Advertisement
Advertisement
The goals of treating patients with coronary artery disease are improving survival rates and reducing the risk of myocardial infarction and relevant symptoms. Situations when revascularization using percutaneous coronary intervention (PCI) or bypass surgery is needed and the optimal means of providing revascularization for patients with multivessel coronary disease, remain somewhat elusive. Much of our evidence base dates from the 1970s to 1980s.1,2 For example, in the BARI trial, from which we thought we had learned the lesson that diabetics with multivessel disease should undergo bypass surgery in preference to PCI, the average on-treatment low-density lipoprotein cholesterol was 145.
Fortunately, several large relevant clinical trials are underway, and early results are available for some. For example, the 1-year results of the SYNTAX and CARDIA trials comparing bypass surgery to PCI with drug-eluting stents have recently been reported. Vigorous debate has erupted regarding the significance given to the endpoints of stroke and repeat PCI; their relative clinical importance seems obvious, but differences in revascularization favoring bypass surgery dominate the primary endpoint. One-year results are inadequate for drawing conclusions about which are the best treatments, however, when the desired beneficial effects should last much longer.
DISCUSSION
What appears clear, however, is that outside the setting of acute coronary syndromes3 and perhaps extensive ischemia,4 stenosis-driven revascularization with PCI as it is usually practiced does not reduce the risk of myocardial infarction or improve survival (Figure 1).5 This is due to the now well-recognized disconnect between stenosis severity and the risk of plaque rupture,6,7 because PCI treats only short segments of the coronary artery tree and because there are occasional complications associated with PCI. With available therapies, reduction of these hard endpoints seems to be the sole purview of medical therapies such as antiplatelet therapy, statins, and, in some contexts, beta-blockers and angiotensin-converting enzyme inhibitors. PCI does, of course, reduce angina. The recently reported COURAGE trial,8 reiterating the results of multiple smaller studies, but with more contemporary therapy, only underscores these facts. However, the trial understates the clinical benefit of PCI because 20% to 25% of patients were asymptomatic by their own admission and their symptoms could not have improved. Even studies suggesting a survival benefit of bypass surgery for patients with unprotected left main or three-vessel disease with diminished left ventricular function can be criticized because the medical therapies used in studies performed several decades ago hardly resemble the optimal medical therapy of today.1,2
The idea that stenting in stable patients does not reduce the risk of myocardial infarction is pivotal to the strategy of PCI. For most patients anyway, it means that the operator should target only the symptom-producing lesion or lesions and take no further risk by refraining from the full use of the oculostenotic reflex. Remarkably, only one small randomized trial utilizing PCI in the context of multivessel disease has attempted to address the issue of whether complete revascularization is required for best results in this setting.9 Although not definitive due to its 219-patient size, the trial found no differences in outcomes throughout 4 years between total and culprit vessel revascularization. This is the background from which the FAME trial investigators sought to clarify whether fractional flow reserve (FFR)-based PCI might outperform PCI as it is commonly done in patients with multivessel disease.10
FFR assessment of the severity of coronary lesions was developed over a decade ago by Nico Pijls, MD, and colleagues, and has been well validated.11 It is important to remember that angiographically determined lesion severity in the 40% to 70% range and flow limitation during hyperemic stress are only weakly correlated (Figure 2).12 With FFR, maximal hyperemia is usually obtained by dilating the distal microvasculature via administration of intravenous or intracoronary adenosine. Once this is achieved, the percentage of distal pressure retained across the stenosis in question is measured using a 0.014-inch guidewire. Values less than 75% (FFR < 0.75) have been correlated with ischemia as assessed by more traditional nuclear imaging. Values in the 75% to 80% range have been considered intermediate. The safety of deferring PCI in patients with relatively simple disease and FFR > 0.75 was previously demonstrated in the DEFER study.13
The FAME investigators randomly assigned 1,005 consenting patients with at least two major epicardial vessels requiring stenting, as assessed by routine coronary angiography, to either FFR-based treatment (intervening only upon lesions with FFR < 0.8) or conventional PCI.11 Recalling that a reduction in the risk of death or myocardial infarction with any form of PCI in this setting would not be expected, a more selective targeting of lesions that were objectively associated with ischemia might be anticipated to reduce costs and complications without losing the beneficial effect on symptoms. Perhaps not surprisingly, major adverse cardiac events at 1 year were seen less frequently with FFR-based treatment compared to standard intervention (13.2% vs 18.4%; P = .02). There were two components to this benefit—an early reduction in myocardial infarction and a later reduction in revascularization.
ASSESSMENT
Although the FAME trial appears to have been carefully performed, the manuscript by Pijls and colleagues falls short in providing certain details that might allow the reader to better put the results into context.11 First, how did the investigators choose which lesions to treat in the usual care group? On what basis did the operators decide whether to treat 50% to 70% lesions or those in smaller vessels? The methods state "the investigator must state which lesions will be stented based on visual assessment of the angiogram."10 What was the institutional perspective on total revascularization? How often was regional perfusion assessed by stress testing and utilized for guidance? Second, was contemporary adjunctive pharmacology utilized? How many patients were pretreated with thienopyridines? What antithrombin agents were used? How often and under what circumstances were glycoprotein IIb/IIIa antagonists utilized? Third, why was the periprocedural myocardial infarction rate in the usual care group so high? Only a third of their patients had acute coronary syndrome and the average SYNTAX score was only 14, yet the periprocedural myocardial infarction incidence was 7.5%. Infarction rates in COURAGE, REPLACE-2, and ISAR–REACT 1, all with largely comparable patient populations, were 3%, 6.6%, and 3.7%, respectively. Might the usual care group in FAME have been unlucky? Fourth, the centers chosen to participate in this study had a long-standing interest in FFR. Can these results be generalized? Lastly, perhaps the question itself was posed against something of a "straw man." The use of FFR certainly does not have to be an all-or-none proposition.
CONCLUSION
History has shown us that many statistically significant results from studies of this size are not repeatable. However, based on results of other relevant trials noted previously, it is likely that the investigators are onto something. A confirmatory study addressing the issues raised here would be very helpful for the interventional community. In the meanwhile, interventionists should recognize the limitations of coronary angiography and PCI, while of course not forgetting to or being intimidated into not offering their benefits.
Stephen G. Ellis, MD, is with the Cleveland Clinic Foundation; Director, Sones Cardiac Laboratories, Department of Cardiovascular Medicine; and Professor of Medicine, Cleveland Clinic Lerner College of Medicine of Case Western Reserve University in Cleveland, Ohio. He has disclosed that he holds no financial interest in any product or manufacturer mentioned herein. Dr. Ellis may be reached at (216) 445-6712; elliss@ccf.org.
Advertisement
Advertisement
