In order to bring you the best possible user experience, this site uses Javascript. If you are seeing this message, it is likely that the Javascript option in your browser is disabled. For optimal viewing of this site, please ensure that Javascript is enabled for your browser.
Did you know that your browser is out of date? To get the best experience using our website we recommend that you upgrade to a newer version. Learn more.

We use cookies to optimise the design of this website and make continuous improvement. By continuing your visit, you consent to the use of cookies. Learn more

PROSPECT: Characteristics of patients enrolled in the Predictors of Response to CRT (PROSPECT) trial - comparison of subgroups according to extent of LV reverse remodeling

Clinical Trial Update I

Heart Failure (HF)

Presenter | see Discussant report

Presentation webcast

Presentation slides

List of Authors:

Rutger J. van Bommel, Jeroen J. Bax, William T. Abraham, Eugene S. Chung, Luis A. Pires, Luigi Tavazzi, Peter J. Zimetbaum, Bart Gerritse, Nina Kristiansen, Stefano Ghio

PROSPECT (Predictors of Response to Cardiac Resynchronization Therapy [CRT]) was the first large-scale, multicenter clinical trial that evaluated the performance of several echocardiographic measures of mechanical dyssynchrony to predict response to CRT. Although various markers of dyssynchrony contributed significantly to prediction of clinical outcome and LV reverse remodeling at 6 months follow-up, the sensitivity and specificity of these markers were modest. Though there is conflicting evidence about the prognostic value of LV reverse remodeling as a surrogate for patient outcome, it is routinely used for monitoring CRT patients. A better understanding of which patient characteristics influence LV reverse remodeling is needed. Consequently, a detailed analysis of LV reverse remodeling at 6 months follow-up was performed in the patients enrolled in PROSPECT. Patients were divided according to the extent of LV reverse remodeling at 6 months follow-up and patients with super-response or negative-response were identified. Differences in clinical and echocardiographic characteristics between the groups were analyzed.

Patients were grouped according to the relative reduction in left ventricular end-systolic volume (LVESV) after 6 months of CRT. These subgroups were defined as follows:
1. super-responders: patients with a reduction in LVESV ≥30%,
2. responders: patients with a reduction in LVESV of 15% to 29%,
3. non-responders: patients with a reduction in LVESV ranging from 0% to 14% and,
4. negative-responders: patients with an increase in LVESV at 6 months follow-up.

Two hundred and eighty-six patients with complete clinical assessment and complete, paired (baseline and 6 months follow-up) LVESV measurements were analyzed. Several baseline characteristics differed significantly between the 4 subgroups and were associated with either super-response or negative-response at 6 months follow-up (Table 1). Super-response was more frequently observed in:
• Females
• Patients with non ischemic heart failure
• Patients with longer QRS duration
• Patients with more baseline mechanical dyssynchrony

Conversely, negative-response after CRT was more frequently observed in:
• Patients in NYHA class IV
• Patients with a history of ventricular tachycardia

Gender, etiology of heart failure, QRS duration, severity of heart failure, a history of VT and presence of baseline mechanical dyssynchrony influence LV reverse remodeling after CRT and are associated with either super-response or negative-response at 6 months follow-up. The current findings help to better understand which characteristics influence the degree of LV reverse remodeling after CRT.

Characteristics of heart failure patients associated with good and poor response to cardiac resynchronization therapy-Table 1

AF: atrial fibrillation, IVMD: inter ventricular mechanical delay, LVEDV: left ventricular end-diastolic volume, LVEF: left ventricular ejection fraction, LVESV: left ventricular end-systolic volume, LVFT/RR: left ventricular filling ratio, NYHA: New York Heart Association, Ts: time to peak systolic velocity, VT: ventricular tachycardia.

Provided p-values are for trend between subgroups, Cochran-Mantel-Haenszel test for categorical variables and least squares regression for continuous variables.

Discussant | see Presenter abstract

Cecilia Linde, FESC (Sweden)

Presentation webcast

Presentation slides

In this sub-study of PROSPECT, Dr van Bommel et al focus on finding a combination of clinical and Doppler-echocardiography parameters to predict a response to CRT. This is a very relevant clinical question.

The present EHRA/ESC guidelines on cardiac pacing and CRT published in 2007 and the HFA/ESC guidelines on heart failure management updated 2008 state that patients in severe heart failure (NYHA III/IV) and left ventricular ejection fraction < 35% despite optimal heart failure medication are indicated for CRT with Class I A recommendation level of evidence A provided they have electrical dyssynchrony with a QRS width of > 120 ms. With these selection criteria 60-70% of patients respond to CRT in randomised controlled studies. But importantly 20% do not improve and another 20% even worsen by CRT.

This is why the PROSPECT (PRedictors Of Response to CRT) studied the addition of mechanical dyssynchrony criteria to classical CRT selection criteria to enhance the response rate to CRT. The patients thus had to be in NYHA III/IV heart failure despite optimal medical therapy and have a LVEF < 35% sinus rhythm and QRS > 130 ms (the cutoff value used in the MIRACLE studies).

This was the first large multi-center open clinical trial on this topic and with core-lab analysis of echo-data. The study published in Circulation (Chung et al Cir 2008;117:2608) assessed various M-mode or Tissue Doppler criteria for mechanical dyssynchrony for six-month improvement. Although some of these measures showed some promise either for clinical and echo-cardiography response, no single criterion was sufficiently robust to firmly predict response to CRT above the expected response rate. The reason for this was to be found in the complex methodology with relatively low feasibility and reproducibility indicating that the methods need further refinement to be put into clinical practice. Whether a combination of these criteria would increase the response rate remains to be addressed.

In this sub-study, the authors looked at a spectrum of clinically relevant baseline and echo parameters including some mechanical dyssynchrony criteria in their ability to predict extent of reverse remodelling on the one hand and left ventricular reverse remodelling combined with clinical improvement after 6 months of treatment on the other hand. The study aimed to give some insights into prediction of a positive and negative response to CRT. Significant univariate predictors for presence and extent of reverse remodelling were non-ischemic aetiology, longer QRS at baseline or greater extent of mechanical dyssynchrony, whereas significant predictors of getting worse were NYHA class IV and a history of ventricular tachycardia. For the combination of reverse remodelling and clinical improvement, non-ischemic aetiology and extent of mechanical dyssynchrony were univariate predictors of response and history of ventricular tachycardia of getting worse during CRT.

My comments on this sub-study are the following:

1. The question asked is clinically relevant since we want more patients to respond to CRT and to avoid CRT in those who cannot be expected to respond or will deteriorate. The observed overall response rate to CRT in PROSPECT is as expected, meaning that the patients studied are representative for CRT recipients to date.

2. The endpoints. Reverse remodelling, although by definition a surrogate endpoint, has been shown to be linked to mortality and morbidity in both drug studies and studies of CRT as recently shown in REVERSE (Linde et al JACC ;52:1834). Therefore the choice of this endpoint is clinically relevant. Overall, there were 56.3% echo-responders (super-responders 37.8% and responders 18.5%). For the combination of clinical improvement using the percent improved by the clinical composite and decrease of LVESV by > 15% , 44.8% improved by both measures and 39.8 % by either of these and 15.4 % did not improve by any measure. It remains somewhat unclear why the investigators chose two ways to assess reverse remodelling, either super response (drop in LVESV of > 30%) and response (drop in LVESV of > 15%) especially as the definition of a super response to CRT remains to be established. The added value of this division for the results appears limited.

3. The observation time. Super-response is most probably linked to time. In my opinion super-response means total restitution of left ventricular function. The observation time in this sub-study is limited to six months. It has been established from CARE-HF and most recently from the REVERSE studies (Daubert et al JACC 2009; in press) that reverse remodelling starts within the first 3 months of CRT but further evolves over a period of at least 18 months. Therefore, the results of the present study are limited by the relatively short observation time.

4. The methodology for establishing mechanical dyssynchrony is difficult and is linked to both inter and intra-observer variability. Paired data for LVESV were only available for 286 patients (67%) and the prevalence and extent of mechanical dyssynchrony varied. It is noteworthy that the simplest techniques such as interventricular mechanical delay (IVMD) were more prevalent and stronger univariate predictors than the more complex Tissue-Doppler techniques. For the baseline criteria, this and randomised controlled CRT studies such as MUSTIC and CARE-HF indicate that non-ischemic aetiology is linked to more extensive reverse remodelling. Some important factors likely to influence the response to CRT such as left ventricular lead position or presence and extent of left ventricular scar tissue were not analysed.

5. The number of patients studied and the uncontrolled study design. The number of patients in this study is relatively low and the study is open, meaning that the results cannot be compared to the natural history of heart failure in these patients.

6. The statistical analysis. Last but not least, most of the baseline factors analysed are interrelated and only a multivariate analysis can establish their predictive value. This analysis remains to be done but is not likely to be conclusive with the low number of patients.


This study addresses a relevant clinical problem but does not answer the question. In my opinion, the only way to get any closer to finding predictors to CRT is to pool the data from already concluded large randomised trials and look for clinically robust parameters to predict who will benefit or not from CRT.




Characteristics of patients enrolled in the Predictors of Response to CRT (PROSPECT) trial: comparison of subgroups according to extent of LV reverse remodeling

Notes to editor

This congress report accompanies a presentation given at the ESC Congress 2009. Written by the author himself/herself, this report does not necessarily reflect the opinion of the European Society of Cardiology.

The content of this article reflects the personal opinion of the author/s and is not necessarily the official position of the European Society of Cardiology.