Time to rethinQ PROSPECT?

Cardiac resynchronization therapy (CRT) is the most successful therapy in treating chronic heart failure patients since the introduction of b-blockers, angiotensin-converting enzyme (ACE) inhibitors, and spironolactone into the therapeutic regimen. Recent studies (REVERSE, MADIT-CRT) will probably expand the indication for CRT to mild and asymptomatic heart failure patients [New York Heart Association (NYHA) functional class I and II], an enormous task for physicians and healthcare providers. CRT with current indications appears to improve the clinical status in 70% of patients. This compares well with the effects of treatments of chronic and life-threatening diseases in other medical fields. However, indications for CRT are currently based to a large degree on QRS width in an attempt to detect mechanical dyssynchrony under the assumption that correction of dyssynchrony is the main mechanism of action. The idea behind the original Predictors of Response to CRT (PROSPECT) study was to identify in a non-randomized observational design the predictive value of predefined echocardiographic parameters of mechanical dyssynchrony with regard to clinical and left ventricular mechanical outcome after CRT. Twelve previously published parameters of dyssynchrony, both conventional and tissue Doppler-based methods, had finally been used in 426 CRT patients at 53 centres. Considering the intended purpose of the PROSPECT study, the overall results were not encouraging, with no single echocardiographic dyssynchrony parameter being highly predictive of the CRT response at follow-up. There seemed to be no ‘holy grail’ of echocardiographic dyssynchrony parameter which by itself could be used for decision making regarding CRT. Is this the end of the ‘mechanical dyssynchrony’ hypothesis? Is echocardiography useless in predicting CRT respsonse despite numerous single-centre studies showing high sensitivity and specificity? Immediately after publication of the PROSPECT results, numerous concerns regarding the design of the study and the way in which it was conducted were raised, especially regarding patient selection as well as echocardiographic assessment. Concerning patient selection, 20.2% had a left ventricular ejection fraction (LVEF) .35%, and 37.8% had a left ventricular end-diastolic diameter (LVEDD) ,65 mm, which ultimately raised the question of how reverse remodelling could be achieved in a non-dilated heart. This, together with a substantial interobserver variability and large percentage of non-assessable echocardiographic studies, indicated a lack of standardized data acquisition and analysis. In addition, tissue Doppler imaging (TDI) data were obtained using echo machines from thee different ultrasound providers without standardization of frame rates, and three different software programs were used for offline data analysis. Furthermore, there was no information on left ventricular lead placement and percentage biventricular pacing. Taken together, important technical limitations as well as suboptimal patient selection may importantly confound and limit the interpretation of the initial PROSPECT study results. Van Bommel et al. have performed a subgroup analysis of 286 of the total of 498 patients which had initially been enrolled into PROSPECT; 426 of those had been originally analysed in the main PROSPECT study after 31 early exits and the exclusion of 41 patients with narrow QRS (,130 ms). Eventually, this substudy analysed 57% of the total PROSPECT population, further excluding 15 patients who died, 18 patients who had been lost for the 6-month follow-up visit, and 44 patients with incomplete data from the 6-month follow-up visit, thus ending up with 286 patients with a complete data set and paired left ventricular end-systolic volume (LVESV) measurements. Overall, this represents quite a substantial reduction compared with the 498 originally enrolled patients. In order to dissect the response to CRT, the authors grouped patients according to both a combined clinical and a pure echocardiographic response. Echocardiographically, subgroups were formed consisting of so-called ‘super-responders’ with a reduction of .30% in LVESV; ‘responders’ with a 15–29% reduction in LVSEV; ‘non-responders’ without evidence of reverse remodelling and reduction of LVESV between 0 and 14%; and ‘negative responders’ with echocardiographic signs of disease progression despite CRT with an increase in LVESV after 6 months. Clinically, they