EACTS Pulls Out of Left Main Guidelines After BBC Bombshell Alleging EXCEL Trial Cover-up

The European Association for Cardio-Thoracic Surgery (EACTS) is formally withdrawing their support for the current treatment recommendations for left main coronary artery disease following an investigative report by BBC Newsnight alleging that critical data were concealed in public reporting of the EXCEL trial.

The EACTS says they signed off on the 2018 guidelines, developed in conjunction with the European Society of Cardiology (ESC), without knowledge of data obtained by the BBC showing a significantly increased risk of MI with PCI when adjudicated using the Third Universal Definition. The ESC, contacted by Newsnight, said they stand by the guidelines.

According to BBC Newsnight, trial investigators downplayed warnings from the Data and Safety Monitoring Board (DSMB) during the course of the trial about an increased risk of death among the PCI-treated patients.

“The Council of the European Association for Cardio-Thoracic Surgery has considered the analysis of the data that BBC Newsnight has shown us,” reads a statement from Domenico Pagano, MD (University Hospitals/Queen Elizabeth Hospital, Birmingham, England), secretary general of EACTS, released yesterday. “It is a matter of serious concern to us that some results in the EXCEL trial appear to have been concealed and that some patients may therefore have received the wrong clinical advice.”

This is the second time in recent months that the EACTS has attempted to draw fresh scrutiny of the EXCEL results. As previously reported by TCTMD, David Taggart, MD, PhD (University of Oxford, England), the chairman of the EXCEL surgical committee during the design and recruitment phase of the trial, previously expressed his concerns about the trial at the EACTS annual meeting, in turn drawing criticism from the EXCEL investigators.

Taggart ultimately refused to sign his name to the 2019 New England Journal of Medicine paper outlining EXCEL’s 5-year findings.

Updated 2018 Guidelines Based on EXCEL and Other Trials

For patients with intermediate and high SYNTAX scores, the 2018 ESC/EACTS guidelines give PCI a class IIa and III recommendation, respectively. CABG, as in prior guidelines, remains a class Ia indication in all patients regardless of anatomic complexity.

But partly on the basis of the Abbott Vascular-sponsored EXCEL trial, published in 2016 in the New England Journal of Medicine, the new 2018 guidance gave PCI a class Ia indication in patients with left main disease and a low SYNTAX score.

EXCEL results at 3 years showed there was no difference in the risk of death, MI, or stroke between patients with left main disease and low-to-intermediate SYNTAX scores treated with PCI and those treated with surgery. Using a single endpoint that combined a modified version of the Society for Cardiovascular Angiography and Interventions definition for periprocedural MI and the Third Universal Definition for spontaneous MI, the EXCEL investigators reported that rates of MI were not significantly different between the PCI- and CABG-treated patients (HR 0.93; 95% CI 0.67-1.28).

In the BBC report, Newsnight journalist Deborah Cohen noted that the trial, by design, had intended to report MI rates according to the Third Universal Definition for MI. Citing leaked data, Cohen reported that the rate of MI in EXCEL using this definition was 80% higher in the PCI arm than in the CABG group. Those numbers were not reported in the NEJM paper or in public presentations, despite this definition being listed as a secondary endpoint in the study protocol.

Nick Freemantle, PhD (University College London, England), who was involved in drafting the 2018 ESC/EACTS guidelines, told the BBC that EXCEL “was absolutely crucial” to their decision to award a class Ia indication to PCI for low-to-intermediate risk patients with left main disease.

“Had the results of the universal definition been available at the point we were making that decision, I know I would never have agreed that the treatments were interchangeable,” Freemantle stressed.

Lead EXCEL investigator Gregg Stone, MD (Icahn School of Medicine at Mount Sinai, New York, NY), has previously defended their endpoint selection, noting that the primary endpoint captures death, stroke, and large periprocedural MIs that are prognostically significant. Contacted by TCTMD today, Stone did not respond to specific questions in time for publication, but said that the EXCEL leadership is currently drafting a response to the BBC allegations. Also contacted by TCTMD, EACTS did not respond to requests for an interview. 

David Kandzari, MD (Piedmont Heart Institute, Atlanta, GA), the principal US interventional EXCEL investigator, expressed dismay at EACTS’ decision to pull support from the European guidelines, telling TCTMD that it undermines the integrity of the heart team and could cripple collaboration between interventionalists and surgeons.

“To me, [this controversy] has evolved in such a short period of time as to be reminiscent of the impeachment hearings,” Kandzari said, with both surgeons and interventionalists seemingly looking at the same data but shaping it to fit their arguments. In contrast, when investigators drafted the EXCEL manuscript, there was enormous consensus and compromise between all collaborators, much like a heart team working together, he said.  

‘Now That’s the Story’

Rod Stables, BMBCh (Liverpool Heart and Chest Hospital, England), who participated in the trial, said the issue is relatively straightforward: the rate of MI as adjudicated by the Third Universal Definition was listed as a secondary endpoint but has not to date been reported anywhere. “We were promised the MI rate by the Universal Definition by protocol specification, but for some reason it isn’t there,” he told TCTMD. “Now that’s the story.”

When clinical trials are designed, investigators are forced to make choices about what endpoints to select (and to define them appropriately), said Stables. The primary endpoint in EXCEL has been reported “faithfully and accurately,” he said, despite the recent accusations that this wasn’t the case. In fact, Stables said he has no criticisms of the modified definition of MI selected by trial leadership, calling EXCEL a well-conducted clinical trial.

Nonetheless, the absence of the MI data is problematic. “There are published papers which show that depending on the definition of MI you choose to use, it will impact the perceived MI rates at the time of surgery and PCI,” said Stables. “This choice has a material effect in the headline results we then discuss as the principal finding of the trial.”

For example, if you ignore periprocedural MIs in EXCEL, the surgical group does really well, he said. In the NOBLE trial also looking at left main disease, as reported by TCTMD, investigators excluded periprocedural MIs in their endpoint definitions and that trial was not nearly as favorable for PCI.

Ajay Kirtane, MD (NewYork-Presbyterian/Columbia University Irving Medical Center, New York, NY), who was not an EXCEL investigator, said the definition of MI is a complicated issue for clinical trialists, something that is very difficult to tease out.

“The choice of MI definition is more complex than it might appear,” said Kirtane. “This was probably one of the most collaborative trials because it involved joint surgical and interventional steering committees, which is somewhat unprecedented,” he said. “It’s a remarkable accomplishment. My view, not being involved in the trial, would be to allow time for the trial investigators to respond in an open and honest way so we can understand all the issues that have been raised.”

Given the complexity of the trial, and all those involved, the allegations require a structured response, he added. “If you’re running a trial, you want the everyone on the steering committee and all the investigators to feel comfortable with [the response] as well. Even with that alone, it takes time,” Kirtane told TCTMD.

The Third Universal Definition, which assesses the rise and fall of cardiac troponin levels, is extremely sensitive for detecting periprocedural MIs, said Kirtane. Additionally, it’s still uncertain as to what constitutes a clinically significant MI, particularly given evolution of biomarker assessment over time and the introduction of higher-sensitivity assays. “It’s one of the reasons doctors in regular clinical practice don’t check enzymes after cardiac surgery or PCI,” he said.

To TCTMD, Stables said in his years as a clinical trialist, he has yet to conduct a perfect study. “There’s always some level of disappointment or regret about how you designed the trial, or conducted it, or what information you collected; it’s more or less universal,” said Stables. “I wouldn’t want to criticize anybody for the way [EXCEL] was set up.”

Case record forms might have been incompetently designed to capture clinical variables, such as cardiac troponin levels, to make a diagnosis of MI using different definitions, he said. Investigators might also be negligent in collecting the detailed information needed for endpoint adjudication. Those mishaps, if this is what happened, should have been caught with early trial monitoring, Stables added.

To TCTMD, Kandzari said the trial leadership plans to further address the issue of MI as defined by the Universal Definition, but he suspects that cardiac troponin measurements were incompletely collected and this made it difficult to publish data on MI rates. “Regardless, the certainty around this definition is much less than the primary endpoint used in the trial,” said Kandzari. 

Taggart Again Expresses Concern Over Mortality

The Newsnight story does not include specific examples of the missing MI data, or show any of the leaked emails pertaining to this or to the concerns of the DSMB. According to Cohen, when the DSMB brought the issue to the EXCEL investigators, the trialists “pushed back” against these concerns and the DSMB ultimately allowed the trial to continue.

Interviewed for the BBC Newsnight report, Taggart said again that he believes the EXCEL investigators downplayed the mortality signal. He also questioned why the NEJM allowed the researchers to ignore the risk of mortality in their concluding remarks. To the BBC, Taggart said he believes the investigators downplayed the increased risk of all-cause mortality with PCI and oversold the reduced risk of the study’s primary composite endpoint of death, stroke, and MI, particularly since the benefit was largely driven by a higher risk of periprocedural MIs in the CABG arm.

As previously reported by TCTMD, there was a trend toward higher mortality in the PCI arm at 3 years (HR 1.34; 95% CI 0.94-1.91), but the difference did not reach statistical significance until 5 years (HR 1.38; 95% CI 1.03-1.85).

To TCTMD, Kandzari said that he is unaware of the DSMB raising any concerns about an excess risk of mortality in the trial. He noted that while there was an increase in the risk of all-cause mortality with PCI, the EXCEL investigators did not observe any signal of increased cardiovascular mortality. He acknowledged that it can be difficult to adjudicate causes of death in trials, though none of this information was presented in the Newsnight report.

“This is a debate that really should occur professionally in a scientific session, not necessarily in the media and not in the context of false accusations of improper conduct or unethical behavior which doesn’t exist,” he said. 

The BBC report also raises the question of whether potential conflicts of interest in EXCEL influenced how the study was reported, in that the study was funded by Abbott Vascular and that many of the trialists have received funding from stent manufacturers; the other principal investigator, A. Pieter Kappetein, MD, PhD, is now Chief Medical Officer and Vice President at Medtronic.

Note: Stone is Co-Director of Medical Research and Education at the Cardiovascular Research Foundation, the publisher of TCTMD.