Cardiac resynchronization therapy (CRT) is now a well-recognized therapeutic option for patients with end-stage heart failure. However, not all patients respond to CRT, and, therefore, preimplantation identification of responders is desirable. The aim of the present study was to investigate whether the degree of left ventricular (LV) dyssynchrony, as assessed with phase analysis from gated myocardial perfusion SPECT (GMPS), can predict which patients will respond to CRT. Methods: Forty-two patients with severe heart failure, depressed LV ejection fraction, and wide QRS complex were prospectively included for implantation of a CRT device and underwent GMPS and 2-dimensional echocardiography as part of the clinical protocol. Clinical status was evaluated using the New York Heart Association (NYHA) classification, 6-min walk test, and quality-of-life score. The histogram bandwidth and phase SD (parameters indicating LV dyssynchrony) were assessed from GMPS, and the clinical status and echocardiographic variables were reassessed at 6-mo followup. Results: Responders (71%) and nonresponders (29%) had comparable baseline characteristics, except for histogram bandwidth (175°6 63°vs. 117°6 51°[P , 0.01]) and phase SD (56.3°6 19.9°vs. 37°.1 6 14.4°[P , 0.01]), which were significantly larger in responders compared with nonresponders. Moreover, receiver-operating-characteristic curve analysis demonstrated an optimal cutoff value of 135°for histogram bandwidth (sensitivity and specificity of 70%) and of 43°for phase SD (sensitivity and specificity of 74%) for the prediction of response to CRT. Conclusion: Response to CRT is related to the presence of LV dyssynchrony assessed by phase analysis with GMPS. A cutoff value of 135°for histogram bandwidth and of 43°for phase SD could be used to predict response to CRT. Larger prospective studies are warranted to confirm the present findings.Key Words: left ventricular dyssynchrony; cardiac resynchronization therapy; SPECT; heart failure J Nucl Med 2007; 48: 1104 -1111 DOI: 10.2967 In the recent years, cardiac resynchronization therapy (CRT) has emerged as a new treatment strategy for a subgroup of patients with end-stage heart failure, a depressed left ventricular ejection fraction (LVEF), and wide QRS complex on the surface electrocardiogram (.120 ms) (1). Although initial studies reported promising results, approximately 20%-30% of the patients do not respond to CRT (1-4). Recent data have indicated that LV dyssynchrony may be mandatory for response to CRT (5,6). A wide QRS complex may not adequately indicate LV dyssynchrony, as 30% of the patients with a wide QRS complex (.120 ms) do not have substantial LV dyssynchrony on echocardiography whereas, on the other hand, approximately one third of the patients with a narrow QRS appear to have substantial LV dyssynchrony on echocardiography (7,8).At present, various imaging techniques are available for the assessment of LV dyssynchrony, including echocardiography with tissue Doppler imaging (TDI), or strain imaging, and MRI (9,10...