Many patients following an acute myocardial infarction (MI) have adverse clinical outcomes despite a successful percutaneous coronary intervention (PCI) (1). Solely revascularizing an obstructed epicardial coronary artery may not be enough to improve microcirculation in acute coronary syndromes (ACS), and persistent microvascular dysfunction leads to worse clinical outcomes (2).

Several vascular resistance indices have been proposed to evaluate microcirculation function invasively, that have the advantage of being available in the catheterization laboratory to delineate the specific contribution of microcirculation to myocardial damage. The index of microvascular resistance (IMR) is a thermodilution technique that allows the quantitative assessment of the microcirculatory resistance in a coronary artery territory. The hyperemic microvascular resistance index (HMR) represents the ratio of mean distal coronary pressure (Pd) and doppler flow peak velocity during hyperemia. Lastly, zero-flow pressure (Pzf) that is extrapolated from pressure-velocity plots, represents Pd at which the coronary blood flow would cease. As opposed to coronary flow reserve (CFR), vascular resistance indices are specific for the microcirculation and independent of hemodynamic variations, making IMR more reproducible than CFR (3). The key characteristics of IMR, HMR, and Pzf are summarized in the Figure 1.

Microvascular obstruction (MVO) is an entity that represents the inability to reperfuse the coronary microcirculation within the infarct zone. It has been associated with major adverse clinical outcomes and it is considered a stronger prognostic marker than infarct size (IS) (3, 4). MVO can be assessed through different procedures at different timings after MI, however, contrast-enhanced cardiac magnetic resonance (CMR) is the gold standard technique (5).

The main purpose of this study is to provide a critical appraisal of the emerging role of microvascular resistance indices in the MI setting, its relation to MVO and IS on CMR, and its relation to clinical outcomes in the infarct related arteries (IRA) and non-IRA.

Several and interconnected pathomechanisms had been proposed to explain myocardial ischemia-reperfusion injury to the microcirculation (Figure 2). There is increasing evidence that ischemia affects not only cardiomyocytes, but also endothelial cells, despite its greater resistance to hypoxemia. Endothelial dysfunction is characterized by (i) endothelial protrusions and blebs, which lead to capillaries obstruction; (ii) endothelial gaps, allowing red blood cells leakage into the myocardial interstitium, causing intramyocardial hemorrhage; and extravasation of fluids, leading to interstitial edema and greater compression of the vessel’s lumen. The capillary dysfunction is further amplified by a proinflammatory and proaggregant environment caused by increased concentrations of vasoconstrictors substances (i.e., thromboxane A2, endothelin), and by the expression of adhesion molecules in vessels and circulating cells promoting the formation of neutrophil-platelet and erythrocytes aggregates causing obstruction of the capillaries. Furthermore, distal embolization originated from micro emboli detached of coronary plaques and/or released during PCI can worsen microcirculatory ischemic injury. Individual’s susceptibility determined by some genetic polymorphisms, pre-existing cardiovascular comorbidities and advanced age augments the risk of coronary microvascular dysfunction associated with MI (5–7).

Cardioprotection refers to interventions that reduce the injury from myocardial ischemia and reperfusion. Mechanical ischemic conditioning approaches, involving brief cycles of ischemia-reperfusion (e.g., arm cuff inflation-deflation cycles) reduce IS and coronary microvascular damage, through a complex signal transduction pathway that target intracellular organelles such as mitochondria and sarcoplasm reticulum (8). Pharmacological cardioprotection largely aimed at either inhibition of injurious pathways or activation of protective processes, such as increased formation of adenosine or nitric oxide. However, most published data did not show a significant impact of adenosine or nitrite therapy on MVO or in clinical outcomes (8, 9). Similarly, several other therapies used early in the acute MI setting (e.g., P2Y12 inhibitors, glycoprotein IIb/IIIa inhibitors, fibrinolytics) that induced cardioprotection in preclinical models, often failed to confirm its efficacy in clinical trials. A promising approach would be the additive cardioprotection, through proven mechanical (combined ischemic pre and postconditioning) and pharmacological strategies, in addition to reperfusion, and focusing on MI patients that would benefit the most from cardioprotection measures, such as those in Killip class III or cardiogenic shock (8).

We provided an overview of currently used resistance invasive techniques to assess coronary microvascular function following an acute MI and examined its relationship with MVO/IS by CMR and its relation to clinical outcomes regarding the culprit and non-culprit IRA. We found substantial differences between reported invasive measurements of microvascular dysfunction and MVO by CMR, which are summarized in Figure 5. These differences reflect the heterogeneity found in the study’s protocols (e.g., type of invasive technique, timing between invasive/non-invasive measurements), which limit the performance of formal meta-analysis or give clinical meaningful cut-off values for the doppler resistance indices (HMR and Pzf). Although, there is no true reference standard for invasive measurement of microvascular function, the vast majority of data derives from IMR and, importantly, HMR and Pzf lack a well-validated normal range.

Previously, several diagnostic tools have been proposed to identify microvascular injury in the catheterization laboratory. Angiographic parameters like the amount of time required contrast dye to reach a distal segment of a coronary artery (corrected TIMI frame count), clearing of myocardial blush after contrast injection (TIMI myocardial perfusion grade, TMPG), or according to myocardial blush intensity (myocardial blush grade, MBG). However, these angiographic methods are not quantitative and its visual assessment has limited reproducibility and does not make possible a real-time evaluation of microvascular function (post-processing). Studies have found that TMPG is the most valuable of these angiographic techniques in assessing MVO and IS (30, 31). Regarding MBG it is correlated with myocardial reperfusion after PCI, and some studies proved that it can also be predictive of MVO (12). However, it is a semiquantitative measure, with a considerable intra and interobserver variability, and presents a weak to moderate correlation strength to IMR (r = –0.42; p = 0.007) (32). Furthermore, Carrick et al. (33) reported that MBG has inferior clinical value for risk stratification than IMR, and data comparing it to HMR and Pzf is lacking.

As a continuous variable, IMR correlation with MVO and IS largely varied, ranging from no (2) to weak (1, 14, 17), moderate (15) or strong strength of relationship (11). These inconsistent results between IMR and the presence of MVO may be likely due to being underpowered studies. Bulluck et al. (34) conducted a meta-analysis to evaluate the role of IMR in detecting MVO at the time of primary PCI in STEMI patients. This study suggested that patients with a mean IMR > 41 were more likely to present MVO. We provided a meta-analysis using resistance measurements of IMR, HMR and Pzf. Our results suggested that MI patients with higher standardized weighted mean values favored the identification of MVO in CMR. The overall effect size was highly significant and with a very low variation across studies. Subsequently, we conducted a meta-analysis using only IMR studies, comprising a small number of cases (N = 142). Nonetheless, our results show that a weighted mean IMR of < 25 (upper limit of the 99% CI in group without MVO) were less likely to have MVO, while those with a weighted mean IMR of > 46 (lower limit of the 99% CI in the group with MVO) were likely to have MVO by CMR. Interestingly, an IMR cut-off value of > 40 had already been validated in previous studies to identify MVO and MACE in the MI setting. Also, the IMR cut-off < 25 is typically used to define patients without microvascular dysfunction in the literature.

Regarding doppler-derived indices (HMR, Pzf), they seemed to relate more consistently with CMR findings than IMR (1, 2), however, no significant differences were found in their diagnostic accuracy to identify MVO. Overall, the resistance indices presented a modest effect size correlation between them: IMR vs. HMR (r = 0.39, p < 0.001), HMR vs. Pzf (r = 0.55, p = 0.002; r = 0.75, p < 0.001) (1, 10, 16), which suggest that the three resistance indices may not be considered equivalent techniques for measuring microvascular function in the MI setting. Furthermore, IMR is known to be a reproducible approach to measure microvascular resistance in the acute setting. HMR is a more challenging technique due to specific doppler-flow-velocity tracing methodological procedures, leading to a more operator-dependent procedure, with a higher likelihood of a wider range of values, making it more difficult to determine a valid normal range of HMR values (see Figure 1) (10).

Novel microcirculatory indices are emerging, such as the Resistive Reserve Ratio (RRR). An intracoronary thermodilution parameter, which derives from IMR (ratio of basal resistance to IMR). Whereas IMR measures at peak hyperemia and reflects microcirculatory structure, RRR quantifies the vasodilator response of coronary microcirculation to a hyperemic stimulus. By expressing vascular resistance during resting conditions and maximal hyperemia, it can be interpreted as the ability of microcirculation to recover after an acute ischemic injury. Scarsini et al. reported that RRR may offer incremental prognostic value compared to other thermodilution-derived indices, such as IMR and CFR, in predicting the extent of MI in patients with STEMI (35). However, the limited sample size (N = 45) and retrospective design of the study, the lack of a validated threshold for RRR, overall limits the reproducibility of their results. Nevertheless, it has been observed that impaired microvascular vasodilatory function occurs in the presence of prolonged ischemia and MVO (36) and can predict HF hospitalizations (17). Larger studies are needed to define whether RRR adds significant information to the IMR in assessing microvascular function after MI.

Often studies that use myocardial hyperemia, do not consider patients with a blunted systemic vasodilation and reflex tachycardia (heart rise < 10 bpm) after an adenosine infusion. Patients with heart failure, diabetes and polymorphisms of adenosine A2a receptors, may present this reduced hemodynamic response to adenosine, but it is uncertain to what extent this lack of peripheral response is reflected in coronary vasodilation. Previously, Mishra et al. have reported that despite the lack of peripheral response to adenosine seen in some patients, coronary vasodilatation remained adequate for the purpose of myocardial perfusion imaging (37). A recent study compared stress myocardial blood flow (MBF) and hemodynamic response with different dosing regimens of adenosine during stress perfusion CMR (38). The authors reported that, although non-responders to standard dose adenosine had a significantly higher stress heart rate following a high dose of adenosine, they showed no significant difference in stress MBF. In contrast, regadenoson is a selective adenosine A2a receptor agonist, more potent vasodilator than adenosine with a better tolerability profile that has been proposed as a simpler alternative to adenosine infusion to achieve myocardial hyperemia. Previous studies have demonstrated regadenoson non-inferiority to adenosine for the detection of perfusion defects (39, 40) and despite regadenoson selective activity, there is considerable inter-patient variability in hemodynamic response, which seems not to be related to common A2a receptor polymorphisms (41).

Post-ischemic coronary microvascular dysfunction is a heterogeneous entity that can be found in up to 65% of STEMI patients. A 1% increase in MVO size associates with 14% relative increase in mortality and 8% increase in HF (42), and IMR has been validated to predict MACE following MI (Table 2). However, it is unclear which of either marker of microvascular dysfunction (MVO or invasive resistance indices) or even both should be used to improve risk stratification of MI patients. As previously discussed, invasive physiology often correlates poorly with MVO and a significant proportion of MI cases show discordance from the presence (or absence) of MVO and IMR values (≤40 or above 40). Furthermore, IMR ≤ 40 was associated with IS reduction at 6 months, irrespective of MVO status (14). The wide variability found between invasive and non-invasive markers of microvascular dysfunction suggests that invasive techniques and MVO on CMR measure distinct features of microvascular ischemic pathophysiology, which often occurs concurrently (Figure 6). The recent study by Scarsini (4), further confirmed the impact of MVO or elevated IMR on adverse events after STEMI. Although patients with IMR above 40 and MVO had the worst IS and clinical outcomes at 1 year follow up, patients with either high IMR or MVO, continued to undergo LV remodeling and tended to develop HF after the first year of MI.

Finally, data suggests that coronary hemodynamics are also altered in non-culprit IRA in the acute MI setting (23). These higher IMR values in non-IRA might be explained by MI size, the remodeling of remote myocardium, which occurs in response to increased overload, and by a higher stimulus of neuronal sympathetic axis which promote structural changes in coronary vessels and higher vascular resistances (28). However, abnormal IMR values in the non-culprit vessel seemed not to impact clinical prognosis of MI patients.

The potential clinical utility of microvascular resistance indices is under continuous investigation. The possibility of having a microcirculation assessment immediately available in the catheterization laboratory, broadens its applicability and interest in the early reperfusion of MI. More research is needed to define the clinical importance of non-IRA microcirculation after an acute MI, and the microcirculation assessment in patients with MINOCA represents an exciting field for future dedicated research.

Since all studies included in this review were observational, non-randomized, with small sample sizes, the readership should anticipate a high risk of type II error. As the lower statistical power of the studies may impede the recognition of microvascular function importance in the MI setting. Regarding the amount of invasive microvascular data available, IMR was disproportionately the most published vascular resistance index. Since HMR and Pzf have no validated cut-off values, different thresholds were used in the clinical trials through ROC curves analysis. The wide range of cut-off values used and the different microcirculation assessment techniques limit the comparison between invasive resistance indices. Moreover, the correlation between CMR findings and hemodynamic markers varied widely (Figure 5), probably due to methodological disparities between studies, raising doubts about the optimal timing for microvascular function assessment and how future clinical trials should be designed. Overall, studies did not consistently report mean ± SD values of microcirculation measurements, which limited the sample size and the performance of meta-analysis for other outcomes in this review. Regarding treatment management, despite most studies reported a similar frequency in the use of standard antithrombotic drugs in primary PCI (aspirin and P₂Y₁₂ inhibitors), the rates of GP IIb/IIIa and aspiration thrombectomy varied significantly between studies (22–71%). Other therapies that potentially impact coronary microcirculation (43) were not consistently reported in our cohort. Moreover, microcirculation studies evaluating clinical outcomes after MI showed a significant heterogeneity in the frequency of use of ACEi/ARA (81–99%), statins (65–94%), and betablockers (74–85%), which unenabled matching patients to their medications.

Microvascular resistance indices are valuable means to evaluate microcirculation function following MI. Microvascular dysfunction relates to the extent of myocardial damage and clinical outcomes after MI.

The original contributions presented in this study are included in the article/Supplementary material, further inquiries can be directed to the corresponding author.

LP and MS conceived and designed the analysis, data collection, and wrote the manuscript. RT performed and reviewed the analysis and critical revision. MF and LG revised the manuscript critically. All authors contributed to the article and approved the submitted version.