Abstract 3555: Late Decline in Left Ventricular Function After Initial Improvement Following Cardiac Resynchronization Therapy
Background: Response to cardiac resynchronization therapy (CRT) is most often reported in clinical trials as a dichotomous outcome, e.g. responder or non-responder. Long-term maintenance of response has not been previously examined. We describe a series of patients in whom a favorable echocardiographic response was temporary.
Methods: Our cardiology database was queried for all patients who underwent CRT from 2000 through 2008, and had at least two post-operative echocardiograms at our institution. Response was considered favorable if there was an absolute increase of 10% in left ventricular ejection fraction (LVEF) compared with the preoperative exam. Re-decline was defined as an absolute decrease of 10% from the peak LVEF. Once re-decline was identified, the echocardiograms were reviewed side-by-side to confirm a true effect.
Results: We identified 110 patients who underwent CRT followed by at least two echocardiograms. Of the 46 patients who had a positive initial response, seven (15%) subsequently developed a re-decline in LVEF. Baseline LVEF was 25% (±7%) in the sustained LVEF group and 29% (±7%) in the re-decline group (p=0.13). Peak LVEF was 46% in both groups (±11% sustained, ±12% re-decline group, respectively, p=0.98). Median time from CRT implant to re-decline was 722 days [range: 139, 1078]. Ischemic etiology, age, sex, hypertension, pre-implant LVEF, LV diastolic diameter and medications were all comparable in both groups. Diabetes was present in 39% with sustained improvement and 86% of patients with re-decline (p=0.02). In multivariate modeling, diabetes was the only significant predictor for re-decline amongst those with initial improvement. There was no increased incidence of death or CHF admissions (p=0.31).
Conclusion: This is the first report of re-decline in LVEF after initially favorable echocardiographic response to CRT. When evaluating the effect of CRT, duration of response should be taken into account. Potential clinical risks associated with LVEF re-decline warrants further investigation.