The final results of the FUTURE (Functional testing underlying revascularisation) trial indicate that more patients with multivessel disease and a high SYNTAX score (>32) undergo percutaneous coronary intervention (PCI)—as opposed to coronary artery bypass grafting (CABG)—when treatment is guided by fractional flow reserve (FFR) than when guided by angiogram. This finding may be why the rate of all-cause death was significantly higher among FFR-guided patients.
Initial results from the first 836 patients enrolled in FUTURE, according to data presented at the 2016 American Heart Association (AHA) meeting, suggested that FFR-guided treatment (PCI, CABG, or optimal medical therapy) of multivessel disease is associated with a significantly higher rate of all-cause death than is angiogram-guided treatment. These results prompted the trial’s data and safety monitoring board (DSMB) to call a halt to the trial (when 938 patients had been recruited).
Therefore, with the final results—which Gilles Rioufol (Interventional Cardiology Department, Hospices Civils de Lyon, Lyon, France) presented at the 2018 European Society of Cardiology (ESC) Congress (25–29 August, Munich, Germany)—the aim was to determine if they confirmed the initial results and to review the potential reasons for the higher rate of all-cause mortality in the FFR group.
According to Rioufol, the goal of FUTURE was to explore the potential of FFR to guide treatment decisions in patients with multivessel disease. He explained that while FFR is already used to guide PCI in patients with multivessel disease, its ability to help to decide among different treatment strategies “remains unknown”. Therefore, Rioufol reported, the study hypothesis posed the question: “In mutlivessel disease patients, does FFR help to guide treatment strategy (PCI, CABG, or optimal medical therapy) and thereby improve clinical prognosis compared to traditional management?”
In the study, multivessel disease (>50% stenosis) patients with stable angina or stabilised angina were randomised to FFR-guided treatment or angiogram-guided treatment; all lesion with a FFR of >0.80 (haemodynamically not significant) were disregarded from the analysis. At one year, based on the outcomes of the 938 patients (460 in FFR group and 467 in angiogram group) who were enrolled in the study before it was halted, the rate of all-cause was two-fold higher in the FFR group compared with the angiogram group. Rioufol said: “More than 70% of the deaths, overall, were cardiovascular related. They were not procedural as they occurred later on.”
To better understand these findings, Rioufol and his co-authors performed an intention-to-treat analysis. This showed that there were no significant differences at baseline in either patient characteristics or angiographic characteristics between the two groups. In the FFR group, 43% of lesions had a FFR value of >0.80 and the mean FFR value was 0.77±0.13.
At one year, there were no significant differences in the rate of the primary endpoint—a composite of all-cause death, myocardial infarction, stroke and revascularisation—between the groups at one year or at two years. However, at one year, the rate of all-cause death was still significantly higher in the FFR group: 3.7% vs. 1.5% (p=0.036).
Rioufol et al also found that choice of treatment was different between the groups. In the FFR group, more patients received optimal medical therapy (17% vs. 9%) and fewer patients underwent PCI (71% vs. 79%) compared with the angiogram group (a significant difference; p=0.002). However, the rate of CABG was identical (12% in both groups) as was the use of ad-hoc PCI (nine out of 10 PCI patients in each group).
In a subgroup analysis, the investigators found that there was two-fold increase risk of major adverse cardiac events (MACE) among patients with stable angina in the FFR group and there was a trend towards worse outcomes for patients with a SYNTAX score of >32. Additionally, in an exploratory analysis, FFR patients who died had higher SYNTAX scores and had more three-vessel disease than angiogram patients who died. Rioufol observed: “Interestingly, in an all-cause mortality analysis, more FFR patients were treated with PCI and significantly more had a SYNTAX score of >32.”
Concluding his presentation, Rioufol said that the FUTURE trial did “not challenge” the current use of FFR for PCI guiding (this was not evaluated in the study) but showed that FFR-guided treatment does not improve outcomes in patients with multivessel disease. “Hypotheses to explain the higher rate of all-cause mortality in the FFR group include a lower than expected rate of CABG in multivessel disease patients, high rate of PCI in severe patients with SYNTAX score >32, and a high rate of ad hoc PCI,” he added.
Patrick Serruys (Imperial College London, London, UK), who was one of the chairs of the late-breaking trial session in which the FUTURE trial was presented, said that the trial was “courageous” but “raised more questions than it answered”. He added that the DSMB should not have been so quick to stop the study, commenting that the higher rate of all-cause mortality could have been a “play of chance” and that the DSMB could have just as easily stopped the trial because of a higher rate of stroke in the angiogram study.
However, Serruys did say that there were “1A” recommendations (i.e. supported by a high level of evidence) that the heart team—specifically, teams that include a surgeon—should be involved in making treatment decisions for patients with three-vessel disease. Agreeing with this view, Rioufol said that he believed that FFR provided important information about functional ischaemia but, in complex patients, “you have to pullback the wire and discuss treatment decisions with the heart team”. He speculated that, in the study, use of FFR may have driven operators to use PCI rather than CABG.