This large, multicenter registry prospectively enrolled 2,000 patients with a total of 3,378 invasive pressure index measurements to analyze the use of coronary physiology in clinical practice, associated revascularization decisions, the rate of complications and potential systematic differences between intracoronary and intravenous administration of adenosine in the context of FFR measurements. In the vast majority of cases, physiological assessment was performed to guide revascularization in angiographically intermediate stenoses of patients presenting with CCS. The preferred pressure index was FFR. It was clearly shown that intracoronary bolus versus continuous intravenous administration of adenosine did not affect the result of FFR measurements, which strengthens the support for the intracoronary route of adenosine administration as a convenient alternative to the intravenous infusion which requires a more elaborate workflow. According to expectations, physiological assessment resulted in the deferral of revascularization in approximately 4 out of 5 lesions. While serious complications of intracoronary pressure measurement were rare, the rate of coronary injury by the pressure wire was not negligible.

Current guidelines recommend the use of coronary physiology to guide the revascularization of angiographically intermediate stenoses in patients presenting with CCS [1,2,3]. Accordingly, over ¾ of patients, in whom an invasive pressure index was measured, presented with CCS in this registry, which is largely in agreement with previous registries [10,11,12,13]. In the vast majority of cases, pressure index measurements were performed in angiographically intermediate stenoses if “intermediate” is defined as the range from 40 to 90% diameter stenosis. This is in accordance with the European guidelines for myocardial revascularization [2]. It should be noted that the US guidelines for coronary artery revascularization only define the stenosis range between 40 and 69% as “angiographically intermediate” [1]. This was the case in just a bit over 60% of lesions of our registry. Similar observations have been made in the ERIS study [22].

If lesions are not physiologically relevant according to the FFR or iFR measurement, current guidelines recommend deferring revascularization in angiographically intermediate stenoses [1, 2]. This has been shown to avoid unnecessary revascularization and thereby its procedure-related complications [4, 5]. Similar to previous registries [10, 13], the use of coronary physiology resulted in the deferral of revascularization in close to 80% of lesions in this registry. In line with previous registries [11, 12], interventional cardiologists decided to proceed in accordance with pressure index measurements in about 95% of cases.

According to guideline recommendations, revascularization was deferred in 77.1% of lesions, including 58.8% of lesions with a visually estimated stenosis degree ≥ 70%. Unfortunately, this study includes no follow-up data regarding the downstream event rate relative to stenosis degree, or relative to the absence or presence of other high-risk lesion characteristics such as angiographic “haziness”. Particularly in consideration of the recently published PREVENT trial, which suggests a potential outcome benefit of revascularization in lesions with high-risk anatomic characteristics [23], such data would be interesting.

The specific rationale of interventionalists to proceed with revascularization in 5.3% of cases even though intracoronary pressure measurements were above the threshold for hemodynamic relevance was not captured as part of the study protocol and remains unknown. With a median FFR of 0.81 and a median iFR of 0.91 in these lesions, measured pressure indices were close to the recommended revascularization threshold and significantly lower than in lesions where revascularization was deferred in accordance with coronary physiology. Therefore, it can be assumed that such decisions were most likely made taking into consideration the clinical context, such as typical symptoms attributed to the coronary lesion. Potentially, the presence of perceived anatomic “high-risk” criteria may also have played a role.

With the results of the DEFINE-FLAIR and iFR-SWEDEHEART trials proving noninferiority of iFR to FFR [6, 7], the novel adenosine-independent pressure index is now equally recommended in current guidelines [1,2,3]. Nevertheless, the use of iFR made up only about 10% of pressure index measurements and the use of Pd/Pa was negligibly low. In comparison, approximately 18% of pressure index measurements were iFR measurements in the SWEDEHEART registry [24]. However, relative acceptance of iFR as compared to FFR has become slightly uncertain, given that iFR was associated with an increased risk of death at 5 years as compared to FFR in the pooled analysis of the DEFINE-FLAIR and iFR-SWEDEHEART trials [8, 9].

In 15% of cases, physiological assessment was performed in patients presenting with ACS. Among patients presenting with ACS and multivessel coronary artery disease, physiological testing with FFR may be useful to guide revascularization in non-culprit lesions [25,26,27,28]. In culprit lesions, however, FFR measurement is not recommended in the acute setting as transient microvascular dysfunction prevents true vasodilation and therefore valid FFR measurement [29]. Accordingly, among patients with myocardial infarction most pressure index measurements were performed in non-culprit lesions even though a pressure index was also occasionally measured in culprit lesions.

Physiological assessment was not exclusively performed in native coronary arteries, but also in 10 coronary artery bypass grafts in this registry. FFR-guided revascularization has been shown to be feasible and provide better clinical outcomes than an angiography-guided revascularization in patients with intermediate stenoses of coronary artery bypass grafts [30].

While coronary physiology is commonly used to guide the decision to perform revascularization, the use of coronary physiology to evaluate the result after PCI is rather uncommon in clinical practice [22]. This was also the case in our registry: about 4% of the invasive pressure index measurements were performed after PCI, even less frequent than reported in the ERIS study [22]. A postinterventional optimization strategy guided by coronary physiology has shown a tendency to improve the physiological result of interventions [31], but its impact on clinical outcomes remains to be investigated. In fact, the ongoing DEFINE GPS (ClinicalTrials.gov identifier NCT04451044) and FFR-REACT (Dutch trial register identifier NL6523) [32] trials currently study the prognostic implications of the postinterventional use of coronary physiology.

Physiological mapping of coronary arteries by the pullback of a pressure wire allows discriminating between focal stenoses and diffuse coronary artery disease and may therefore facilitate identifying the optimal PCI target [33,34,35,36]. Despite the hemodynamic interaction between serial stenoses during hyperemia [37] limiting the value of FFR for pullback pressure registration, FFR was far more frequently used than iFR in this registry.

In our registry, intracoronary adenosine was used in the majority of cases. Intracoronary adenosine has been shown to induce hyperemia for FFR measurements equally well as a continuous intravenous infusion of adenosine [38]. The use of intracoronary adenosine is also supported by a recent meta-analysis, in which intracoronary adenosine demonstrated equivalent diagnostic accuracy, but was associated with less frequent side effects compared to intravenous adenosine [18]. Likewise, the route of adenosine administration did not affect FFR results in this registry either. We furthermore demonstrated that agreement between hemodynamic assessment and the subsequent revascularization decision did not differ significantly depending on the route of adenosine administration. Hence, our data further strengthens the evidence that the route of adenosine administration does not relevantly affect revascularization decisions in clinical practice.

The use of adenosine-independent pressure indices obviates side effects caused by the administration of adenosine for hyperemia. Consequently, patients experience chest pain and dyspnea less frequently [6, 7]. Prolonged chest pain was only observed once during FFR measurement in this registry, but never with adenosine-independent pressure indices. Furthermore, adenosine sometimes causes arrhythmia, especially transient third-degree atrioventricular block [7, 13, 39, 40]. Likewise, third-degree atrioventricular block or asystole was reported in about 1% of FFR measurements, but not for adenosine-independent pressure indices in this registry.

Severe complications of invasive pressure index measurement were only rarely observed. In this registry, ventricular fibrillation occurred once during iFR measurement. However, ventricular fibrillation has been described previously in the context of the physiological assessment of coronary arteries [7, 39,40,41]. The pressure wire caused coronary artery dissection in 0.2% and coronary artery occlusion in 0.1% of our patients. Similar rates of coronary artery injury have been reported by previous registries and studies [10, 11, 13, 28, 39, 41].

Consequently, the rate of coronary artery injury is not negligible and should be borne in mind when considering invasive pressure index measurement. Coronary artery injury has even been reported to be fatal once in this registry and once in the Compare-Acute trial [28]. Clearly, a thoughtful use of coronary physiology to guide revascularization is required and interventional cardiologists should be aware of its potential complications.

This registry reflects the use of coronary physiology in clinical practice in the participating 8 interventional cardiology centers. However, it seems likely the use of invasive pressure indices varies between interventional cardiology centers depending on preferences. Substantial heterogeneity was observed even between the participating centers and the registry is therefore not necessarily representative of the clinical use of coronary physiology in other centers.

The study protocol did not prescribe a specific algorithm to address the hemodynamic evaluation of serial lesions or lesions followed by a bifurcation, including left main coronary artery stenosis. For the latter, and particularly in left main bifurcation lesions, it is assumed that the investigators followed the recommended approach to measure pressure in both branch vessels as long as neither shows an angiographic lumen reduction [42, 43].

Novel methods for the evaluation of coronary physiology appear to be underrepresented in this study. The use of adenosine-independent pressure indices was relatively low while noninvasive, angiography-based pressure indices without the use of pressure wires or adenosine were not included in the registry.

The registry-based nature of this study obviously does not allow a randomized comparison between intracoronary and intravenous adenosine, hence making it subject to bias. Furthermore, the observative and nonrandomized design of a registry makes analysis, especially for correlation, vulnerable to confounding. Additionally, small systematic differences may have evaded detection as too few patients were included. While this study was able to compare intracoronary pressure measurement results, clinical decisions and complications following intracoronary or intravenous administration of adenosine, this study did not collect data on procedural duration and therefore is not able to analyze differences between the two approaches.

Since this registry primarily aimed to analyze the use of coronary physiology in clinical practice, no follow up was performed. Therefore, long-term clinical implications of the current practice regarding invasive pressure index measurement remain uncertain.