Author + information
- Nico H.J. Pijls, MD, PhD⁎ (, )
- Marcel Van ‘t Veer, PhD,
- Keith G. Oldroyd, MD,
- Colin Berry, MD,
- William F. Fearon, MD,
- Petr Kala, MD, PhD,
- Otakar Bocek, MD,
- Nils Witt, MD,
- Bernard De Bruyne, MD, PhD and
- Stelios Pyxaras
- ↵⁎Department of Cardiology, Catharina Hospital Eindhoven, Michelangelolaan 2, P.O. Box 1350, Eindhoven, Noord-Brabant 5602 ZA, the Netherlands
We read the paper by Sen et al. (1) with great interest. We have a number of concerns regarding the proposed index, instantaneous wave-free ratio (iFR). First, the validity of iFR depends on the assumption that minimum resting myocardial resistance during diastole is equivalent to the mean resistance during maximum hyperemia. We believe that this assumption is not correct. Numerous experimental studies performed over the last 4 decades using true volumetric flow measurement and calculating absolute resistance have provided incontrovertible proof that blood flow at rest in a normal coronary artery is very low during systole (because of the high resistance) and occurs primarily during diastole. During maximum hyperemia, flow increases during both phases of the cardiac cycle, but much more so during diastole. Because blood pressure remains either unchanged or decreases by approximately 10% to 15% (depending on the hyperemic stimulus used), both systolic and diastolic resistance will fall accordingly. Consequently, the minimal diastolic resistance at rest (regardless of whether the entire diastole or the so-called wave-free period is taken) generally is 50% to 100% higher than the average resistance over the complete heart cycle during hyperemia (2,3).
For example, in a normal left anterior descending coronary artery in a human at rest, diastolic flow is approximately 75 ml/min. During maximal hyperemia, flow during the complete heart cycle increases to 350 ml/min with still 75% to 80% of blood flow during diastole. Because there are only small changes in blood pressure or heart rate, the minimum diastolic resistance at rest, is more than twice as great as the average hyperemic resistance. Close examination of Figure 2 in the paper by Sen et al. (1) reveals that the lowest value of resistance during any moment of the heart cycle at rest is approximately 40% higher than the average resistance at hyperemia.
These differences may explain why the correlation between iFR and FFR as presented in Figure 8 of the paper by Sen et al. (1) is weak. For a given FFR value of 0.60, iFR values range between 0.4 and 0.9. This is obscured in the Bland-Altmann diagram by the extreme compression of the vertical axis. Certainly, in a mild stenosis with minimal gradient during rest or hyperemia, as a matter of fact, all indexes are equal, and the correlation between iFR and fractional flow reserve (FFR) is excellent. The same is true for severe stenosis where vasodilator reserve is exhausted and no additional changes in resistance can occur. However, in the clinically relevant range of FFR between 0.60 and 0.90, the correlation between iFR and FFR is rather poor. In addition, taking a threshold value of 0.83 is based on retrospective receiver-operating characteristic curve analysis and reflects the systematic overestimation of true FFR by iFR, rather than, as suggested, any issue related to physiological reproducibility.
Second, it is unclear why the wave intensity analysis theory is introduced to justify the use of iFR. In practice, iFR is calculated simply as the ratio of mean distal coronary pressure to mean aortic pressure during a predefined fixed part of diastole. There does not seem to be any instantaneous component to the index. We do not understand why the differential equation in the paper by Sen et al. (1) is necessary to define iFR.
Finally, we have calculated iFR in a large number of FFR tracings with appropriate resting and hyperemic pressure recordings obtained in our laboratories during the last few months (n = 555). We have found a weak correlation with a diagnostic accuracy of 69% for all data and 60% in the relevant FFR range between 0.60 and 0.90. Moreover, even larger differences were found between iFR at rest and iFR at hyperemia, despite the fact that by definition, iFR should be independent of hyperemia. This is particularly important because true resting conditions are difficult to obtain during cardiac catheterization (4,5).
Have we overlooked something or done anything wrong? More specifically, is there a correction factor used by the authors in their algorithm not reported in the article? We are in favor of introducing new mechanisms for facilitating the application of coronary physiology to guide procedures in the cardiac catheterization laboratory. However, we urge caution before applying a new index routinely in clinical practice until it can be understood adequately and validated prospectively in larger and more diverse groups of patients.
Please note: Dr. Pijls is a consultant for St. Jude Medical; and has received grant support from St. Jude Medical and Maquet. Dr. Oldroyd has received speaker and research fees from St. Jude Medical and speaker fees from Volcano. Dr. Berry has received grant/research support and consultant fees/speaker honoraria from St. Jude Medical. Dr. Fearon has received research support from St. Jude Medical. Dr. Witt has received speaker honoraria from St. Jude Medical. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- American College of Cardiology Foundation
- Gould K.L.
- Spaan J.
- De Bruyne B.,
- Bartunek J.,
- Sys S.U.,
- Pijls N.H.J.,
- Heyndrickx G.R.,
- Wijns W.
- Barbato E.,
- Aarnoudse W.,
- Aengevaeren W.R.,
- et al.