Morton J. Kern, MD

This was a very well-conducted study. It showed what happens to patients with a discordance between CFR and FFR. People with abnormal values of both FFR and CFR had the highest event rates, even after PCI. People with normal CFR and FFR values had the lowest event rate, typical of people who have stable asymptomatic coronary disease. The discordant groups — abnormal FFR/normal CFR and abnormal CFR/normal FFR — had nearly the same in event rates, higher than FFR-/CFR- but lower than FFR+/CFR+.

In this analysis, the abnormal FFR/normal CFR group was not noninferior to the group with normal values of both, meaning that if your FFR was positive even if your CFR was negative, you still had a slightly worse event rate.

What does this mean to guide us in practice? While we don’t routinely measure CFR, but if we do and it’s low, the patient is at greater odds for events regardless whether the FFR is positive or negative. CFR carries its own adverse prognosis. It’s not certain why that is, but people with low CFRs often have hypertrophy, diabetes, myocardial scarring, fibrosis, advanced age and other comorbidities. In the past several decades, the Amsterdam group, which studied CFR extensively, found that even in patients without coronary disease, those who had abnormal CFR still had more adverse events.

Decision-making for stable ischemic heart disease relies on what one trusts most. Since we don’t measure CFR routinely, we have to trust FFR or a nonhyperemic pressure ratio such as instantaneous wave-free ratio, diastolic resting pressure ratio, resting full-cycle ratio, etc. All of those physiologic indices are similar for stable angina and will likely result in similar outcomes. If you throw CFR in the mix, it becomes even more complicated.

Part of our problem with the data in DEFINE-FLOW is that we do not sort out the population with coronary disease well enough. Someone with stable CAD and diabetes is not the same as someone with stable CAD and no other risk factors or someone with stable CAD and a prior CV event. To make blanket statements for the stable CAD population, we would need evidence in thousands of patients.

Nonetheless, the first step toward optimizing decision-making is to be sure we are treating a lesion that needs to be treated. We can look to the ISCHEMIA trial to help. Can we justify continued medical therapy based on the patient’s clinical course or will our patient get more mileage out of revascularization? That decision requires an individual discussion with the patient. A 50-year-old executive with an FFR of 0.82 and symptoms might do better with revascularization. But for an 85-year-old sedentary individual with an FFR of 0.82, giving them a stent may increase their chance of stent thrombosis or bleeding due to dual antiplatelet therapy, and maybe they should be treated medically.

I still trust FFR. I believe it works and the longstanding data are consistent. I also believe iFR (and other nonhyperemic pressure ratios) works in this population. If CFR is normal, I am comfortable that the patient will do well if they do not have much coronary disease. But patients with normal CFR and lots of coronary disease (ie, FFR+) may have higher event rates, and FFR will tell you that. I would trust FFR for revascularization decisions and predict outcomes based on CFR in patients without CAD. If your CFR is normal, you are in good shape. If it is abnormal, then look at the lesion-specific problem.

Unfortunately, the study did not give us a stronger answer to the question of whether flow is better than pressure, because it depends in whom you are making the measurements.