INOCA is a major cause of chest pain in patients without hemodynamically significant coronary lesions, as assessed by invasive or CT coronary angiography [5]. The associated CMD can also worsen existing hemodynamically significant epicardial coronary disease [27,28,29]. Patients with INOCA are often misdiagnosed, leading to a negative impact on their physical and mental well-being and an increase in healthcare costs [30]. Abnormalities in the microcirculation have also been implicated in the pathogenesis of several conditions, including apical ballooning (Takotsubo) syndrome [31], hypertension, diabetes, obesity, metabolic syndrome [32], and the cardiovascular manifestations of COVID-19 [33,34,35,36,37]. Therefore, it is essential to establish an appropriate diagnosis to meet the therapeutic needs of patients with INOCA [38].

The increasing recognition of INOCA as a significant cause of ischemic chest pain has led to the development of several indices derived from angiography that measure the microvascular resistance and aid with the assessment of coronary microcirculation. Angiography-based methods calculate the microvascular resistance by estimating coronary flow from angiographic frame counting and subsequently deriving distal coronary pressure using CFD or cQFR. In the case of caIMR, A-IMR and Angio-IMR, resting frame counting was utilized to determine the Tmn value corresponding to coronary flow, whereas both AccuIMR and IMRangio utilized hyperemic frame counting. These indices of microvascular resistance are virtually derived from angiography and lack direct physiological measurements, which could misrepresent the actual state of the microcirculation due to the potential sources of error summarized in Table 1.

In this review, angiography-based methods showed unacceptably high limits of agreement on Bland–Altman analysis. In contrast, combined angiography- and pressure-based methods showed more acceptable levels of agreement. Both RRRCFD and RRRp-3D utilized invasively measured intracoronary pressure gradients to determine coronary flow in their calculation of microvascular resistance. The integration of accurate pressure measurements provides a more physiological basis for the calculations and reduces the risk of bias. This patient- and vessel-specific approach may account for the superior accuracy of these combined methods in assessing the microcirculation.

A recent meta-analysis comprising seven studies found that angiography-derived IMR demonstrated good overall diagnostic accuracy in predicting abnormal invasive IMR, with a sensitivity of 82% and a specificity of 83% [39]. However, Morris et al. point out that diagnostic accuracy alone does not reflect the degree of agreement between the two methods and may be imprecise in borderline cases with values close to the cut-off [40]. Instead, a Bland–Altman plot offers a better indication of how accurately angiography-derived IMR agrees with invasive IMR.

Angiography-based methods to quantify coronary microvascular function (caIMR, A-IMR, Angio-IMR, IMRangio)

The angiography-based indices of microvascular resistance identified in this review show wide limits of agreement despite having a reasonable diagnostic performance at identifying abnormal cut-off values in reference to thermodilution-derived IMR.

Table 1 summarizes the ± 1.96 SD limits of agreement, which reflect the potential magnitude of discordance between angiography-based methods and the reference method. Such large discordance can be misleading and may directly impact decision-making in the catheterization laboratory.

The central paradox of adenosine- and pressure wire-free methods is that distal pressure is calculated using fluid dynamic equations that assume hyperemic coronary flow velocity. As summarized in Table 1, caIMR relies on diastolic flow to extrapolate hyperemic flow velocity (Vhyp). A-IMR uses resting frame counts to derive resting flow velocity, which in turn is significantly lower than hyperemic velocity. Similarly, Angio-IMR is calculated by extrapolating hyperemic flow from resting flow analysis. In these cases, flow velocity is determined without achieving maximal hyperemia. However, the patient's microvascular function can affect the assumed hyperemic velocity, leading to deviations in calculated QFR values from the patient-specific flow velocities [41, 42]. These deviations can lead to errors affecting equations determining the distal pressure and the resulting error will be multiplied in all subsequent calculations, leading to erroneously large IMR values.

Conversely, IMRangio and AccuIMR rely on hyperemic frame counts to derive flow velocity [18, 20], but this approach may also introduce bias and result in inaccurate Tmn values that deviate from patient-specific values. Although a state of hyperemia theoretically enables more precise detection of microvascular functional abnormalities, reading a hyperemic frame count is challenging due to the difficulty in discerning the contrast wavefront during a high flow rate compared to the resting angiogram. Furthermore, the variability of the detected contrast transport time may be more pronounced during hyperemia as it is heavily influenced by the timing of contrast injection in the cardiac cycle compared to the resting state [42]. This can result in discrepancies between the measured contrast velocity and the actual blood flow velocity within the vessel. In a subsequent study, a non-hyperemic version of IMRangio (NH-IMRangio) was proposed, with a cut-off value of > 30 U for detecting abnormal thermodilution-derived IMR in STEMI patients [26]. However, the diagnostic performance of NH-IMRangio was suboptimal in patients with non-ST segment elevation acute coronary syndrome and CCS, possibly due to the inability of a non-hyperemic index to reflect the minimal level of resistance attainable at maximal hyperemia when the microvascular vasodilatory capacity is preserved [26].

Angiography- and pressure-based methods to quantify coronary microvascular function (MVRCFD, RRRp-3D)

The angiography-based techniques discussed in the previous section rely solely on angiography to quantify IMR. In contrast, angiography- and pressure-based methods estimate microvascular resistance by deriving coronary flow from invasive intracoronary pressure gradients measured with standard pressure wires (Table 2).

Morris et al. proposed a computational fluid dynamics model to calculate absolute volumetric flow (QCFD) from invasive pressure data and 3-D anatomic reconstructions of coronary angiographic images [21]. This enabled the subsequent calculation of MVRQCFD from the ratio of distal pressure (Pd) and QCFD.

In this systematic review, the resistance reserve ratio (RRR) was calculated from the basal and hyperemic MVRQCFD values obtained from the Morris et al. study to facilitate a direct comparison with the limits of agreement calculated from the study by Tar et al. Both studies utilized invasively measured pressure data and compared their results to Doppler-derived RRR.

RRR is an integrated index of microvascular resistance, defined as the ratio between basal and hyperemic microvascular resistance (bMR/hMR) or the ratio of distal coronary pressure (Pd) and distal coronary flow rate (Q) during resting and hyperemic conditions [43, 44]. Alternatively, RRR can also be represented as coronary flow reserve (CFR) divided by the ratio between resting and hyperemic distal pressure (Pd). In contrast with IMR, which does not provide information on the vasodilatory capacity of the microcirculation, RRR reliably reflects the ability of the coronary microcirculation to adjust its resistance in response to adenosine and provides prognostic value in both acute myocardial infarction and nonobstructive coronary artery disease [45,46,47].

Pressure- and 3D-derived CFR (CFRp−3D) was proposed by Tar et al. to calculate CFR using invasive intracoronary pressure data and 3D anatomic reconstructions of the target vessel from angiography [22]. Measuring CFRp−3D facilitates the subsequent calculation of the RRR by incorporating distal coronary pressure through the aforementioned formula. Their combined angiography- and pressure-based approach also factored in individual variations in hydrostatic pressure, where distal pressure was corrected for hydrostatic pressure variations caused by the level difference between the tip of the catheter and the pressure wire sensor. RRRP-3D showed a good correlation with Doppler-derived RRR, and better limits of agreement with the Doppler-based method was also reported compared to all methods included in this review. This highlights the importance of correcting distal pressure for variations in hydrostatic pressure to avoid inaccuracies in calculating the driving pressure gradient.

During functional assessment of coronary arteries, hydrostatic pressure variations occur due to the height difference between the pressure sensor and the catheter tip at the vessel orifice in the supine position, where the LAD usually runs upwards while the RCA and LCX run downwards [48]. These variations can influence intracoronary pressure measurements, but their impact has largely been ignored in clinical practice up until recently. In 2019, Kawaguchi et al. examined intracoronary pressures in 23 patients and reported significant differences between FFR and resting Pd/Pa values measured in the supine and prone positions. These differences were mitigated by hydrostatic pressure correction [49]. Üveges et al. investigated the effect of hydrostatic pressure on resting Pd/Pa and FFR based on height differences calculated with 3D coronary reconstruction. In their study, 41 intermediate-severity coronary lesions with FFR values between 0.7 and 0.9 were evaluated and pressure measurements were corrected for height differences by subtracting the hydrostatic pressure gradient from the distal pressure. This correction changed the interpretation of the measurements in 12% and 27% of cases for FFR and resting Pd/Pa, respectively, highlighting the potential clinical significance of hydrostatic pressure measurement [50].

Hydrostatic pressure variations are even more pronounced when invasive pressure data is used to derive coronary flow and subsequently calculate microvascular resistance from the ratio of coronary flow and distal pressure. It is worth noting that, in the study by Morris et al., pressure-derived CFR (CFRpd) closely correlated (R2 0.92, P < 0.001) but systematically underestimated (mean delta − 0.16 ± 0.17) QCFD-derived CFR in their in vivo assessment. In turn, Doppler-derived CFR overestimated CFRpd (mean delta − 0.35 ± 0.46) and a very weak correlation was reported (R2 0.32, P = 0.1) [21]. We posit that the poor correlation between pressure- and Doppler-derived CFR could be attributed, at least partly, to their lack of correcting the distal coronary pressure (Pd) for variations in hydrostatic pressure, which could significantly impact the calculated driving pressure gradient. In light of the above, the inclusion of invasive pressure data and its correction for hydrostatic pressure in hemodynamic calculations provides a stronger physiological basis for the derived parameters and may help overcome the aforementioned challenges and assumptions of deriving physiology merely from anatomy.

Clinical implications of combined angiography- and pressure-based microvascular assessment

The 2019 European Society of Cardiology (ESC) guidelines recommend invasive guidewire-based pressure and flow measurements to diagnose a microcirculatory origin of angina in patients with persistent symptoms and either angiographically normal coronary arteries or moderate stenoses with preserved FFR or instantaneous wave-free ratio (iwFR) [9]. Additionally, pharmacological testing with intracoronary ACh injection may be performed to test endothelial function and rule out vasospastic angina or microvascular spasm [9, 51]. The CorMicA trial demonstrated that a tailored treatment strategy based on CFR, IMR, and Ach testing significantly improved angina scores and quality of life in patients with INOCA [52].

Despite these recommendations, invasive microvascular assessment is not widely used due to a lack of consensus on a uniform testing protocol and a general fear of associated complications [53]. Angiography-based methods may facilitate the routine assessment of the coronary microcirculation and help identify underlying pathomechanisms of INOCA, ultimately aiding in the selection of optimal medical therapy [32]. The diagnostic accuracy of these methods could be improved with the inclusion of invasive pressure measurements and accounting for hydrostatic pressure variations during the calculation of distal pressure.

Combined angiography- and pressure-based methods in the catheterization laboratory could provide a quick and comprehensive anatomical and functional assessment of both epicardial coronary arteries and the microcirculation. A diagnostic algorithm that incorporates combined angiographic- and pressure-based evaluation of coronary physiology in patients with clinically indicated invasive measurement of FFR is proposed in Fig. 3. Patients without significant epicardial disease (FFR > 0.80 or iwFR > 0.89) could benefit from angiography- and pressure-based evaluation of CFR, RRR, or MRR.

Fig. 3

Proposed flow chart for investigating microvascular disease using combined angiography- and pressure-based methods. In cases of epicardial stenosis ranging from 50 to 90% diameter reduction, the initial evaluation of intracoronary pressure gradients (FFR) using a standard pressure enables the differentiation of hemodynamically significant lesions that require revascularization (FFR ≤ 0.80) from those necessitating further investigation to rule out underlying microvascular disease (FFR ≥ 0.80). In patients with persistent symptoms, pressure wire-based microcirculatory resistance measurements could be considered even in the absence of angiographic stenosis. RRR values ≤ 2.5 confirm the existence of micovascular disease, while negative values prompt additional investigation through intracoronary acetylcholine injection to rule out vasospastic angina or exclude a cardiac origin of angina altogether. FFR Fractional Flow Reserve, CFR Coronary Flow Reserve; RRR Resistive Reserve Ratio

After conducting the proposed investigations, the identification of pathological values suggests the presence of microvascular disease, while patients with normal values may benefit from intracoronary Ach testing to assess micro- and macrovascular reactivity. Accordingly, patients with INOCA can be further classified into those with abnormal vasoconstriction, abnormal vasodilation, or a mixed disease type [1]. In case of negative testing, myocardial ischemia could be ruled out altogether.