“History has time and again shown that unlimited power in the hands of one person or group in most cases means that others are suppressed or their powers curtailed. The separation of powers in a democracy is to prevent abuse of power and to safeguard freedom for all.” The legislative power is in charge of making laws, the executive power implements these laws and the judicial power administers justice. When the 3 powers merge into one, we have an autocratic regime.
In cardiology the executive power is wielded by cardiologists and surgeons who perform their daily practice on its “citizens” (patients). The legislative power are the societies in charge of drafting the guidelines and recommendations on how to act depending on the presentation of our citizens. Finally, the judicial power is who defines what should be taken into account when assessing the effect of the treatment of our citizens and how should we interpret these findings. When the same group of people make definitions, decide relevant outcomes in trials, write the guidelines, run the trials and practice cardiology, we are in an autocratic regime.
The Valve Academic Research Consortium (VARC) was founded in 2010 with the objective of standardizing definitions related to clinical research. In its original version, a diverse group of stakeholders were represented. Interventional cardiologists, surgeons, the Food and Drug Administration (FDA), cardiac imaging, industry, clinicians. Definitions proposed by VARC were rapidly accepted and taken into consideration in major trials of transcatheter aortic valve implantation (TAVI). If we use our previous analogy, VARC represents the judicial power of cardiology.
The third version of VARC has recently been published in the European Heart Journal. In its authorship, it is not surprising to see experienced cardiologists who represent key opinion leaders from our community. In contrast to previous versions of
VARC, in which members of the FDA were mentioned at some point in their manuscript or supplementary material, in this case the FDA is not present anywhere in the main manuscript nor supplement. All authors have either received personal fees or research grants to their institution from Industry which stands to gain from studies favorable to TAVI. This, along with the fact that 91% of the authors are related to either one of the big TAVI vs SAVR trials (PARTNER, SURTAVI and Portico), suggest that fusion of powers has been taken to the highest level. Of these, 76% belong to PARTNER trial. One of its authors, J. Popma recently joined Medtronic as Vice President and Chief Medical Officer for Coronary, Renal Denervation, and Structural Heart Businesses (it is interesti
ng to note his affiliation, as at the time of submission, was noted as Beth Israel Deaconess; this was not updated before revision). Rebecca T Hahn is Chief Scientific Officer for the Echocardiography Core Laboratory at the Cardiovascular Research Foundation. David J Cohen is Director of Clinical and Outcomes Research at Cardiovascular Research Foundation (public communication of his position was released after the final acceptance of this manuscript).
In the VARC-2 document, the funding statement mentioned that grants have been provided to the VARC Board including representatives of Cardialysis, the Cardiovascular Research Foundation and others, by Abbott Vascular, Boston Scientific, Direct Flow Medical, Edwards Lifesciences, Heart Leaflet Technologies, Medtronic Corporation, and St. Jude Medical. Let’s not forget that Cardialysis and Cardiovascular Research Foundation were the trial units involved in EXCEL trial and Cardiovascular Research Foundation in PARTNER. No information regarding funding is declared in the VARC-3 article.
Can one ever mitigate for these COI, even if declared?
No one will doubt the level of expertise of each of the co-authors involved in VARC-3. But, is it not possible to find key opinion leaders without COI to be part of this important task?
Three authors of VARC-3 are simultaneously investigators of PARTNER and members of the ACC/AHA or ESC/EACTS guidelines committee. When you have authors under the payroll of industry with key positions in the 3 powers of government, it is industry who directs the course of practice.
Hospitalization or re-hospitalization may represent the most controversial definition included in the new VARC-3. This outcome has been part of the primary composite outcome of PARTNER 3 and the main driver for non-inferiority and superiority at one year of TAVI vs SAVR. Having the principal investigators of PARTNER 3 as co-authors calls into question the motivation for including this definition into VARC-3. I acknowledge the importance of re-hospitalization to patients, but let’s remember this outcome in open label trials is subject to observer bias and should therefore not be included in a primary composite outcome.
The authors propose dividing hospitalization into cardiovascular hospitalizations related to the index procedure, cardiovascular hospitalizations not related to the index procedure and non-cardiovascular hospitalization. In cardiovascular hospitalizations not related to the index procedure, they include acute myocardial infarction (AMI). Many more patients in the SAVR arm of the PARTNER trials had concomitant coronary revascularization. This means that if a patient is discharged after TAVR and is hospitalized for an AMI, this is not considered related to the procedure. I believe this is an important omission which affects surgical and TAVI patients. When a patient has concomitant coronary artery disease, the decision to undergo SAVR or TAVI has an implicit therapeutic approach (incomplete revascularization in TAVI patients compared to SAVR). Not acknowledging that future ischemic events might be a direct result of an unsatisfactory revascularization approach or mechanical distortion of the aortic root during the index procedure is controversial. In a recent paper published in Journal of the American College of Cardiology by one of the authors of VARC-3 (JRC), “about one-third of STEMI events after TAVR occurred within the few months following the procedure, and a non-atherothrombotic mechanism (coronary embolism, late prosthesis migration) was involved in 15% of cases. Finally, as outlined by Gaudino et al. regarding all-cause mortality, using cause-specific outcomes introduces ascertainment bias and doesn’t improve the prognostic effect. A more bias-resistant approach would be to focus on all-cause rehospitalization as a secondary outcome, although this is more prevalent after SAVR, due to the more invasive nature of the intervention, but with no mid- and long-term prognostic impact.
Defining valve thrombosis
Valve thrombosis in VARC-2 has been defined as the presence of a thrombus in the prosthesis with partial obstruction to the flow. The use of this definition explains the significantly higher incidence of valve thrombosis at 2 years in PARTNER 3. Instead of adding a new definition of clinically significant valve thrombosis, VARC-3 decided to update this definition, in other words to replace it with clinically significant valve thrombosis and bury valve thrombosis as part of Stage 1: Morphological valve deterioration. The significance of valve thrombosis is downgraded without sufficient justification and without any evidence that a thrombus without hemodynamic obstruction in the prosthesis has no long-term effects. There is no solid evidence in the role of anticoagulation in presence of HALT (hypoattenuated leaflet thickening), but what would you do in presence of a thrombus in the prosthesis? What would a cardiologist do in presence of a thrombus in the left atrium? Therefore, in future trials, a patient with a thrombus in the prosthesis will not be considered as having valve thrombosis unless additional criteria are present (clinically evident or hemodynamic obstruction). Removing such a definition without any evidence at the moment that long-term outcome of these patients (with a thrombus in the prosthesis) is the same as those without a thrombus is again, controversial.
Significant bleeding as a composite endpoint of early safety has been considered differently for surgery and TAVI. In the words of the authors, “As a result of the frequency and relative benign nature of VARC-3 type 2 bleeding in the setting of surgery and its significant impact on prognosis after TAVR, early safety should include VARC-3 type 3–4 bleeding in the setting of surgery but VARC-3 type 2–4 bleeding in the setting of TAVR.” This is a completely sound argument. Now, why not hold this same argument of MI?
The definition for MI has also been updated in VARC-3. One might recall months of controversy around the EXCEL trial and its questionable periprocedural MI definition. Besides the issue around the best unit of measurement (CK or TnT), the main point of discussion was based on the use of biochemical ONLY criteria to define with the same threshold a MI after PCI or CABG. The authors in VARC-3, which some of them hold key positions at Cardiovascular Research Foundation (entity in charge of conducting PARTNER and EXCEL), decided to modify their VARC-2 definition (which was aligned to the Universal Definition of MI) and adopted the modified SCAI definition. In it, patients after TAVI and SAVR are considered under the same biochemical ONLY criteria in order to define procedural related MI. The same threshold is used for a procedure in which the heart is arrested (biochemical released is inevitable) for a procedure in which TAVR is placed on a beating heart. I agree there is a need to remove ascertainment bias and that using different cut-off values might affect it, therefore, the best way is to include additional criteria besides biochemical markers (this is not an apparent problem in case of bleeding for the authors). In the same document, the authors argue that clinically significant valve thrombosis should rule and on the other hand support an AMI definition without clinical adjunct criteria. In total disrespect to the surgical community (which goes in hand with the low % of practicing surgeons as co-authors), VARC-3 embraces a controversial definition.
Downplaying new LBBB
The need for a new permanent pacemaker has been added to the composite early safety which is something for which the authors should be congratulated. Nonetheless, although clinical evidence has been growing regarding the negative impact of new left branch bundle branch block (LBBB), the authors state that “new LBBB was not included in the safety composite, but VARC-3 recognizes that this may become an important endpoint to consider in the future”. Furthermore, one of the authors of VARC-3 (JRC), was senior author of a meta-analysis which concluded that new onset LBBB was a marker of an increased risk of cardiac death and need for permanent pacemaker at 1 year.
Mysterious removal of NYHA class
Finally, NYHA class has been removed as part of the clinical efficacy composite outcomes under the justification that is a subjective measure of functional status. On the other hand, the use health related QOL questionnaires (such as Kansas City Cardiomyopathy Questionnaire, KCCQ) is promoted. Aren’t these questionnaires subjective as well? Do these questionnaires provide an objective reality as opposed to the subjective nature of NYHA class? Let’s not forget that the PARTNER 2 trial has shown that the percentage of patients in NYHA III-IV class was higher for TAVR at 2 (p=0.04) and 5 years (p=0.08). On the contrary, no differences were seen when using the KCCQ-OS. These results make one wonder why VARC-3 decided to abandon NYHA class and embrace KCCQ.
Suggestions for democratization
Industry is a great ally of medical practice, but medical practice should not be a great ally of industry. Great trials will still be sponsored by industry but to circumvent these and future controversies:
- Industry should strive to include in their trials surgeons and cardiologists with no COI
- Guidelines committees should be made up by authors with no COI nor participation in trials by industry
- Consortiums such as VARC, need to be integrated by authors with no COI nor participation in trials by industry
History has taught us that autocracy benefits a selected group of people (the establishment or oligarchy) and in the medical arena it is prone to medical reversal. Although sooner or later evidence will show its true nature, in the meantime patients lose.
Victor Dayan is Prof Adj Cardiac Surgery at Centro Cardiovascular Universitario, Montevideo, Uruguay. No conflicts of interest to declare.