the paced ECG of 213 patients post‐CRT device implantation with disputes resolved by a third reviewer. The primary end points of all‐cause death, unplanned hospitalization, left ventricular assist device implant, or transplant within a 1‐year follow‐up were blindly adjudicated according to standard definitions. Groups were compared via Kaplan‐Meier estimates and Cox proportional hazards models to determine association with event‐free survival.
Results
Among CRT patients postimplantation, 56 (26.3%) exhibited the RV1SI pattern on ECG. Patients with the RV1SI pattern were significantly less likely to achieve the primary end point as compared to patients without the RV1SI pattern (33.9% vs 52.2%; Log Rank P = 0.022). This difference was driven by a significantly lower risk for unplanned hospitalization among patients with the RV1SI pattern (hazard ratio = 0.510; confidence interval [0.298, 0.876]). The predictive value remained after adjustment for potential confounders (P = 0.004).
Conclusions
The 12‐lead ECG postimplantation predicts clinical outcomes of BiV pacing. Such prediction may be useful in predicting the need for alternative or advanced heart failure therapies. Further study into ECG patterns may help to prospectively guide CRT.