Fractional flow reserve (FFR) is a technique that has clearly revolutionized what we do in the cath lab. It has taken us away from a look-and-treat approach, where if you see a stenosis, you stent it, which we know from clinical studies is not a good thing to do, to an approach where a much more detailed, measured physiological assessment can be made in order to see if it is worth putting a stent into a lesion. Recent publications by Nico Pijls and Pim Tinino have shown that one of the challenges in measuring FFR is the administration of adenosine. Adenosine is administered via large central access tubes put into the femoral vein and groin. It adds time to the procedure, and of course, it also adds additional cost. The technology we have developed essentially circumnavigates the administration of adenosine through the use of established, basic physiological principles. The parts of the cardiac cycle that we call the “wave-free period,” where resistance is naturally very stable and minimal, are isolated. As shown by Pijls, De Bruyne and others before them, in a period where resistance is stable, pressure is proportional to flow. That is important, because flow is very difficult to measure in the cath lab, requiring a great deal of expertise, whereas pressure is very simple and straightforward to measure. We isolate this window in the cardiac cycle by using automated algorithms. During this window, where resistance is naturally stable and minimized, we can make an automatic measure, much in the same way as FFR, which gives a value that is very similar to FFR. It takes about 5 seconds to make the measurement, and without the requirement to give adenosine.
In a typical case, an FFR assessment is made, a stent is put in, and another assessment is done to view the result. In most labs, it will take 8-10 minutes to get the tubes into the leg, to mix up the adenosine, and get the infusion pumps ready. It is an estimated 2 minutes per run, with operators making three runs (assessment, before stenting, and after stenting), totaling about sixteen minutes. iFR takes probably less than three minutes per run, minus the adenosine requirement, which is a big time savings. In addition, it offers benefits to the patient: there is no need to have another tube placed in them and an adenosine infusion is not required. Adenosine not only gives patients a tight-chested feeling of anxiety and impending doom, but also, some patients cannot tolerate it. Asthmatic patients can get very bad bronchospasms. If a patient has heart block, the electrical wiring of the heart is abnormal, and adenosine can cause the heart block to become worse. For people with low blood pressure, adenosine can push it down further, making them unstable. FFR measured at low blood pressure values is potentially unreliable, as previously reported in the guidelines outlined by Pijls and De Bruyne.1
For the clinician, there are other advantages with iFR for the workflow in the cath lab. There is no requirement to put an additional line into the leg and femoral vein. For radial operators, putting a tube in someone’s groin, a femoral access site, is something they would prefer not to do. With iFR, everything can be done through the radial artery, without the additional problems involved with adenosine. Administrative managers are also keen on techniques such as iFR, because it speeds up the workflow, allowing more patients to come through the cath lab.
What kind of equipment is involved?
At a basic level, the technique is identical to making a measurement of FFR. A sheath is put into the leg, the pressure wire is passed distal to the stenosis, and it is normalized exactly the same way. While the same equipment is used, it is the software on the console that is different. It uses the algorithms we developed. The nice thing about the technique is that it requires equipment that interventionalists use day in and day out, so there is nothing new to get used to or new techniques to learn. It is the back end, the software on the console, which is different.
iFR is not currently available, but it is being developed and commercialized by Volcano Corporation.
At the November TCT meeting, you presented data from the ADVISE (Adenosine Vasodilator Independent Stenosis Evaluation) study. Can you tell us about it?
The ADVISE study is the validation work for our algorithms. The results of this study were announced at TCT, but they are also published in JACC.2
Demographically, ADVISE was an all-comers trial, very similar to the FAME study. Ninety-six percent had stable disease. There were a variety of arteries involved in equal proportion. Seventy percent had single-vessel disease, while the rest had double- or triple-vessel disease, a typical mix for the cath lab. Eighty percent were male, fairly representative for our population. We had a high proportion of cardiac risk factors: 34% with diabetes, 22% were smokers, and 56% had hypertension. In addition, we also had around 8% with an ejection fraction of <30%.
In total, there were 157 lesions studied in 131 patients. The study had two parts. Part one was a proof of concept study, which had 39 lesions, and it sought to identify whether we could get a measure of resistance as stable as the resistance obtained with adenosine hyperemia. We took patient measurements at rest, identified the wave-free period, and made our measurements of resistance over that period without giving any adenosine. These measurements were compared to ones from patients who had adenosine over the whole cardiac cycle. We found that the iFR-period of measured resistance is statistically identical to that measured with adenosine. Part two of the study compared our measurements over the wave-free period against the gold standard of fractional flow reserve. We did a correlation, a receiver operating characteristic (ROC) analysis, and looked at diagnostic characteristics. There was a strong correlation between the two methods, across all ranges, so it was not just at the mild or the tight end of the stenosis, but we found good correlation across all ranges. These relationships also were consistent between the left and right coronary arteries, so we found no differences in measurements between the arteries. There was a high degree of diagnostic accuracy with an area under the ROC curve of 93%. Measured diagnostic accuracy was 88%, with a positive predictive value of 91%, and a negative predictive value of around 85%. We then adjusted for this statistically, taking into account the variability of FFR, as reported by De Bruyne and colleagues. Anomalies improved, as would be expected when we take into account the variability of FFR.
Could iFR overcome the low rate of FFR use in the United States?
Many articles have discussed the limitations and complications associated with the FFR procedure. All these things have been drivers to simplify the technique. iFR probably comes at a very good time, certainly in the U.S., and in the wider world as well, because there is an ever-greater scrutiny of medical practice. Particularly at the moment, cardiologists are in the spotlight. Previously, operators were at their own leisure to make assessments of the arteries to decide who needed a stent or not. In current guidelines going forward — and this is at a state-by-state and county-by-county level — angiograms are being assessed, and some physicians are being asked to write down if an FFR was done and if it reached the critical level to warrant putting in a stent, or whether an intravascular ultrasound assessment was made. Reimbursement in the U.S. is increasingly going to be made on the basis of whether an angioplasty was done appropriately. It will certainly drive the use of iFR technology, coupled with certain aspects that make it easier to do. Hopefully, it is a tool which physicians find helpful and will increase the use of this technology in the cath lab. Ultimately, anything we can do to ensure we put stents only into the people that actually need them serves all of us, as operators or as patients. That is why we have study data from FAME and DEFER, showing that if stents are placed appropriately, patients do better than if we just implant them using fluoroscopic assessment only. We know that we make mistakes if we do that.
Can you tell us more about your involvement with iFR?
I am the inventor of the algorithm. I have been involved in wave modeling and hypertension propagation models in the coronary arteries and aorta for about 6-7 years. I do computer programming, so I developed all the algorithms, but I am also an interventional cardiologist — admittedly, a strange mix! I was the principal investigator for the ADVISE study. The trial had three sites: where I am based, out of Imperial College in Hammersmith (London), at Hospital Clínico San Carlos, under Javier Escaned, in Madrid, and in conjunction with Carlo Di Mario from The Royal Brompton Hospital in London. They were very keen participants and enrollment was very quick for the study.
In 2012, we are going on to do further validation work with larger studies. ADVISE was a pilot study and we want to collect as much data as possible to validate our algorithms in larger groups. From what we have found so far, from a mathematical and statistical perspective, it is pretty sound.
Dr. Davies can be contacted at firstname.lastname@example.org.
- Pijls NH, Kern MJ, Yock PG, De Bruyne B. Practice and potential pitfalls of coronary pressure measurement. Catheter Cardiovasc Interv 2000 Jan;49(1):1-16.
- Sen S, Escaned J, Malik IS, Mikhail GW, et al. Development and validation of a new adenosine-independent index of stenosis severity from coronary wave-intensity analysis. Results of the ADVISE (ADenosine Vasodilator Independent Stenosis Evaluation) study. J Am Coll Cardiol, doi:10.1016/j.jacc.2011.11.003 (Published online 7 December 2011).