A prospective evaluation of cardiovascular magnetic resonance measures of dyssynchrony in the prediction of response to cardiac resynchronization therapy

Manav Sohal, Simon G Duckett, Xiahai Zhuang, Wenzhe Shi, Matthew Ginks, Anoop Shetty, Eva Sammut, Sebastian Kozerke, Steven Niederer, Nic Smith, Sebastien Ourselin, Christopher Aldo Rinaldi, Daniel Rueckert, Gerald Carr-White, Reza Razavi

Research output: Contribution to journalArticle (Academic Journal)peer-review


BACKGROUND: Many patients with electrical dyssynchrony who undergo cardiac resynchronization therapy (CRT) do not obtain substantial benefit. Assessing mechanical dyssynchrony may improve patient selection. Results from studies using echocardiographic imaging to measure dyssynchrony have ultimately proved disappointing. We sought to evaluate cardiac motion in patients with heart failure and electrical dyssynchrony using cardiovascular magnetic resonance (CMR). We developed a framework for comparing measures of myocardial mechanics and evaluated how well they predicted response to CRT.

METHODS: CMR was performed at 1.5 Tesla prior to CRT. Steady-state free precession (SSFP) cine images and complementary modulation of magnetization (CSPAMM) tagged cine images were acquired. Images were processed using a novel framework to extract regional ventricular volume-change, thickening and deformation fields (strain). A systolic dyssynchrony index (SDI) for all parameters within a 16-segment model of the ventricle was computed with high SDI denoting more dyssynchrony. Once identified, the optimal measure was applied to a second patient population to determine its utility as a predictor of CRT response compared to current accepted predictors (QRS duration, LBBB morphology and scar burden).

RESULTS: Forty-four patients were recruited in the first phase (91% male, 63.3 ± 14.1 years; 80% NYHA class III) with mean QRSd 154 ± 24 ms. Twenty-one out of 44 (48%) patients showed reverse remodelling (RR) with a decrease in end systolic volume (ESV) ≥ 15% at 6 months. Volume-change SDI was the strongest predictor of RR (PR 5.67; 95% CI 1.95-16.5; P = 0.003). SDI derived from myocardial strain was least predictive. Volume-change SDI was applied as a predictor of RR to a second population of 50 patients (70% male, mean age 68.6 ± 12.2 years, 76% NYHA class III) with mean QRSd 146 ± 21 ms. When compared to QRSd, LBBB morphology and scar burden, volume-change SDI was the only statistically significant predictor of RR in this group.

CONCLUSION: A systolic dyssynchrony index derived from volume-change is a highly reproducible measurement that can be derived from routinely acquired SSFP cine images and predicts RR following CRT whilst an SDI of regional strain does not.

Original languageEnglish
Pages (from-to)58
JournalJournal of Cardiovascular Magnetic Resonance
Publication statusPublished - 1 Aug 2014


  • Aged
  • Aged, 80 and over
  • Biomechanical Phenomena
  • Cardiac Resynchronization Therapy
  • Female
  • Heart Failure/diagnosis
  • Humans
  • Image Interpretation, Computer-Assisted
  • Magnetic Resonance Imaging, Cine
  • Male
  • Middle Aged
  • Myocardial Contraction
  • Patient Selection
  • Predictive Value of Tests
  • Prospective Studies
  • Recovery of Function
  • Reproducibility of Results
  • Time Factors
  • Treatment Outcome
  • Ventricular Dysfunction, Left/diagnosis
  • Ventricular Function, Left
  • Ventricular Remodeling


Dive into the research topics of 'A prospective evaluation of cardiovascular magnetic resonance measures of dyssynchrony in the prediction of response to cardiac resynchronization therapy'. Together they form a unique fingerprint.

Cite this