br Conclusions br Financial support br Conflict of interest

Conclusions
Financial support
Conflict of interest
Acknowledgment
Introduction Cardiac resynchronization therapy (CRT) has been demonstrated to improve symptoms, quality of life, functional status, and reduce mortality and olda failure-related hospitalizations in patients with New York Heart Association (NYHA) class III or IV heart failure and electrocardiographic evidence of ventricular dyssynchrony (QRS duration ≥120ms) [1–3]. However, approximately one-third of patients receiving CRT fail to respond to treatment [4]. Analyses that accounted for the placebo effect and the use of remodeling rather than functional endpoints found an even higher non-response rate [5]. Given the number of device and/or procedure-related adverse events as well as the economic burden of CRT, numerous efforts have been made to identify parameters that could be used to improve patient selection and reduce the incidence of non-response. Recently, it has been shown that the presence of posterolateral scar burden can predict CRT non-response [10,11]. Birnie et al. showed that lateral wall scarring/fibrosis is important in determining CRT response in non-ischemic cardiomyopathy patients [10]. Bleeker et al. found that the patients most likely to improve after CRT were those without transmural scar tissue in the posterolateral wall and severe left ventricular mechanical dyssynchrony (LVMD) at baseline [11]. These findings were attributed to ineffective pacing due to nonviable scar tissue in the region of the LV pacing lead. This observation supports combined assessment of both LVMD and scar burden prior to CRT implantation to verify the characteristics of the area targeted for LV pacing. Given these findings, there has been increased interest in the use of nuclear imaging for predicting CRT response [12,13]. Myocardial perfusion assessment with single photon computed tomography (MP-SPECT) provides information regarding LV myocardial viability and could enable evaluation of LV scar burden. A software tool that enables the simultaneous estimation of regional LV function and wall motion synchrony in combination with MP-SPECT has been validated [14–16]. Since the phase analysis tool is largely automatic, it is less prone to inter observer errors than echocardiography. Additionally, it does not require the considerable specialized expertise required to perform these assessments by echocardiography.
Methods
Results
Discussion As with other phase analysis tools, computing quantified cardiac function parameters with cardioGRAF is largely automatic. Therefore, the considerable level of expertise required to achieve reproducible results with echocardiography is not required. In an evaluation of the cardioGRAF tool, Takahashi et al. [22] demonstrated that the inter- and intra-observer reproducibility (i.e., coefficient of variation error) for the time to contraction (i.e., time from the onset of the R-wave to end systole) was 4.8% and 3.2%, respectively. Patients with ischemic cardiomyopathy were not included in this study because of the potential difficulties in accurate depiction of the LV endocardium with gated MP-SPECT due to severe perfusion defects. The results of the present study, however, suggested that there could be severe perfusion defects even in the hearts of patients with non-ischemic cardiomyopathy. The growing numbers of patients eligible for CRT due to an aging population and expanded indications, including mild heart failure [30,31], in combination with evidence that the actual non-responder rate may be even higher than initial estimates [32], underscores the ongoing need for methods to predict the CRT response. Reducing the non-response rate requires identification of appropriate parameters, cutoff values, and assessment techniques. Starting with the early stages of CRT, the use of a wide QRS duration as a patient selection criterion has been questioned; rather, the presence of LVMD was proposed as an important predictor of a firm response to CRT [6]. Direct estimation of LVMD using an imaging modality has been proposed as a method to improve patient selection for CRT and reduce the incidence of non-responses. Multiple small single center studies have demonstrated the application of a variety of echocardiographic measures of LVMD to predict CRT response with a high degree of sensitivity and specificity [6–8]. However, these results have not been duplicated in a large-scale multicenter study. The Predictors of Response to CRT (PROSPECT) trial, a prospective, multicenter study evaluating 12 echocardiographic parameters of LVMD, found only modest sensitivity and specificity in predicting CRT response [9]. Although each of the 53 participating centers was required to obtain accreditation from an echocardiography core laboratory, Chung et al. found that differences in technology platforms, training standards, and measurement analytics limited the reproducibility of the imaging methods in a non-specialized multicenter setting [9]. In Japan, Seo et al. concluded that echocardiographic parameters did not show significant power to detect CRT responders independently [35].