Page 165 - Simplicity is Key in CRT

This thesis evaluates patient selection and management in cardiac resynchronization therapy (CRT), in order to maximize the benefit for heart failure (HF) patients treated with CRT. This chapter will discuss the findings of the studies presented in this thesis with respect to their scientific and clinical relevance, limitations and will give an outlook towards the future. The work in this thesis can be divided in parts related to selection of patients for CRT (part I) and improving management of patients treated with CRT (part II)
Part I. Patient selection
The research, presented in part I of this thesis, evaluates criteria for defining left bundle branch (LBBB) QRS morphology on the 12-lead ECG, used for patient selection in CRT. Since several years guidelines dictate that LBBB morphology is the most important electrocardiographic selection criterion for patient selection for CRT [REF]. However, the studies presented in this thesis, show that LBBB morphology is not as straight forward as it seems, and propose a new selection criterion to overcome these limitations.
Variability in classification of LBBB
Chapter 2 summarizes the available evidence for the currently recommended ECG markers for patient selection in CRT. From this summary it becomes clear that the use of QRS duration, and especially QRS (LBBB) morphology comes with uncertainties that need further evaluation. The fact that meta-analyses of the studies evaluating the value of LBBB as a predictor of outcome in CRT show no association with clinical outcome, while individual randomized studies do, gives room for thought about how different study-groups defined LBBB and whether this issue could play a significant role in clinical decision-making as well.
In chapter 3 of this thesis we show that indeed there is a profound difference in the subjective identification of LBBB between implanting cardiologists, and the interpretation of the LBBB definitions in the various clinical guidelines and landmark CRT trials. In this study we showed that rigorous application of the exact definitions, used in the guidelines and landmark papers, yielded significantly different “LBBB patient” populations, compared to those established through routine clinical observation by experienced CRT-device implanting physicians. This is reflected by the large differences in the probabilities to be classified as LBBB, (ranging from 0.20 to 0.76) and in phi-coefficients of correlation between LBBB definitions ranging from 0.10 to 0.68. In addition, inter-observer variability in clinical diagnosis compared to application of the exact LBBB definition led to re-classification in one in five patients. Assuming that the implanting cardiologists, who acted as observers in this investigation, are a good representation of clinical practice, patient selection in practice may differ markedly from that in the landmark trials and between landmark trials.
The level of evidence of the recommendations in current guidelines suggests a sound scientific basis for the use of LBBB in patients selection [1, 2]. Two large meta-analyses, including the large randomized controlled trials in CRT, however show no independent association of LBBB morphology with outcome in CRT. While one analysis clearly states the lack of an independent association of LBBB morphology to outcome in CRT as its’ conclusion [3], the other analysis included it in the final model even though it did not increase the ‘goodness of fit’ [4]. An important consideration is that these meta-analyses did not consider how the included trials classified LBBB. Both studies acknowledged the lack of detailed ECG pattern information as a limitation to their conclusions. Yet, this might be the
165