Author + information
- Thomas H. Marwick, MBBS, PhD, MPH∗ (, )
- Paaladinesh Thavendiranathan, MD, SM,
- Tomoko Negishi, MD,
- Kazuaki Negishi, MD, PhD,
- on behalf of the SUCCOUR investigators
- ↵∗Baker Heart and Diabetes Institute, 75 Commercial Road, Melbourne, Victoria 3004, Australia
As authors of a paper about the optimal performance of myocardial strain for the assessment of cardiotoxicity (1), we were excited to learn that our work had been the subject of an editorial comment (2). Our work seeks to determine whether left ventricular strain obtained from a single view has agreement and reproducibility comparable to global longitudinal strain (GLS), the current standard of care, obtained from multiple views. Because there have recently been proposals that single-view strain may be a potential alternative for increasing the efficiency of the potentially high-volume activity of cardiotoxicity surveillance, we believed a systemic comparison with GLS would be very topical. We recognize the editorialists’ pre-eminence in governance and trial design, but as a group, the SUCCOUR (Strain sUrveillance of Chemotherapy for improving Cardiovascular Outcomes) investigators hoped to clarify some issues which might have questioned performance of the SUCCOUR trial (3). Randomized controlled trial(s) (RCT) of imaging techniques are difficult to carry out and hard to fund. Alas, contrary to expectations, there are few funds available from diagnostic imaging companies and none from pharmaceutical manufacturers (as the relevant drugs are off-patent). Moreover, the reason why imaging trials are important is that outcomes are determined by the links between appropriate selection, correct interpretation, appropriate therapeutic selection, and therapeutic response. This is exactly the reason why information from the SUCCOUR RCT will provide material that the European registry mentioned by the editorialists will not provide.
In the context of clinical trials, equipoise describes the uncertainty that there is a superior option, either with respect to harm or benefit (4). The SUCCOUR RCT is relevant because the superiority of GLS, compared with ejection fraction (EF)-guided management, has been inferred from observational studies but never shown in a randomized trial. Showing that 1 test result is more predictive of cardiotoxicity in an observational study simply does not provide the requisite evidence. We would posit that one needs much more robust proof to conclude that “there is already sufficient evidence to banish equipoise” (2).
Selection of the appropriate endpoints is a cornerstone of good trial design. The SUCCOUR trial (3) seeks to show better preservation of EF in patients in whom GLS is used for surveillance. Much as we are also strong supporters of large trials with clinical endpoints, this design is difficult to integrate with the available timeframe and the nature of cardiotoxicity. First, the current definition of cardiotoxicity is based on EF (5). Second, the development of heart failure is usually a downstream risk, perhaps years hence, based upon multiple cardiotoxic exposures.
Pre-specified “subgroup analysis” refers to evaluation of treatment or intervention for a specific endpoint in subgroup of patients defined by baseline characteristics. Our analysis of 2 approaches to strain measurement does not correspond to a subgroup analysis. Indeed, this report is unrelated to the primary question (GLS vs. EF) in this trial.
Our group, including our trial design experts, do not see how “The SUCCOUR investigators now run the risk that their study, if not the care of individual patients, is compromised” (2). The concern that the analysis of GLS versus single-plane strain will somehow allow clinicians to “figure out” who is in the EF versus the GLS arm can be assuaged by the way the study is designed. This is not a double-blind study, the clinicians already know which patients are in each arm, it is the outcome determination that is blinded (PROBE design [Prospective Randomized Open Blinded Endpoint]). Our group thus believes there is no need for recalculation of power based on our observations, nor a need for any alternative trial design.
The SUCCOUR study is being run on the goodwill and interest of colleagues at more than 25 sites across the world. We hope that it will testify to the feasibility and benefit of trials of new imaging technologies, and we are grateful for the guidance of the editorialists and welcome their offers of help.
Please note: The SUCCOUR study was supported by a research grant from General Electric Medical Systems. All authors have reported that they have no relationships with industry relevant to the contents of this paper to disclose.
- 2018 American College of Cardiology Foundation
- Thavendiranathan P.,
- Negishi T.,
- Coté M.A.,
- et al.
- Fraser A.G.,
- Zaidi A.
- Negishi T.,
- Thavendiranathan P.,
- Negishi K.,
- Marwick T.H.
- Alderson P.