Biomarkers to Predict the Response to Cardiac Resynchronization Therapy

Ward Heggermont; Angelo Auricchio; Marc Vanderheyden


Europace. 2019;21(11):1609-1620. 

In This Article

Abstract and Introduction


Cardiac resynchronization therapy (CRT) is an established non-pharmacological treatment for selected heart failure patients with wide QRS duration. However, there is a persistent number of non-responders throughout. The prediction of the CRT response is paramount to adequately select the correct patients for CRT. One of the expanding fields of research is the development of biomarkers that predict the response to CRT. A review of the available literature on biomarkers in CRT patients has been performed to formulate a critical appraisal of the available data. The main conclusion of our review is that biomarker research in this patient population is very fragmented and broad. This results in the use of non-uniform endpoints to define the CRT response, which precludes an in-depth comparison of the available data. To improve research development in this field, a uniform definition of the CRT response and relevant endpoints is necessary to better predict the CRT response.


Cardiac resynchronization therapy (CRT) is an extensively validated treatment for symptomatic heart failure (HF) patients with reduced left ventricular ejection fraction (LVEF), prolonged QRS duration, and abnormal QRS morphology. Following the most recent European Society of Cardiology (ESC) guidelines,[1] there is strong evidence for CRT implantation in HF patients [New York Heart Association (NYHA Class II–IV)] with a LVEF below 35%, and with a QRS duration of more than 150 ms with left bundle branch (LBB) block morphology. For other subgroups of patients, CRT might still be useful although evidence is less compelling.[1] However, despite a moderate improvement over the last two decades, the response to CRT is still rather mixed. The recent Clinical Trial of the SonRtip lead and automatic AV-VV optimization (RESPOND-CRT) trial[2] showed a clinical composite response rate of 77% in CRT patients with LBB block and only 66% in CRT patients without LBB block. Furthermore, the proportion of so-called 'super-responders' to CRT defined by almost complete normalization of ventricular function and volumes) has remained constant over time, still representing not more than 30% of all CRT patients. The timely prediction of a CRT response, especially a super-response, is therefore of paramount importance. Although numerous other factors do determine the success of CRT (e.g. optimization of the device, lead positioning, etc.), early identification, or even prediction, of CRT response might have long-term beneficial therapeutic consequences.

Numerous attempts have already been made to find a reliable biomarker(s) to predict the response to CRT. A good, clinically useful biomarker has to be specific for the condition that needs to be detected, yet at the same time display adequate sensitivity for the pathologic condition. Ideally, it has strong predictive value, is robust and reproducible over time. Preferentially, the biomarker can be assessed non-invasively, is readily accessible (e.g. a simple blood test) and its use should be supported by both pre-clinical and clinical data.

In this review article, we aimed to present a summary of the studies on biomarkers that have been performed so far, highlighting the fact that more focused research is mandatory to find a reliable biomarker (panel) to predict the CRT response.