Editor's Corner: Interventional Notes from ACC.17; Big Splashes, But Few Ripples

ACC.17 had a few Late-Breaking Clinical Trials that are of particular interest. The non-interventional FOURIER trial produced the most hype. In short, the FOURIER results support the so-called “LDL hypothesis.” LDL is now the latest coronary culprit, lowering it to 20 to 30 mg/dL is possible, and by doing so cardiac events are reduced. Interestingly, reduced mortality did not emerge as significant, but long-term results were, of course, not yet in hand. The major issue for consumers is cost (purportedly $14,000 a year) and whether the reduction in events is large enough to justify the expense. These questions will be answered in coming years as more data are gathered, but for high-risk patients with high LDL levels, PCSK9 inhibition is available and works to reduce major adverse cardiac events.

I have picked three major interventional trial topics – chronic total occlusion (CTO) interventions, fractional flow reserve (FFR), and bioresorbable vascular scaffolds (BVS) – as having had the potential for changing our interventional practices. However, in fact, the trials (DECISION-CTO, DEFINE-FLAIR, iFR-SWEDEHEART, ABSORB III) all end up being cautionary.


Whether CTO interventions are “useful” has been abundantly discussed, and there has been no randomized trial to help us understand whether mortality is affected by opening a CTO. Multiple registries, reports, editorials, etc. have indicated that CTO interventions reduce angina and increase well-being. However, there has been no difference in the adjusted risks of all-cause mortality or the composite of death or myocardial infarction (MI) between patients with successful versus failed CTO-PCI.

Successful CTO-PCI seems to reduce the need for CABG (relieves symptoms). The downside is not minor. Experience is needed to properly open a CTO. How to achieve that in mid- and low-volume centers is problematic. Perforation is rare (1.5 percent), but serious: tamponade occurs in about 15 percent of perforations. Curiously, there is a legacy effect of a perforation – continued excess mortality after 30-days. CTO-PCI is labor/resource-intensive. The time for CTO-PCI is at least double that of a PCI (usually more). Average balloon and stent use is increased compared with a non-CTO case. Fluoroscopy time is increased, which produces radiation dose concerns, and contrast use increases, making CTO-interventionalists concerned about contrast nephropathy. Lastly, collateral vessel regression starts immediately post successful CTO-PCI. Only 18 percent of patients maintain recruitable collateral function adequate to prevent ischemia or an acute coronary syndrome in the event of an acute re-occlusion.

All these issues are compounded by the results of the DECISION-CTO trial from South Korea, which randomized patients to CTO-PCI plus optimal medical therapy (OMT) or to OMT alone. Major adverse events (MACE) at three years were 21 percent and 20 percent, respectively. Secondary endpoints (MACE at 5 years, on a per-protocol analysis; cardiovascular mortality, MI, repeat revascularization, and quality of life measures) were all similar between the two groups. Much more will be learned from further analyses. Are there patient subgroups that achieve better outcomes with CTO-PCI? A subgroup analysis of the EXPLORE trial showed left ventricular ejection fraction improvement if the CTO-PCI was for the left anterior descending (LAD). Patients with high ischemic burden might be better treated with CTO-PCI, but for now, it seems OMT is non-inferior to CTO-PCI.


DEFINE-FLAIR and iFR-SWEDEHEART were collaboratively designed to quickly compare a large number of patients. The aim was to determine whether functional lesion assessment by instantaneous wave-free ratio (iFR) would be non-inferior to traditional FFR. The primary outcome (incidence of all-cause death, MI, or unplanned revascularization at 12 months) showed non-inferiority as did secondary outcomes among tested subgroups. The threshold to revascularize was ≤0.89 for iFR and ≤0.80 for FFR. Adverse cardiac events were similar between study groups. However, symptoms (not surprisingly) during the procedure were fewer in the iFR group.

The investigators pointed out that as a result of these trials, the use of iFR alone (not a hybrid iFR/FFR approach) over FFR for assessment of indeterminant lesions should be considered. Will these results really change the way we think about evaluating indeterminate coronary lesions? For those of us who have access to iFR, the technique does indeed have some advantages: it is faster than FFR, does not demand intravenous infusion of adenosine, and seems pretty reliable.

However, we need to learn more about the new technique. FFR has been more rigorously tested in large trials confirming its usefulness, and iFR is the new kid on the block. The delta between 100 percent and 0.89 is far narrower than the delta for standard FFR, and might vary enough such that an ischemia-producing lesion might be missed by iFR. There are advantages, however, for iFR. Adenosine is not available in some countries, and the side effects are, for some patients, uncomfortable, though transient. Do these two issues convince me to use iFR alone? Not as long as I have FFR to back me up. The wire is already down, and if I have doubts about the iFR outcome, I can quickly switch over and ask for an adenosine infusion to confirm, say a borderline iFR outcome of 0.90, using “standard” FFR.

ABSORB III: 2-Year Results

The ABSORB III two-year results were widely discussed at ACC 17. The Absorb BVS produced an increased risk of target lesion failure (TLF) compared with the group of patients receiving the conventional Xience drug-eluting stent. TLF occurred in 11 percent and 8 percent of patients, respectively. The difference was driven by target vessel MI (7.3 vs. 4.9 percent). Interestingly, there was zero stent thrombosis between year one and two in the Xience group and 0.3 percent in the Absorb group. These data led the U. S. Food and Drug Administration to issue a letter to physicians (notably not a Black Box warning) pointing out the increased risk of the Absorb platform compared with Xience.

Is the use/usefulness of BVS platforms now gone? I doubt it. The concept of bioresorbable stents is so attractive that further data must be collected before the concept is abandoned. Bioresorbable stents are theoretically attractive for use in younger patients (there is no residual foreign body after two years) or in patients with long lesions in the anterior descending artery. In such patients, if left internal mammary artery/LAD CABG surgery is needed after a few years, the stent has disappeared, making surgery easier. However, Absorb is costlier, takes longer to implant, and requires intravascular ultrasound, with its slight risk and expense.

Analyses from the ABSORB III trial already indicate that explanations for the outcomes are available. The device is a first-generation device and its struts are simply thicker than the best drug-eluting stents in use today. That may play an important role in deployment and outcomes. Smaller vessels (<2.5 mm) simply do not fare as well with implantation of the Absorb platform. Absorb should be used only in vessels >2.5 mm in diameter. Technique of implantation also appears to influence outcomes. The concept of “PSP” (preparation, sizing and post-dilation) is critical to promote good outcomes. When Absorb was implanted in vessels >2.5 mm there was no difference in outcomes. In larger vessels, the TLF rate was 9.4 percent and 7 percent in the Absorb and Xience arms (not statistically different).

What intrigues me is that the ABSORB III trial outcomes that differentiate the Absorb from the Xience stent are the late outcomes. They come at a time when the Absorb platform is either going or gone. Why should that lead to worse late outcomes? Perhaps the Absorb platform does not simply disappear in place, but is displaced as it breaks up, leading to prothrombotic elements in or near the coronary lumen. Perhaps there are other, more reasonable explanations. We still have a lot to learn, but for now, I doubt that many will reach for an Absorb stent when the ABSORB III data show how low the event rate is with well-designed drug-eluting metal stents such as Xience.

Peter C. Block, MD, FACC is a professor of medicine and cardiology at Emory University Hospital and School of Medicine in Atlanta, GA.

Clinical Topics: Acute Coronary Syndromes, Dyslipidemia, Invasive Cardiovascular Angiography and Intervention, Noninvasive Imaging, Lipid Metabolism, Nonstatins, Interventions and ACS, Interventions and Imaging, Nuclear Imaging

Keywords: Cardiology Interventions, ACC Publications, Absorbable Implants, Acute Coronary Syndrome, Angina Pectoris, Fluoroscopy, Myocardial Infarction, Quality of Life, Registries, Stents, Stroke Volume, Cholesterol, LDL, Cholesterol

< Back to Listings