The news programme BBC Newsnight has made further claims that the EXCEL investigators “held back” data from the published results of the left main trial. In a programme that aired on 18 February, reporter Deborah Cohen claims that Gregg Stone (The Zena and Michael A. Wiener Cardiovascular Institute, Icahn School of Medicine at Mount Sinai, New York, USA) and colleagues could have published mortality data in 2016 that showed that “more people had died” with stents.
Last year, a Newsnight investigation claimed that unpublished data (based on the universal definition of myocardial infarction), from EXCEL, indicated that “80% more patients had a heart attack” in the percutaneous coronary intervention (PCI) group than in the coronary artery bypass grafting (CABG) group. Cohen acknowledged that Stone et al dispute this result (calling it “fake”) and say that the methodology used in the left main trial was the “right one”.
However, on the latest programme, she said: “But, it was not just heart attack data that were not published. People had died and the researchers held that information back for years.” She goes on to explain that while the first patients were enrolled in EXCEL in 2011, some patients “joined it as late as 2013” because “it takes time to recruit people for a medical trial”. “The researchers followed the patients from the time that they had the procedure, so they could have chosen to see how some patients were faring up to four or five years after their treatment. But, they chose to see only two or three years. So when they published the data [in 2016] on what happened to the patients, it was only on the basis of the first three years,” Cohen noted.
These mortality findings, as Cohen noted, were “eventually published”—the five-year data were presented at the 2019 Transcatheter Cardiovascular Therapeutics (TCT) meeting and published in the New England Journal of Medicine. The five-year data, like the three-year data, overall showed that PCI was non-inferior to CABG for selected patients with left main disease. However, all-cause mortality was higher with PCI (albeit not significantly).
According to Cohen, Stone et al claim that the trial was “carried out rigorously and to accepted academic standards” and they “analysed as they said they would at the start of the trial”. Stone has previously told Cardiovascular News that: “There was no significant difference in cardiovascular mortality, which is what would be expected between two cardiac procedures if one was to improve survival” and the all-cause mortality finding was “much more likely” to be a play of chance. Furthermore, in a talk at the recent Joint Interventional Meeting (JIM 2020; 13–15 February, Milan, Italy), he described mortality as an “underpowered exploratory endpoint” that was not specified for hypothesis testing and said that the observed difference of 87% survival for PCI vs. 90% for CABG, a difference of 3% over two years (0.6% per year) was “modest”.
Reaction to new claim
Nick Freemantle (Institute of Clinical Trials and Methodology, University College London, London, UK) and David Taggart—who both featured on the original Newsnight programme—made strong statements about the new claims. Freemantle said on the programme on 18 February: “So if somebody had died at three years and one day, their death would not have been counted in the [three-year] results. I am absolutely appalled that they have done this. I have taken a straw poll of my professional colleagues and it draws disbelief that someone would do this.” Taggart, who withdrew his name as an author from the five-year paper because he disagreed with its conclusion, said: “I should have been made aware of these data. Had I been made aware of these data, I would have insisted that this was discussed with the full trial steering committee because that is what we expect them [the lead investigators] to do.”
Additionally, Cohen stated that when the European Society of Cardiology (ESC) and the European Society for Cardio-Thoracic Surgery (EACTS) were drawing up their 2018 guidelines for myocardial revascularisation there was such a debate about the left main chapter (which EACTS has since withdrawn support for) that the ESC brought in an external reviewer to resolve the conflict. This reviewer apparently said: “Most patients would find these differences to be clinically meaningful, I do not believe that both these procedures should receive the same class of recommendation.”
Freemantle, who was on the ESC/EACTS guideline committee, says he did not see this review and is now calling into question the neutrality of the guidelines process. “It feels like you are stacking the decks because you have reached the conclusion you have already decided you wanted to make… discard the stats that do not fit into that narrative. It also feels a bit grubby to be honest,” he commented.
The ESC has said that the society did things “entirely properly” and it did actually make some changes after the external review; whereas EACTS has published the following statement on its website: “The allegations in BBC Newsnight’s investigation are disturbing and underline the need for the recommendations on left main disease in the 2018 myocardial revascularisation guidelines to be reviewed urgently by an independent group. In the meantime, we recommend that patients seek the advice of the multidisciplinary heart team at their hospital before deciding which treatment option is most appropriate for them. The latest revelations corroborate the EACTS Council decision of December 2019 to withdraw support from the recommendations on left main disease in the 2018 myocardial revascularisation guidelines.”
The ongoing debate about EXCEL prompted, last year, the Cardiovascular Research Foundation (CRF) to commission an independent review of the data. A CRF statement affirmed its commitment to clinical research “using high academic and operational standards”, adding: “We fully support the balanced and transparent presentation and dissemination of all clinical trial results.” However, Newsnight has questioned whether this review can truly be independent and reiterated previous concerns about the EXCEL investigators and the ESC guideline committee having conflicts of interests (i.e. financial ties with stent companies).
Fiona Godlee, the editor-in-chief of the British Medical Journal (BMJ), was featured on the programme and claimed that the New England Journal of Medicine (NEJM) has “not done a good job” by publishing the EXCEL results (the journal published both the three-year and five-year results) and should have put the trial under closer scrutiny. She added that while the BMJ was “very, very choosy” about what papers it published, it had not “solved the problem” of biased data being published—a problem she said has been going on “for years” and one that is “getting worse not better”. “Patients are right to be sceptical about clinical trials—ones that are funded by industry [ones that have] principal trial investigators that are funded by industry…. we need to have more independent research.”
In a separate development, Cardiovascular News has also learnt that the Society for Cardio-Thoracic Surgery in Great Britain and Ireland (SCTS) and the American Association of Thoracic Surgery (AATS) have written a joint letter to the NEJM, expressing concerns over the conduct of the trial and its conclusions, in particular that the inferior survival for patients treated with stenting in EXCEL did not receive sufficient emphasis. The letter is believed to have raised the issue of potential conflicts of interest generated by industry-sponsored trials and how these should be recognised and mitigated by triallists, reviewers, and editors of manuscripts submitted for publication, and clinicians engaged in writing patient management guidelines. It is understood that both societies emphasised that all data in such trials should be made available for public scrutiny and independent analysis, to avoid any reporting bias, and that when the SCTS receives a reply from the NEJM it will circulate the letter to its membership.
Cardiovascular News asked Stone for a comment on the latest Newsnight programme. However, last year, Stone and colleagues published a 3,500 statement refuting the previous programme’s claims about the trial and said that they would not respond to further requests for comment. They did not respond to a request by Newsnight to comment on the latest claims.