← Back to January 2016

Studying the Studies or Surfing PubMed

links image

Evgenij Potapov, MD, PhD
Deutsches Herzzentrum
Berlin, Germany

"Long-term mechanical circulatory support is an established treatment of end stage heart failure." This is how many, if not the majority of papers start. And the words are correct. In the past, there have been many papers reporting that implanting MCS saves lives. This has now become an established scientific fact, but it is based on only one prospective randomized study for destination therapy. It is also clear that no surgeon would randomize dying patients on the waiting list for heart transplantation (HTx).

The focus of prospective "survival" studies has moved from dying to stable NYHA III patients. One example is a Germany based multicenter "Early LVAD" study, which should help us to understand the value of the LVAD during waiting times for HTx. This study is supported by the DZHK - German Center for Cardiovascular Research. Another, although observational and nonrandomized, is the recently published "ROADMAP" prospective multicenter study - the new edition of the REMATCH study, supported by Thoratec. Then there is the randomized study ENDURANCE, which has led to improvements in patient management and pump settings.

Following the implantation of thousands of devices another point of interest has arisen: the complications associated with LVADs. Work on this seems to come in wave sets - to name just some, right heart failure and the tricuspid valve, then von Willebrand factor (vWF), the role of pulsatility regarding the aortic valve, and GI bleeding. Most recently, pump thrombosis is en vogue. All these studies are of a retrospective nature, the analysis of single or multicenter registries or of international registries such as Intermacs or Euromacs.

Few of these studies have actually changed our practice of VAD use in reality. If a surgeon repairs the tricuspid valve for moderate tricuspid valve regurgitation during LVAD implantation, that person believes in this approach and will continue to follow it. On the other hand, I do not know of a single center that still routinely measures vWF in VAD patients. The debate about LVAD settings in terms of opening of the aortic valve is still controversial and the pulsatile mode of HeartMate III or HeartWare MVAD does not add clarity. Even the optimal anticoagulation is not defined, and all recommendations are based on retrospective analysis or are more or less arbitrary, relying on institutional experience. And the most important question of all, LVAD vs. HTx, is also still open.

Recommendations for all aspects of perioperative management could ultimately be provided on the basis of prospective randomized trials (such as the DZHK supported study mentioned above) and I am sure that all surgeons and heart failure cardiologists would be happy to participate in such trials. However, due to the large number of patients required for the sake of adequate statistical power, such studies are only possible with multicenter efforts. We, as a Society in which most end-stage heart failure specialists are involved, should initiate such studies and urge industry and national research organizations to support them. And major VAD companies should continue to support prospective randomized studies like those cited above to find the optimal pump for the individual patient and the optimal settings for each pump operation.

However, I am a realist and I know that such dreams will never come true until the number of implants reaches a critical mass - in my opinion, 5-10 times more implantations would force physicians to ask fundamental questions and to perform prospective studies. In the meantime, the analysis of VAD registries for basic questions, and of large institutional databases for more sophisticated issues, remains the only tool available to us today and in the near future. This opinion is shared with us in the October issue of the Journal of Heart and Lung Transplantation by Dr. J. Stehlik et al. The quality of registry analysis is dependent on us - how precisely and completely we feed the registries - and on the registries themselves - quality control of the submitted data is indispensable and should be evident for researchers and users of the registry based studies. However, a larger problem arises from the limitations in conception of the data collection and amenability to analysis. There is a lack of uniformity of diagnostic and therapeutic approaches between the reporting centers - this is the view taken by Dr. S. Urschel et al. in the same journal one month earlier. It made me smile to read this paper immediately after that by Stehlik et al. It is, of course, not a completely controversial opinion, but it urges us against meaningless use of registry based studies for developing clinically relevant decisions. I strongly recommend that Urschel's paper is read before registers or register based studies are designed and before decisions based on such studies are made. ■

Disclosure Statement: The author has no conflicts of interest to disclose.

Share via:

links image    links image    links image    links image