Clinical Editor's Corner: Kern

Ten Things I Hate About FFR

Morton J. Kern, MD, MSCAI, FACC, FAHA

Clinical Editor; Chief of Medicine, Long Beach VA Medical Center, Long Beach, California; Associate Chief Cardiology, University of California, Irvine Medical Center, Orange, California 

Morton J. Kern, MD, MSCAI, FACC, FAHA

Clinical Editor; Chief of Medicine, Long Beach VA Medical Center, Long Beach, California; Associate Chief Cardiology, University of California, Irvine Medical Center, Orange, California 

An upcoming meeting organizer gave me the above title as a topic to talk about. Really? C’mon, I don’t hate fractional flow reserve (FFR); I can’t hate FFR (you all know that), but there are things that some people do hate about it. At the Scripps 2019 Cardiovascular Interventions annual meeting, course co-director Matt Price shared a slide with a few things he hated about complex FFR (Figure 1). These included losing the guidewire when trying to equalize at the ostium, crossing and re-crossing tandem lesions with the pressure wire, having to remeasure FFR distal to a newly placed stent, and of course, finding drift after doing a challenging pressure wire FFR (and having to do it all over again). I know Matt does not really hate FFR either. I’m grateful to him for helping us emulate good practice in the use of physiology as a frontline tool in his lab, and educating the meeting attendees on the importance of good decisions with and without FFR/non‑hyperemic pressure ratio (NHPR). I’ve sat next to Matt during many discussions of complex percutaneous coronary intervention (PCI) cases where FFR/NHPR was a critical tool for decisions to either proceed with or defer intervention. It’s also a well-known fact that at times we all struggle with using FFR, in part because of the technology, and in part because of the hassle factor (“let’s just stent and be done — outcomes are the same usually, aren’t they?”).

Like anything in medicine, there are lingering misconceptions and bad attitudes about coronary physiology in some of our labs. Despite the fact that the field has advanced with remarkably improved technology over this decade, in-lab physiology use remains low (about 15% of cases of all PCIs), due in part to low reimbursement, poor guidewire handling, signal drift, and the need for adenosine. However, the biggest barrier to use is still the physicians’ perceived lack of need as some remain wedded to the angiogram as their gold standard. There are over a decade’s worth of major clinical trials showing the benefit of using physiology to improve outcomes and reduce major adverse cardiac events (MACE).1 The debate on the value of FFR should be over by now.  

The major reasons why we might hate FFR (Table 1) are concerns that FFR doesn’t work well or is too hard to perform routinely. In describing the 10 things I hate about FFR, I hope to make a couple points: (1) that FFR/NHPR is not hard to do; (2) that’s it is valuable to our patients, and (3) that I do not hate FFR/NHPR, and neither should you. With my tongue firmly in my cheek, here are 10 things that I hate about FFR. 

I Hate Drift

It is irritating to have to do things twice. After matching the guide catheter and guidewire pressure at the beginning of the case, finding at the end that the two pressures no longer match is frustrating, deflating, and painful. Drift can be due either to the guidewire pressure changing or to the aortic pressure changing, and often we don’t know which, but we assume it’s mostly the guidewire. We must re-normalize the guidewire pressure with the aortic pressure, and re-cross the lesion and do the FFR again. What’s the solution?

First, we should have a solid setup for aortic pressure with tubing that is tightly connected and bubble free. Next, set both the aortic (guide catheter) pressure and pressure wire to atmosphere (ie, zero) on the table. Mark the zero point on the drapes so that you can check it again later at the same spot, if needed. Lastly, consider using equipment (wires/microcatheters) with optical sensors, which are reported to have significantly less drift than some piezoresistive pressure wire sensors (see “Comparing FFR Tools”2). Unfortunately, sometimes drift is unavoidable, but it should occur much less frequently than in the earlier years of FFR.  

Any change in the pressure ratio of >0.03 from 1.00 is considered significant and strictly speaking, should lead to repeating the FFR measurement. Some might think you could just add or subtract any difference in drift to the measured FFR value, but differences in ratios cannot be combined in this fashion. You can determine whether the drift would have impacted your measurement, though, such that a negative FFR of 0.82 would still be negative (and more negative) if there was a 0.98 drift check at the end. 

It is worth recalling that the impact of drift on NHPRs is much greater than the impact it has on hyperemic gradients (ie, FFR). The effects of 2 mmHg pressure drift on Pd/Pa [the ratio of mean distal coronary pressure (Pd) to the mean pressure observed in the aorta (Pa)], instantaneous wave-free ratio (iFR), and FFR are shown in Figure 2. For a Pd/Pa and iFR of 10 and 13 mmHg, a 2 mmHg pressure drift has a significant impact, whereas for an FFR, a 2 mm pressure drift would have a minimal effect on a 30 mmHg pressure gradient (at hyperemia). Thus, pressure drift is typically more important to watch for when using NHPRs than hyperemic Pd/Pa measurements.  

I Hate Adenosine

Like the majority of interventionalists, I hate unnecessary complexity and embrace simplicity. The use of adenosine is not complex, but it requires mixing and preparing pumps for infusions.  Preparation can be time consuming and costly. In addition, adenosine (IV) produces occasional short-lived but uncomfortable side effects. Eliminating adenosine was the major rationale for developing non-hyperemic pressure ratios (NHPRs) like iFR. Nonetheless, I don’t really hate giving adenosine, but I do prefer intracoronary to intravenous adenosine for two reasons. First, it’s quick and can be repeated several times over a couple minutes for a reliability check, and second, it avoids the infusion time as well as the ‘confusion time’ to pick the ‘smart minimum’ FFR (Figure 3) (smFFR). The smFFR is the lowest value of Pd/Pa during infusion, no matter when it occurs (the machines usually pick the right spot, provided there are no artifacts recorded).3 Currently there are now 4 commercially identified NHPRs equivalent to iFR (Figure 4).4 Using NHPRs is certainly simple, but still requires performing all the steps of FFR without adenosine. While I hate waiting for things, like mixing the adenosine in the lab, I like the confidence of using the gold standard technique for invasive physiology, namely hyperemic (adenosine) Pd/Pa, when I have a doubt.  

Don’t get me wrong, I also like the ease and speed of using NHPR. NHPRs are divided into whole cycle (Pd/Pa) or diastolic sub-cycle [diastolic hyperemia-free ratio (DFR) (Boston Scientific), iFR (Philips), resting full-cycle ratio (RFR) (Abbott Vascular), and diastolic pressure ratio (dPR) (OpSens)]. Available to all labs all the time is Pd/Pa, either at rest or with submaximal hyperemia using contrast media (cFFR). The cFFR threshold is ≤0.83 (for best correlation with FFR). Pd/Pa uses a threshold of ≤0.91. All diastolic sub-cycle NHPRs have the same threshold of ≤0.89 with near identical clinical outcomes.5 We should recall that the NHPRs were initially validated against FFR, then the later indices demonstrated identity with iFR. All NHPRs have about an 80% concordance with FFR. In our lab, we follow an algorithm for lesion assessment, often without needing adenosine (Figure 5). The takeaway message is that simplicity of use makes the NHPRs attractive, as well as clinically acceptable for daily use.   

I Hate Poor Pressure Wire Handling 

This is probably the most common complaint about doing FFR. The construction of piezoresistive sensor wires requires 3 small internal electrical micro cables next to a small, central core torque wire.2 Recently developed optical-fiber nitinol pressure wires have a large core wire and a very small central optic fiber, making the handling characteristics almost identical to standard workhorse PCI wires, like a Runthrough (Terumo) or Balance Middle Weight (BMW) wire (Abbott Vascular).  

Moreover, if you really hate the pressure wire handling, you can use a micro-monorail pressure catheter. Over your favorite standard wire, slide a micro-monorail catheter (ACIST Medical) beyond the lesion to get the Pd/Pa and FFR. The newest generation of the piezoresistive wires are better than their predecessors, with improved torque response, handling, and support, and can be used in most cases without problems. 

I Hate Double-Checking My FFR/NHPR Technique

Good hemodynamics require good technique. The measurement of translesional hemodynamics is no different. There are several common sources of errors: equipment or technical factors, and procedural factors (Table 2). Equipment and technical factors include connections to the monitors, tight tubing connections with de-bubbling, and signal display monitor accuracy for the correct setup (zero, scale, timing, signal recording). Poor pressure wire signals may be due to faulty connections, due in large part to moisture or blood in the wire interface connector. For best confidence in signal accuracy, confirm the absence of signal drift during pullback after measurement.  

With regard to procedural factors, guide catheter damping, incorrect pressure sensor position in the artery, inadequate hyperemia, and changing basal flows can impact either FFR, NHPR, or both. Guide catheter pressure may be lower when measured with a wire introducer into the Tuohy Borst Y connector (Figure 6). For every drop of blood leaking out, a small amount of pressure is lost. Measure guide catheter pressure with the wire introducer removed. Additional artifacts that need to be recognized include pressure waveform damping and IV adenosine pressure variance (discussed above with the smart minimum FFR). It is important to observe the guide catheter pressure signal and recognize a damped pressure with matching of the distal wire pressure with the guide catheter pressure (Figure 3, lower right). Guide pressure damping is one of the few causes of a false-negative FFR. Pressure damping is easily remedied by removing the guide from the coronary ostium.  

I Hate Re-Crossing Stents at the End of the Procedure 

Although >20% of patients still have an ischemic FFR/iFR after PCI, many operators hate having to re-cross stents with pressure wires to obtain a final post-PCI FFR. The first-generation pressure wires were difficult to torque and negotiate challenging anatomy. The discussion about wire improvements2 above is particularly relevant to re-crossing newly implanted stents. Newer fiber optic nitinol wires perform better than older wires and can be used to perform the initial lesion assessment, and then be kept in place to be used as the workhorse wire for stents, intravascular ultrasound (IVUS), and other maneuvers. The wire can also remain in place after all interventions to measure the final FFR. Our lab and other labs have been using these pressure wires as workhorse wires to test a ‘one-wire’ approach using the same guide wire from beginning to end. A ‘one-wire’ concept obviates the need for re-crossing stents at the end of the procedure. Another approach to re-cross stents with a pressure wire is to use the workhorse wire as a buddy wire to help move the pressure wire distally for a final FFR.  

I Hate Tandem or Serial Lesion FFR 

Much has been said about how to assess tandem lesions6 and as above, no one really likes complexity. Assessment of tandem lesions requires 4 laborious steps: (1) cross both lesions for a summed FFR (A+B); (2) perform a pressure wire pull back during hyperemia to identify the largest ΔP; (3) treat the lesion with the largest ΔP, and then (4) re-measure FFR of the remaining lesion and treat if still <0.80. Using NHPR, the tandem lesion assessment is simplified: (1) cross both lesions; (2) perform (if available) angiographically co-registered pressure wire pull back without hyperemia; (3) treat each lesion according to their NHPR. However, controversy remains about outcomes based on these measurements. Whichever method one chooses, the identification of gradients in the coronary artery is superior to angiographic estimates of which lesions may or may not need treatment.   

I Hate Thinking About Microvascular Disease and FFR

Many operators have expressed concern that microvascular disease prevents maximal hyperemia and produces a false negative FFR. I don’t believe this is correct. FFR provides the accurate answers for patients with microvascular disease for any degree of hyperemia that can occur in that individual. Stenting of a negative FFR (ie, >0.80, a non-obstructive epicardial stenosis) lesion in a patient with microvascular disease will not improve flow for that individual. Flow will remain limited by the microvascular disease and the patient will have the added problem of an unneeded stent in the vessel.  

I Hate Struggling With the Discordance Between FFR and NHPR

You measure Pd/Pa at 0.94 and then perform FFR, which registers 0.80 (Figure 7). What causes such a large disparity and what measurement do you trust? The second part of this question we’ll talk about later. Several postulated mechanisms attribute the disparity to difference in flow and pressure loss across focal lesions compared with that across a diffusely diseased segment. The formula for calculating the pressure gradient, ΔP=f*Q+S*Q2, where Q is flow, has 2 terms indicating the proportion of pressure loss due to a frictional coefficient (f*Q) and the portion due to a separation coefficient (S*Q2). The pressure-flow relationship of a stenosis is a curvilinear line that may reflect mild, moderate, or severe ischemia. The steeper the curve slope, the greater the ischemia (Figure 8). Focal lesions produce a larger change pressure for a given flow, because the S*Q2 coefficient has the squared Q2 term, whereas the diffuse lesions often produce larger resting gradients because of the larger f*Q term.  

Other factors associated with the disparity between FFR/NHPR include lesion severity, location (left main/left anterior descending coronary artery), heart rate (beta blockers), and age.7 Waraisaw et al7 found that approximately 60% of patients with focal disease had FFR+/iFR-, whereas 80% of patients with FFR-/iFR+ had diffuse disease. So discordant FFR/NHPR values tells you something about the lesion and the flow, and could be considered in your decision. We often use the algorithm when confronted by disparate information at the initial assessment with NHPR or cFFR.    

I Hate Not Having Enough Data

This pet peeve is easy. I want to see more data to quench my academic thirst and continue to help the skeptics by showing them that physiology for clinical care really works. 

I Hate Waiting for Angiographic FFR

I have to be careful what I wish for with this one. Angiographic FFR is coming on strong and will have a large role in interventions even though it may only be 80% correlative with invasive wire-based FFR.9 I may be out of business regarding invasive FFR, but I don’t think so (at least in the short term).  Nonetheless, angiographic FFR will be adopted as it fulfills the innate desire of the interventionalist to make things simple and quick (hopefully).  

The Bottom Line

I hope you liked this short rant on FFR. Also, I hope it encourages you to delve into the technology continuing to improve cath lab results and our patients’ lives. Of course, I don’t really hate FFR, as coronary physiology is one of my passions (as many of our readers know from previous editor’s pages). As with every technical and scientific advance in our cath labs, don’t fight (hate) it; enjoy the challenge and embrace it. 

Disclosures: Dr. Morton Kern reports he is a consultant for Abiomed, Abbott Vascular, Philips Volcano, ACIST Medical, Opsens Inc., and Heartflow Inc. 

Dr. Kern can be contacted at mortonkern2007@gmail.com.

On Twitter: @drmortkern

References
  1. Parikh RV, Liu G, Plomondon ME, et al. Utilization and outcomes of measuring fractional flow reserve in patients with stable ischemic heart disease. J Am Coll Cardiol. 2020 Feb 4; 75(4): 409-419. doi: 10.1016/j.jacc.2019.10.060.
  2. Kern M. Comparing FFR tools: new wires and a pressure microcatheter. Cath Lab Digest. 2016 May; 24(5): 4-9. Available online at http://www.cathlabdigest.com/article/Comparing-FFR-Tools-New-Wires-Pressure-Microcatheter. Accessed February 11, 2020.
  3. Johnson NP, Johnson DT, Kirkeeide RL, et al. Repeatability of fractional flow reserve (FFR) despite variations in systemic and coronary hemodynamics. JACC Cardiovasc Interv. 2015 Jul; 8(8): 1018-1027. doi: 10.1016/j.jcin.2015.01.039.
  4. Kern MJ, Berry C, deBruyne B, et al. Conversation in cardiology — is there a need for clinical trials for the non-hyperemic pressure ratios (NHPR)? Catheter Cardiovasc Interv. 2019 Aug 1;94(2):227-232. doi: 10.1002/ccd.28336.
  5. Ahn JM. IRIS FFR: prognostic performance of five resting pressure-derived indexes of coronary physiology. Presentation: TCT 2018. TCTMD. Available online at https://www.tctmd.com/slide/iris-ffr-prognostic-performance-five-resting-pressure-derived-indexes-coronary-physiology. Accessed February 11, 2020.
  6. Pijls NHJ, De Bruyne B, Bech JW, et al. Coronary pressure measurement to assess the hemodynamic significance of serial stenoses within one coronary artery: validation in humans. Circulation. 2000 Nov 7; 102(19): 2371-2377.
  7. Warisawa T, Cook CM, Howard JP, et al. Physiological pattern of disease assessed by pressure-wire pullback has an influence on fractional flow reserve/instantaneous wave-free ratio discordance. Circ Cardiovasc Interv. 2019 May; 12(5): e007494. doi: 10.1161/CIRCINTERVENTIONS.118.007494.
  8. Fearon WF, Achenbach S, Engstrom T, et al. Accuracy of fractional flow reserve derived from coronary angiography. Circulation. 2019 Jan 22; 139(4): 477-484. doi: 10.1161/CIRCULATIONAHA.118.037350.