is also possible to measure iFR during administration of a vasodilator such as adenosine (referred to as iFRa). This consistently produces lower iFR values, the majority even lower than FFR [61]. Notably, stability of microvascular resistance remains unchanged meaning there is minimal incremental diagnostic yield by adding adenosine. Indeed, comparing iFR, FFR, and iFRa against a combined pressure and flow based parameter such as hyperemic stenosis resistance (HSR), did not show any added value [62].
IFR in Clinical Trials
IFR has been compared to FFR in a number of studies, notably the ADVISE family of studies [63–66]. The degree of classification match is strongly dependent upon the distribution of lesions included with clinical cohorts demonstrating 80–88% match. The limit of match is driven by the capacity for FFR to match itself when repeated; this natural variability is highest close to the threshold with nearly 15% of cases showing a change in classification on repeat measurement [67]. When compared to third party markers of ischemia, including HSR, CFR, SPECT, and PET imaging, iFR and FFR are equal in their capacity to detect ischemia [68–72].
Clinical outcome studies have compared the use of iFR and FFR to determine revascularization decisions for moderate stenoses. Two large randomized controlled trials: DEFINE‐FLAIR (blinded) and the iFR‐Swedeheart (open‐label) have demonstrated equivalent event rates for both iFR and FFR at one and two years [18,73]. The two studies had design co‐alignment allowing combination of the datasets with additional power; combined analysis confirms the overall findings and provides reassurance that managing patients with either index is safe and efficient approach.
Novel Non‐Hyperemic Indices
A number of other non‐hyperemic pressure ratios have been developed and they show a very strong and close relationship to iFR [74]. The diastolic pressure ratio (dPR) equals the resting ratio of the mean diastolic pressure distal to the stenosis to the mean diastolic aortic pressure. A number of iterations exist and it is numerically equivalent to iFR [75] Diastolic hyperemia‐free ratio (DFR) calculates the ratio of Pd/Pa over five beats when Pa is less than the average of Pa across a heart‐beat (ie. When the Pa is falling). This index has limited data to date. The resting full‐cycle ratio (RFR) seeks the lowest instantaneous Pd/Pa ratio within the entire cardiac cycle. RFR has a strong relationship to iFR. In some cases, this value appears to be calculated during systole, particularly in the right coronary artery [76]. The significance of this unclear. For dPR, DFR and RFR, the same threshold for revascularization has been established (0.89). Pullback approaches are under development. To date, none of the novel indices have outcome data but have been utilized in clinical practice according to their availability and numerical equivalence to iFR, which has outcome data from multiple studies and has been accepted into Guidelines.
Resting whole Pd/Pa has also been assessed and has comparable capacity to assess ischemia compared to FFR using a threshold of 0.92 [77,78]. However, Pd/Pa has a limited dynamic range to assess stenoses, has lower reproducibility and a higher susceptibility to wire drift. The addition of contrast to generate sub‐maximal hyperemia extends the accuracy of resting Pd/Pa and has a threshold of 0.83. This index cannot, however, be used for pullback assessment and there will be crosstalk between stenoses.
NHPR use scenarios
NHPRs can simplify stenosis assessment by removing the absolute need for a vasodilator. This reduces cost, time and patient discomfort from vasodilators [73]. There is additional value in patients cannot be given a vasodilator, such as those with allergy, severe airways disease or with conditions which cannot tolerate hypotension – including aortic stenosis. IFR measurement in aortic stenosis before and after percutaneous valve replacement has shown minimal shifts with the change in after load conditions [79]. This means iFR has utility in making coronary revascularization decisions in patients with concomitant aortic stenosis and coronary artery disease. NHPR studies to date have not routinely included those with left main stem lesions, but observational data in this important cohort demonstrates similar outcomes when patients are deferred or revascularized according to iFR [80].
iFR has also been assessed in non‐culprit vessels during NSTEMI and STEMI [81–83]. Studies have suggested good correlation between acutely measured values and those performed at an interval. Resting coronary flow is somewhat elevated during STEMI meaning that acutely measured iFR values may be artificially elevated and may subsequently become negative during follow‐up [83]. The data suggests a high negative predictive value: those vessels that are negative during the acute admission, are less likely to become positive during interval. More studies are being performed.
NHPR‐Pullback
A key advantage of the NHPRs is that there is less flow interaction between stenoses. Under hyperemia, a significant distal stenosis can cause under‐estimation of a proximal lesion. Alternatively, a significant proximal stenosis can cause over‐estimation of a distal stenosis. As each stenosis alters hyperemic flow conditions it means a pressure gradient measured between two lesions does not predict the pressure gradient after a stenosis is removed. Under resting conditions, this interaction between stenoses occurs much less. Resting coronary flow velocity is preserved despite the presence of even severe stenoses up to 90%. The result is that the transtenotic pressure drop observed distal to a stenosis is resultant only due stenoses proximal to the pressure sensor. In the context of a resting pressure wire pullback, the gradient observed at a given point will be valid for all the disease proximal to that location. If a given stenosis is removed, then resting flow velocity does not change significantly meaning residual pressure gradients are preserved. Therefore it becomes possible to predict the physiology after PCI [22,43]. The expected iFR or other NHPR value is computed under the assumption that the intervention is ideal without any residual gradient in the treated segment. In that regard, it can provide a target physiological value for the real‐world intervention (Figure 7.8). It can also permit multiple different stenting strategies to evaluate their potential physiological benefit. In some cases, a limited interventional approach may be sufficient to achieve a good result, while in others more extensive stenting may be required.
Choosing between NHPR and FFR: when is one better than the other?
There has been considerable debate over the merits of resting and hyperemic indices. Given the totality of evidence, it is clear that patients managed with either approach have good outcomes with little to discern between the approaches at one year [18,84]. As FFR has been in routine use for longer, many physicians feel a greater degree of comfort in its application, albeit it at greater expense, procedure time and intra‐procedure patient discomfort. Operators should be reassured that outcomes are equivalent and systematic assessment against other ischemic parameters show there is little to discern between the indexes. Where there is doubt, a resting index can always be supplemented by hyperemia using a “Hybrid” approach.
There are physiological reasons that resting indices have a closer relationship with flow‐based parameters [85]. There appears a closer relationship between iFR and CFR – a parameter that integrates the assessment of both an epicardial stenosis and microvascular function (Figure 7.9). Truly ischemic patients are those where there is concordance between pressure gradients and reduction in CFR [85,86]. These patients benefit most from revascularization and have high event rates if deferred. In those in which there is pressure loss (FFR and NHPR), but CFR is preserved, there may be merit in deferring the lesion [86]. This is being assessed prospectively in the DEFINE‐FLOW study. In patients in which iFR is non‐ischemic and FFR is ischemic, CFR is likely to be in the normal range, as measured by flow velocity [85] or PET [87]. In studies following patients who are deferred with this physiological pattern, the event rate is the same
