Diastolic Pressure Ratio and Resting Full Cycle Ratio Are Similar, But Not Exactly the Same
Article ID: CJ-20-0485
Details
Article ID: CJ-20-0485
The advent and successful development of instantaneous wave free ratio (iFR) has led to the burgeoning of other resting pressure-derived physiological indices such as the resting full-cycle ratio (RFR) and the diastolic pressure ratio (dPR).1–4 Although no randomized clinical studies dedicated to RFR and dPR have been conducted to date, these indices have been used as an alternative to iFR in clinical practice because diagnostic test studies have reported that the correlation coefficients between RFR and iFR or between dPR and iFR are over 0.99 and the area under the curve (AUC) in receiver-operating characteristic (ROC) curve analysis using iFR ≤0.89 as a reference standard is also more than 0.99 with the same threshold of 0.89 for both RFR and dPR as iFR.3,4 These extremely high correlation coefficients and AUCs by ROC analysis against iFR allow RFR and dPR to be used as numerical equivalents to iFR despite their different calculation methods (Figure 1).
Differences in the calculation methods for resting indices. RFR is calculated as the point at which the ratio of Pd and Pa is lowest during the entire cardiac cycle, dPR is an averaged ratio of Pd and Pa during the entire diastolic period, and iFR is an averaged ratio of Pd and Pa during the wave-free period in the diastole. dPR, diastolic pressure ratio; iFR, instantaneous wave free ratio; Pa, aortic pressure; Pd, distal pressure; RFR, resting full-cycle ratio.
Article p ????
In this issue of the Journal, Hoshino et al5 examine the frequency of the diagnostic discordance between dPR and RFR and investigate the causes of that discordance. In terms of prevalence, they found that discordance occurred in 4.7% of lesions when cutoff values ≤0.89 were used for both dPR and RFR, although their correlation coefficients were as high as 0.997. Interestingly, all the cases of discordance showed positive RFR but negative dPR and none showed the reverse. This is a novel finding that interventional cardiologists need to bear in mind when using dPR or RFR to assess the hemodynamic significance of coronary stenoses, because this difference might influence the decision-making process for revascularization. Actually, this discordance has resulted in a significantly higher prevalence of positivity for myocardial ischemia when assessed by RFR than by dPR (54.3% vs. 49.6%, P=0.047), and could mean physiological assessments by RFR are more likely to result in revascularization than by dPR.
With regard to the causes of RFR/dPR discordance, some might argue that it is just due to pressure drifts or biological viability, because all the discordance occurs around the cutoff values. But this seems not to be the case. Sensitivity analysis in which all cases with pressure drifts were excluded demonstrated discordance in 5.1%. In addition, the fact that all cases of discordance showed positive RFR but negative dPR debunks the idea of biological variability, because if the discordance had been due to variability, discordance with negative RFR but positive dPR would have happened as well. However, the causes of the discordance remain unknown. Two multivariable logistic models were built to identify contributing factors to the discordance in this study, but there were conflicting results. In the first model, right coronary artery (RCA) location and higher FFR were predictors for RFR/dPR discordance, whereas in the second model, RCA location and higher FFR were also predictors, but not for discordance.
Although further investigations are necessary to clarify the causes of discordance, a possible mechanism perhaps lies in the different calculation algorithms for RFR and dPR (Figure 2): RFR is calculated as the lowest Pd/Pa during the entire cardiac cycle whereas dPR is calculated as the averaged Pd/Pa during the diastole, so RFR is intrinsically likely to be lower than dPR. This would explain the study finding that the mean value of RFR was smaller than that of dPR and the result that all cases of discordance had positive RFR but negative dPR.
Possible mechanism for the difference between RFR and dPR. RFR, defined as the lowest point of Pd/Pa during the entire cardiac cycle (open blue circle), might intrinsically tend to be lower than dPR of the averaged Pd/Pa in diastole (orange line). *The blue smoothed line for RFR was estimated by using LOWESS because the smoothing algorithm for RFR is not publicly available. dPR, diastolic pressure ratio; Pa, aortic pressure; Pd, distal pressure; RFR, resting full-cycle ratio.
Another interesting point in this study is that when RFR or dPR were borderline, defined as between 0.86 and 0.93, a borderline value in one of them showed a borderline value in the other as well, but borderline RFR or dPR did not necessarily show a gray-zone FFR (0.75–0.80) when hyperemia was induced. Actually, only 23.5% of the lesions with borderline RFR or dPR showed a gray-zone FFR and 76.5% showed otherwise. These findings tie in with a previous study comparing iFR and FFR, which showed lesions with negative iFR could show positive FFR and vice versa.6 This discrepancy between iFR and FFR is explained by different coronary flow responses to vasodilators such as adenosine. For example, a coronary stenosis showing negative iFR could have positive FFR when coronary flow through the stenosis considerably increases in response to adenosine. On the other hand, a coronary stenosis showing positive iFR could have negative FFR when the increase of coronary flow going through the stenosis is modest. Likewise, a coronary stenosis with borderline RFR and dPR could show either positive or negative FFR depending on the response to the vasodilator. These findings mean that the resting and hyperemic physiological indices are not the same and their results can be inconsistent on many occasions. From the clinical point of view, decision-making on revascularization is not straightforward when physiological indices have borderline values. This is even more so for RFR and dPR because clinical outcome studies are limited for both of them.7–9 In this context, FFR may serve as an adjudicator in cases of borderline RFR or dPR, as in the SYNTAX II study.10 The results of the study by Hoshino et al suggest that this hybrid strategy would work in the three-quarters of cases with borderline RFR or dPR values, but for the remaining one-quarter, we need to make the decision on a case-by-case basis with careful thought.
In summary, RFR and dPR are similar, but not exactly the same and can show different results in terms of their positivity in ≈5% of cases. When the indices differ, they lie within their borderline values and RFR is invariably positive whereas dPR is negative. The article by Hoshino et al will help to better understand the results of dPR and RFR and to make a better decision on revascularization, particularly when the values are borderline.
Y.S. has received lecture fees from Phillips Volcano, and T.K. has received lecture fees from Abbott Vascular.
T.A. has received lecture fees and research funding from Abbott Vascular and is also serving as the Associate Editor of Circulation Journal.
