Author + information
- Received September 15, 2015
- Revision received November 30, 2015
- Accepted January 1, 2016
- Published online March 28, 2016.
- Guus A. de Waard, MDa,
- Maurits R. Hollander, MDa,
- Paul F.A. Teunissen, MDa,
- Matthijs F. Jansen, MDa,
- Elise S. Eerenberg, MDb,
- Aernout M. Beek, MD, PhDa,
- Koen M. Marques, MD, PhDa,
- Peter M. van de Ven, PhDc,
- Ingrid M. Garrelds, MScd,
- A.H. Jan Danser, PhDd,
- Dirk J. Duncker, MD, PhDe and
- Niels van Royen, MD, PhDa,∗ ()
- aDepartment of Cardiology, ICaR-VU, VU University Medical Center, Amsterdam, the Netherlands
- bDepartment of Vascular Medicine, Academic Medical Center, Amsterdam, the Netherlands
- cDepartment of Epidemiology and Biostatistics, ICaR-VU, VU University Medical Center, Amsterdam, the Netherlands
- dDepartment of Internal Medicine, Division Vascular Pharmacology, Erasmus Medical Center, Rotterdam, the Netherlands
- eDepartment of Experimental Cardiology, Thoraxcenter, Erasmus Medical Center, Rotterdam, the Netherlands
- ↵∗Reprint requests and correspondence:
Dr. Niels van Royen, VU University Medical Center, Department of Cardiology, De Boelelaan 1117, 1081 HV Amsterdam, the Netherlands.
Objectives The aim of this study was to determine the effects of an acute myocardial infarction (AMI) on baseline and hyperemic flow in both culprit and nonculprit arteries.
Background An impaired coronary flow reserve (CFR) after AMI is related to worse outcomes. The individual contribution of resting and hyperemic flow to the reduction of CFR is unknown. Furthermore, it is unclear whether currently used experimental models of AMI resemble the clinical situation with respect to coronary flow parameters.
Methods Intracoronary Doppler flow velocity measurements were obtained in culprit and nonculprit arteries immediately after successfully revascularized ST-segment elevation myocardial infarction (n = 40). Stable patients without obstructive coronary artery disease served as control subjects and were selected by propensity-score matching (n = 40). Similar measurements in an AMI porcine model were taken both before and immediately after 75-min balloon occlusion of the left circumflex artery (n = 11).
Results In the culprit artery, CFR was 36% lower than in matched control subjects (Δ = −0.9; 1.8 ± 0.9 vs. 2.8 ± 0.7; p < 0.001) with consistent observations in swine (Δ = −0.9; 1.5 ± 0.4 vs. 2.4 ± 0.9 for after and before AMI, respectively; p = 0.04). An increased baseline and a decreased hyperemic flow contributed to the reduction in CFR in both patients (baseline flow: Δ = +5 and hyperemic flow: Δ = −7 cm/s) and swine (baseline flow: Δ = +8 and hyperemic flow: Δ = −6 cm/s). Similar changes were observed in nonculprit arteries (CFR: 2.8 ± 0.7 vs. 2.0 ± 0.7 for STEMI patients and control subjects; p < 0.001). CFR significantly correlated with infarct size as a percentage of the left ventricle in both patients (r = −0.48; p = 0.001) and swine (r = −0.61; p = 0.047).
Conclusions CFR in both culprit and nonculprit coronary arteries decreases after AMI with contributions from both an increased baseline flow and a decreased hyperemic flow. The decreased CFR after AMI in culprit and nonculprit vessels is not a result of pre-existing microvascular dysfunction, but represents a combination of post-occlusive hyperemia, myocardial necrosis, hemorrhagic microvascular injury, compensatory hyperkinesis, and neurohumoral vasoconstriction.
- acute myocardial infarction
- coronary flow reserve
- coronary microcirculation
- microcirculatory dysfunction
Outcomes after acute myocardial infarction (AMI) have been drastically improved owing to timely primary percutaneous coronary intervention (PCI), as recommended by clinical guidelines (1). Although primary PCI successfully restores epicardial coronary blood flow in the majority of the cases, microcirculatory disturbances can still adversely affect myocardial perfusion (2,3). Patients with a blunted vasodilatory response after successful revascularization of the culprit artery, in other words, a decreased coronary flow reserve (CFR), have worse functional and clinical outcomes (4,5). The development of coronary flow before acute myocardial infarction (AMI) to immediately after revascularization remains unclarified in patients, however. Knowledge of the changes in flow could aid in better understanding the value of intracoronary physiological parameters in the acute phase after ST-segment elevation myocardial infarction (STEMI) for 2 reasons. First, in order to use CFR as a risk stratification tool and select patients who might benefit from adjunctive therapy, it is imperative that the foundations of the development of CFR due to AMI are fully understood. Second, recent trials suggest that revascularization of not only the culprit artery but all significantly obstructed coronary arteries is associated with an improved clinical outcome (6–8). In intermediate stenoses, revascularization should be guided by the fractional flow reserve (FFR) (7). Because FFR measurements are influenced by microvascular behavior (9), knowledge of the changes in flow during AMI in noninfarct-related arteries could be helpful to better interpret the intracoronary hemodynamic measurements in the acute phase.
In addition, studies aiming to improve outcomes after STEMI have recently received much attention for reducing microvascular injury (2,3). Many of these studies were conducted in animal models of reperfused AMI. It remains unclear to what extent these experimental models resemble the coronary and systemic hemodynamics in patients experiencing STEMI. In particular, it is unclear how these measurements relate to the final infarct size.
The primary objective of the present study was to investigate the changes in coronary blood flow before and after AMI in order to gain further insight into the microcirculatory behavior after AMI. This was done by studying Doppler flow velocity measurements obtained immediately after successful revascularization for STEMI in stable control subjects with a similar risk profile but without obstructive coronary artery disease. Similar measurements were obtained in a porcine model before and after reperfused AMI, which permitted investigation of the histopathological characteristics in relation to the flow responses after AMI. Second, we aimed to compare the coronary and systemic hemodynamics in this frequently used porcine model of reperfused AMI with the changes observed in patients. Finally, we investigated the development of neurohumoral activation as a possible contributor to the reduced CFR after AMI.
For the present study, we compared intracoronary Doppler flow velocity measurements in 3 settings: 1) immediately after successful primary percutaneous coronary intervention (PCI) for STEMI (n = 40); 2) from a propensity score–matched group of control subjects without coronary artery disease on angiography (n = 40); and 3) in an experimental porcine AMI model, before and after reperfusion (n = 11). Details of the flow acquisition protocol are provided in the section on hemodynamic recordings in animals and patients. Flow measurements were compared with infarct size as a percentage of the left ventricle (IS%LV) in patients using cardiac magnetic resonance imaging (MRI) and in swine by histopathology. Examples of these measurements are shown in Figure 1. All patients gave written informed consent, and the local ethics committee approved the study. Animal experiments conformed to the Position of the American Heart Association on research animal use (10) and were approved by the local Animal Ethics Committee.
For the STEMI cohort, data from the previously published PREDICT-MVO (Pressure-Flow Measurements Directly after Primary PCI to Predict Late Occurrence of Microvascular Obstruction) study was used (NTR3164) (11). In brief, STEMI patients presenting within 6 h after symptom onset and successfully treated with primary PCI according to standard procedures and clinical guidelines were enrolled (1). Successful PCI was defined as a Thrombolysis In Myocardial Infarction flow grade III (12) after revascularization. Exclusion criteria were an inability to provide informed consent, hemodynamic instability, a history of AMI in the culprit artery, coronary artery bypass grafting, 3-vessel disease, and poor renal function. Immediately after PCI, all patients gave oral informed consent, which was witnessed by an independent person, and written informed consent was obtained within 24 h after oral consent. In the ambulance, all patients received 5,000 IE heparin intravenously, 500 mg acetylsalicylic acid intravenously, and 60 mg prasugrel orally along with 1.75 mg/kg/h bivalirudin during the PCI procedure. Immediately after primary PCI, intracoronary Doppler flow velocity measurements were obtained in both the culprit artery and an angiographically unobstructed nonculprit artery supplying noninfarcted myocardium. After the procedure, standard maintenance medication was prescribed. Corrected Thrombolysis In Myocardial Infarction frame count immediately after primary PCI and ST-segment resolution at 1 h after primary PCI were analyzed as previously described (13,14). Because acquiring high-quality Doppler flow velocity measurements is challenging in the acute setting, the 40 measurements with the highest quality were selected from the 60 patients enrolled in the PREDICT-MVO study before data analysis. Cardiac MRI was performed between 4 and 6 days after PCI using a 1.5-T MR-scanner (Magnetom Avanto, Siemens, Erlangen, Germany). Cardiac MRI defined IS%LV; left ventricular end-diastolic and end-systolic volumes, left ventricular ejection fraction, and the presence of microvascular injury were determined using cine and late gadolinium enhancement imaging, as previously described (11). The myocardial salvage index (MSI) was calculated as the difference between edema (in grams), which was manually delineated as the hyperintense area on T2-weighted turbo spin-echo with fat suppression, and scar size (in grams) quantified on late gadolinium enhancement, divided by edema, and multiplied by 100. Cardiac MRI could not be obtained in 6 patients due to technical failure of cardiac magnetic resonance image acquisition in 1 patient, waist circumference exceeding the maximum of the MRI unit bore in 1 patient, claustrophobia in 3 patients, and because the 1 patient declined to undergo cardiac MRI.
Forty control patients with thoracic symptoms suspected to be coronary artery disease, without a history of ischemic, structural, or valvular heart disease and with a good left ventricular function, were selected by propensity-score matching. Control patients were drawn from a cohort of simultaneous intracoronary Doppler flow velocity and pressure measurements in 519 coronary arteries in 218 patients. In 182 arteries, there was an absence of focal stenoses, angiographically detectable collateral artery formation, severe calcification, and diffuse atherosclerotic narrowing, and these measurements were used for propensity-score matching. Propensity-score matching was based on sex, age, and the classic cardiovascular risk factors to select control patients who best matched the patients in the STEMI cohort.
Porcine model of reperfused AMI
The development of intracoronary Doppler flow velocity was studied in an AMI model of 75-min balloon occlusion of the left circumflex artery with subsequent reperfusion in 11 female Yorkshire swine. Doppler flow velocity measurements were taken in the infarct-related artery before and immediately after AMI. IS%LV was determined histopathologically at 7 days after AMI. To be able to compare animal results with patient results, a similar medication regimen was given. The evolution of neurohumoral activation was determined by consecutive measurement of endothelin-1 levels and plasma renin activity (PRA) at different time points. A detailed methodology for anesthesia, the experimental procedure, termination of life, quantification of IS%LV, and hemorrhage as a percentage of the infarct size are provided in the Online Appendix.
Hemodynamic recordings in animals and patients
To measure Doppler flow velocity, a similar methodology was used for both patient cohorts and the porcine reperfused AMI model. A 0.014-inch guidewire equipped with both a pressure and Doppler flow velocity sensor (ComboWire XT, Volcano Corporation, San Diego, California) was placed in the coronary artery. The guidewire was placed just distal to the site of balloon occlusion (porcine model) or just distal to the site of PCI in the culprit artery (STEMI patients), and in the distal third of the vessel in the nonobstructed nonculprit arteries (STEMI patients) as well as vessels of the propensity score–matched control patients. Before all measurements, intracoronary nitrates (200 to 300 μg) were administered. The Doppler guidewire was manipulated until an optimal and stable Doppler flow signal was obtained. Doppler flow velocity and aortic and distal pressure traces were recorded during both baseline and hyperemic conditions induced by intracoronary bolus injection of 150 μg adenosine. Instantaneous values of electrocardiographic voltage, peak Doppler flow velocity, and aortic and distal coronary pressure were continuously acquired at a sampling rate of 200 Hz using the ComboMap console (Volcano Corporation) and extracted for offline analysis. Mean Doppler flow velocity was averaged over a minimum of 3 consecutive heart beats (average peak velocity [APV] for both baseline [b-APV] and hyperemic [h-APV] conditions using custom software (written in Delphi v. 2010; Embarcadero, San Francisco, California). CFR was defined as the ratio between h-APV and b-APV.
Categorical data are presented as frequencies and percentages, whereas continuous data are presented as mean and SD. To test for differences between 2 groups of unpaired categorical variables, either the chi-square or Fisher exact test was used. Means of continuous variables were compared between groups using the independent-sample t test or paired-sample t test in case of paired data. Propensity scores were used for matching each patient in the STEMI cohort with a single patient in the control cohort (1:1 matching). Propensity-score matching was performed with SPSS version 22 (SPSS Inc., Chicago, Illinois) with the propensity scores based on the following variables: sex; age; and the classic risk factors of smoking history, diabetes mellitus, hypertension, family history of coronary artery disease, and hypercholesterolemia. Differences in the outcome variables CFR, b-APV, and h-APV were compared between the STEMI patients and the propensity score–matched control patients using the paired-sample t test. Associations between 2 continuous variables were tested using linear regression analysis, and Pearson’s correlation coefficient was determined to quantify the strength of the relationships. Temporal development of endothelin-1 and PRA levels was assessed by repeated-measures analysis of variance. A 2-sided p value of <0.05 was considered statistically significant. All statistical analyses were performed using SPSS version 20 (SPSS Inc.).
STEMI and control patient characteristics
Baseline variables of the STEMI patients and the matched control patients are shown in Table 1. Of the 40 STEMI patients included, 31 were male (78%) with a mean ± SD age of 59.3 ± 9.4 years. In control patients, 33 were male (83%) with a mean ± SD age of 58.0 ± 8.8 years. No significant differences were observed between any of the baseline variables for which matching was applied. Scar tissue was 23.6 ± 18.8 g on average, whereas IS%LV was 17.5 ± 12.4%.
Changes in coronary flow and CFR in the culprit artery
In STEMI patients, mean b-APV was 32% higher in STEMI patients compared with nonoccluded control vessels in propensity score–matched stable patients (21.2 ± 11.7 vs. 16.1 ± 6.3 cm/s, respectively; p = 0.009). The mean h-APV was significantly lower by 6.6 cm/s (18%) in STEMI patients compared with control subjects (42.5 ± 13.0 vs. 35.9 ± 18.3 cm/s, respectively; p = 0.03).
CFR is calculated using both b-APV and h-APV, and the increased b-APV together with the trend toward a lower h-APV resulted in a significant decrease in CFR in both patients and swine. In the culprit artery, CFR was 36% lower compared with stable controls (1.8 ± 0.9 vs. 2.7 ± 0.7, respectively; p < 0.001).
Changes in coronary flow and CFR in noninfarct-related territories
The changes in CFR in the nonculprit coronary artery showed a similar, but less pronounced, reduction as in the culprit artery. CFR was 2.0 ± 0.7 versus 2.8 ± 0.7 in nonculprit versus control arteries (p < 0.001), which amounted to a 29% reduction after STEMI. This reduction was driven by a nonsignificant 12% increase in b-APV (18.0 ± 7.6 vs. 16.1 ± 6.3 cm/s for nonculprit and control arteries, respectively; p = 0.24) in combination with a significant 16% reduction in h-APV (34.1 ± 14.8 vs. 42.5 ± 13.0 cm/s for nonculprit and control arteries, respectively; p = 0.003). Figure 2 shows blood flow for both the culprit and the unobstructed nonculprit arteries for the STEMI patients compared with the propensity score–matched stable control subjects. No significant differences were observed between the infarct-related artery and nonculprit CFR. When only patients with a large myocardial infarction (≥15 IS%LV) were evaluated, infarct-related CFR was significantly lower than nonculprit artery CFR (1.50 ± 0.40 vs. 1.78 ± 0.51 respectively; p = 0.04).
Characteristics of the porcine model
Swine had an age ranging between 11 and 14 weeks with a weight ranging between 27 and 40 kg. In 2 pigs, ventricular fibrillation occurred during AMI, which was successfully terminated by cardiac defibrillation and administration of amiodarone. After swift restoration of sinus rhythm, the remaining study procedures were carried out as described in the protocol. Before AMI, the mean heart rate was 65 ± 10 beats/min, which increased to 87 ± 24 beat/min after AMI (p = 0.02). Mean arterial pressure was 87 ± 13 mm Hg before and 74 ± 14 mm Hg immediately after reperfusion (p = 0.06).
For the porcine model, b-APV increased by 44% from pre-AMI to post-reperfusion (17.1 ± 5.5 vs. 24.6 ± 9.3 cm/s, respectively; p = 0.06). The same was observed in the porcine model, where h-APV decreased by 14%, from 39.9 ± 6.3 cm/s pre-AMI to 34.2 ± 3.0 cm/s post-AMI, but this was not statistically significant (p = 0.38). Similarly, in the porcine setting, CFR decreased by 37%, from 2.4 ± 0.9 pre-AMI to 1.5 ± 0.4 post-AMI (p = 0.04). Figure 3 summarizes the changes in flow and CFR induced by AMI.
Comparison of hemodynamics between patient study and porcine model
Tables 2 and 3 describe the hemodynamic characteristics in both the 2 patient cohorts and the porcine AMI model. Hemodynamic characteristics were compared between the controls and the situation before AMI in the porcine model, and no significant differences in the variables of heart rate, mean arterial pressure, b-APV, h-APV, and CFR were noted. This comparison was also made between AMI patients and the situation post-AMI in the porcine model, where mean arterial pressure was the only different hemodynamic parameter (86 ± 10 vs. 74 ± 14 mm Hg, respectively; p = 0.04).
Prognostic implications of CFR and its individual components
A significant inverse relationship between the culprit post-AMI CFR value and IS%LV was observed in both patients: r = −0.48; p = 0.005, and swine r = −0.61; p = 0.047 (Figure 4). A significant correlation was also found between the CFR after AMI in the nonculprit artery and IS%LV in patients (r = −0.37; p = 0.04). There was no significant relationship between the MSI and culprit CFR after AMI in patients (r = 0.22; p = 0.23). For patients, outcomes were stratified according to the median of cardiac MRI–defined IS%LV in Table 4. Compared with the subgroup with IS%LV below the median, the creatine kinase-myocardial band peak was higher, the left ventricular ejection fraction and the MSI were decreased, and cardiac MRI–defined microvascular injury occurred more often in the subgroup with IS%LV above the median. Of the flow parameters specifically investigated in this study, CFR was 1.5 ± 0.4 in the subgroup with IS%LV above the median and 2.2 ± 1.2 in the subgroup with IS%LV below the median (p = 0.02). This difference in CFR could mainly be attributed to a significantly lower h-APV in the subgroup with IS%LV above the median (29.7 ± 11.0 vs. 42.5 ± 22.7 cm/s; p = 0.04), whereas b-APV was statistically equivalent in both subgroups (20.9 ± 8.0 vs. 21.7 ± 13.9 cm/s for the upper vs. lower median of IS%LV, respectively; p = 0.84). For the nonculprit coronary artery, similar observations for CFR, h-APV, and b-APV were made when stratified according to the median of IS%LV. This stratification according to functional outcome could not be performed for the porcine model due to insufficient data.
In the porcine model, concentrations of the vasoactive substances endothelin-1 and PRA (as a marker for activation of the renin-angiotensin-aldosterone system) were determined at consecutive time points (Figure 5). Endothelin-1 concentration showed a significant rise and fall over time (repeated-measures analysis of variance, p = 0.02). Immediately after AMI, a significant increase occurred compared with before AMI (4.4 ± 1.3 vs. 2.7 ± 0.7 pg/ml; p = 0.001), which remained significant after 3 h (5.9 ± 2.9 vs. 2.7 ± 0.7 pg/ml; p = 0.02). After 24 h, endothelin-1 values returned to normal (3.1 ± 1.1 vs. 2.7 ± 0.7 pg/ml; p = 0.46), which persisted at 7 days after AMI (3.1 ± 1.1 vs. 2.7 ± 0.7 pg/ml; p = 0.19). PRA concentration remained stable over time (ranging between 0.20 ± 0.13 and 1.03 ± 1.13 pmol Ang 1/ml∙h; repeated-measures analysis of variance; p = 0.08).
Development of coronary flow in relation to intramyocardial hemorrhage
Intramyocardial hemorrhage was present in all swine, and the mean hemorrhage as a percentage of the infarct size was 30.7 ± 13.4%. Hemorrhage as a percentage of the infarct size significantly correlated with CFR after AMI (r = −0.71; p = 0.02) (Figure 6).
In the present study, we investigated changes in myocardial blood flow in successfully reperfused AMI by using paired observations in patients with STEMI and a matched control cohort, as well as an experimental porcine model of reperfused AMI. Our findings are as follows. 1) A reduction in CFR is observed after AMI both because of an increase in b-APV and a decrease in h-APV. 2) CFR is significantly and inversely related to infarct size, driven by a lower h-APV in large infarctions. 3) In nonculprit vessels, b-APV, h-APV, and CFR developed similarly as in the culprit vessel, but with less pronounced changes. 4) The porcine model and patient study showed remarkable concordance with respect to the coronary blood flow responses to AMI. The clinical implications of each of these findings are discussed in the following.
Changes in hyperemic and resting component of CFR after AMI
h-APV was significantly lower immediately after AMI in the patient study and showed a numeric, nonsignificant decrease in the experimental porcine study. When stratified according to the median of cardiac MRI–defined IS%LV, patients with larger infarctions had a significantly lower h-APV with more frequent occurrence of cardiac MRI–defined microvascular injury compared with patients with smaller infarctions. The reduced hyperemic flow in these large infarctions is likely governed by injury to the microvasculature, whereas in smaller infarctions, h-APV remained similar (2). Previously we found that cardiac MRI–defined microvascular injury comprises almost exclusively of intramyocardial hemorrhage (15). Histopathological examination of the porcine model revealed that the percentage of hemorrhage within the infarcted myocardium was related to lower values of CFR, pointing to an important role of hemorrhagic microvascular injury in the observed reduction of CFR after AMI.
b-APV, in contrast to its hyperemic counterpart, markedly increased immediately after AMI in both the experimental model and patients. This may be caused by reactive hyperemia in which metabolites with vasodilatory properties, such as endogenous adenosine, build up due to prolonged ischemia (16). The degree of resting myocardial perfusion after AMI is not static, but develops over time and varies depending on the timing of the measurement. In an experimental animal study, Ambrosio et al. (17) demonstrated that on reperfusion after ischemia, an initial steep increase in myocardial blood flow occurs that dissipates over time and eventually decreases to less than pre-ischemia myocardial flow. In both the patient and porcine studies, we obtained Doppler flow velocity measurements immediately after reperfusion was established, and as such, the baseline flow was likely increased due to post-occlusive vasodilation. Interestingly, b-APV was similar in small and large infarctions (as defined by the median of IS%LV), suggesting that the increase in b-APV due to myocardial ischemia is not necessarily related to myocardial necrosis.
Together, the decrease in h-APV and the increase in b-APV resulted in a 36% mean decrease in CFR after AMI. Possibly the increased vasoactive state, as indicated by the temporary increase in endothelin-1 levels immediately after AMI, could have acted to globally reduce CFR by invoking arteriolar vasoconstriction and thereby contribute to the no-reflow phenomenon. PRA levels remained constant, however, and only showed a modest (nonsignificant) increase at 7 days post-AMI. As such, it is unlikely that the renin-angiotensin-aldosterone system–mediated arteriolar vasoconstriction contributed much to the immediate reduction in CFR.
Prognostic implications of CFR
Our study demonstrates that culprit Doppler flow velocity–derived CFR measured immediately after revascularization significantly correlated with functional outcome after STEMI, not only in patients, but this finding was also confirmed in the experimental porcine model. Although the correlation between CFR and IS%LV was significant, it was only modest, likely because CFR is influenced by a multitude of factors. In a recent study, Cuculi et al. (18) did not find invasive CFR measured immediately after revascularized STEMI by thermodilution to predict functional outcome. Although methodological differences in the quantification of CFR could perhaps explain the discrepant results, the overall variability in CFR might preclude clinical adoption as a risk stratification tool for AMI patients. Culprit CFR showed no significant relationship with cardiac MRI–defined MSI in patients. It is expected that CFR is lower in the infarctions with the least amount of myocardium salvaged due to the no-reflow phenomenon. The fact that no such relationship was found can also be of a methodological nature because Kim et al. (19) recently showed that cardiac MRI–defined edema (used in the calculation of the MSI) is actually a poor reflection of true area at risk.
Changes in flow parameters in nonculprit vessels
CFR was not only decreased in the culprit artery, but also in the unobstructed nonculprit coronary artery compared with stable control patients. This reduction was predominantly driven by an attenuated h-APV, especially in the infarctions with IS%LV above the median. This finding has implications for physiology-guided revascularization of nonculprit arteries in the setting of STEMI. Three recent randomized clinical trials (6–8) found favorable results for complete revascularization compared with culprit artery–only revascularization. In 1 of these studies, the DANAMI-3—PRIMULTI (Primary PCI in Patients With ST-elevation Myocardial Infarction and Multivessel Disease: Treatment of Culprit Lesion Only or Complete Revascularization) trial (6), revascularization of nonculprit arteries was guided by physiological interrogation using FFR. Because lesion assessment by FFR relies on induction of the maximal hyperemic state to correctly estimate the flow downstream of a stenosis (20), an attenuated microcirculatory response to adenosine in the nonculprit arteries could lead to underestimation of stenosis severity. Hyperemic flow steadily normalizes in the months after AMI (21), and an initial negative FFR result may become positive as the coronary flow response to hyperemia is restored. Revascularization guided by resting physiological indexes such as the instantaneous wave-free ratio, however (22), could lead to overestimation of hemodynamic stenosis significance in the acute phase because our findings show that nonculprit resting flow is increased after AMI. In this study, a numerically lower CFR after AMI was in the infarct related compared with the noninfarct-related artery; however, this difference was not statistically significant, whereas previous work had demonstrated significant differences (21,23). It is conceivable that the wide range of infarct sizes in this study (range, 1% to 50% IS%LV) causes heterogeneity in CFR values as observed by large SDs, precluding statistical significance. This is evidenced by a significantly lower culprit CFR compared with nonculprit CFR, when only patients with ≥15 IS%LV are evaluated.
Porcine model of AMI is a reliable method to study myocardial reperfusion injury
Both the initial hemodynamic status before AMI as well as the hemodynamic changes after AMI showed remarkable concordance between the experimental porcine and the patient studies. This concordance supports the validity of the conclusions drawn on the changes in coronary blood flow before AMI and after reperfusion. Furthermore, it shows that the experimental porcine myocardial infarction and reperfusion model using balloon occlusion of the left circumflex artery is a robust method with which to study the microcirculatory disturbances after reperfused AMI.
Reduced CFR can exist independently of pre-existing microcirculatory dysfunction
CFR in an unobstructed nonculprit coronary artery was significantly lower than CFR in propensity score–matched control patients. This implies that microcirculatory function in a nonculprit artery does not accurately reflect the microcirculatory status before AMI. Studies have shown that alterations of microvascular tonus and structure after AMI extend to noninfarcted territory. We find that CFR after AMI in the nonculprit artery was also decreased and significantly related to IS%LV. In the acute phase, increased left ventricular filling pressures, compensatory hyperkinesis of remote areas takes place (24), and neurohumoral activation causes global coronary arteriolar vasoconstriction (25), which might explain the decreased CFR in nonculprit arteries. In the chronic phase, altered neurohumoral activation persists (26), and ventricular remodeling of infarcted as well as noninfarcted myocardium occurs (27). In 1994, Uren et al. (23) already noted that CFR in remote, noninfarcted regions was lower than in healthy control subjects and did not fully recover over time. However, due to the design of that study, it was not clear whether this could be attributed to pre-existing impairment of microvascular function or to the neurohumoral and structural changes from the myocardial infarction itself. In the present study, we studied CFR after AMI, but used patients with a comparable risk profile for microcirculatory dysfunction for comparison with the STEMI patients instead of healthy control patients. Similar alterations for CFR and its individual components were found in the patient study, with unknown pre-existing microcirculatory status and in the healthy porcine model, where coronary microvascular dysfunction is certainly nonexistent. In addition, similar relationships between a reduced CFR and functional outcome defined by IS%LV were found in patients and animals. These 2 observations suggest that a reduction in CFR attributable to AMI can occur independently of pre-existing microvascular dysfunction.
An a priori imperfect match due to an abundance of unknown variables that influence microcirculatory function inherently limits the use of propensity-score matching in this study. The validity of the study results, however, are supported by closely matching baseline variables between STEMI and control patients. Further support of our conclusions is provided by similar results in the porcine model and patient study. A second limitation of the study is that Doppler flow velocity measurements in an unobstructed nonculprit artery were not obtained in the porcine model. Third, due to the complexity of the study protocol, the sample size was relatively small. This should be taken into account when interpreting the results of the study, especially in the analyses that stratified patients according to the median of the IS%LV. Finally, it would be of interest to have intracoronary physiological measurements at later points in time. The invasive nature of these measurements, however, precludes follow-up measurements due to ethical reasons.
CFR in both culprit and nonculprit coronary arteries decreases after AMI with contributions of both an increased b-APV and a decreased h-APV. The decreased culprit artery CFR after AMI is not a result of pre-existing microvascular dysfunction, but rather represents a combination of post-occlusive hyperemia, the extent of myocardial necrosis, and hemorrhagic microvascular injury. In nonculprit arteries, the CFR was also related to the size of the infarction, which may be attributed to compensatory hyperkinesis and neurohumoral-mediated vasoconstriction.
WHAT IS KNOWN? The CFR after AMI is frequently impaired, and this is related to a worse functional and clinical outcome. Flow changes before AMI and after revascularization remain undescribed.
WHAT IS NEW? In this study, an increased baseline flow and a reduction in hyperemic flow contributed to the impaired CFR in both culprit and nonculprit arteries. This could influence hemodynamic lesion assessment in the subacute phase of AMI. Furthermore, coronary blood flow after AMI was similar in human patients and the experimental AMI porcine model.
WHAT IS NEXT? The results of our study confirm that microcirculatory impairment as determined by CFR predicts adverse outcome and indicates that the porcine model is a useful way to study microcirculatory dysfunction after AMI.
For supplemental material, please see the online version of this article.
Unrestricted research grants from Volcano Corporation and Biotronik supported this work. Dr. Royen received an unrestricted research grant from Volcano Corporation and Biotronik. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- acute myocardial infarction
- baseline average peak velocity
- coronary flow reserve
- hyperemic average peak velocity
- fractional flow reserve
- infarct size as a percentage of the left ventricle
- magnetic resonance imaging
- myocardial salvage index
- percutaneous coronary intervention
- plasma renin activity
- ST-segment elevation myocardial infarction
- Received September 15, 2015.
- Revision received November 30, 2015.
- Accepted January 1, 2016.
- American College of Cardiology Foundation
- O’Gara P.T.,
- Kushner F.G.,
- Ascheim D.D.,
- et al.
- Ibanez B.,
- Heusch G.,
- Ovize M.,
- Van de Werf F.
- Wakatsuki T.,
- Nakamura M.,
- Tsunoda T.,
- et al.
- Engstrom T.,
- Kelbaek H.,
- Helqvist S.,
- et al.
- Gershlick A.H.,
- Khan J.N.,
- Kelly D.J.,
- et al.
- van de Hoef T.P.,
- Nolte F.,
- EchavarrIa-Pinto M.,
- et al.
- Teunissen P.F.,
- de Waard G.A.,
- Hollander M.R.,
- et al.
- Chesebro J.H.,
- Knatterud G.,
- Roberts R.,
- et al.
- Gibson C.M.,
- Cannon C.P.,
- Daley W.L.,
- et al.
- Nijveldt R.,
- Beek A.M.,
- Hirsch A.,
- et al.
- Robbers L.F.,
- Eerenberg E.S.,
- Teunissen P.F.,
- et al.
- Eikens E.,
- Wilcken D.E.
- Ambrosio G.,
- Weisman H.F.,
- Mannisi J.A.,
- Becker L.C.
- Cuculi F.,
- Dall'Armellina E.,
- Manlhiot C.,
- et al.
- Kim H.W.,
- Van A.L.,
- Jennings R.B.,
- et al.
- Pijls N.H.,
- van Son J.A.,
- Kirkeeide R.L.,
- De Bruyne B.,
- Gould K.L.
- Teunissen P.F.,
- Timmer S.A.,
- Danad I.,
- et al.
- Sen S.,
- Escaned J.,
- Malik I.S.,
- et al.
- Lew W.Y.,
- Chen Z.Y.,
- Guth B.,
- Covell J.W.
- Haitsma D.B.,
- Bac D.,
- Raja N.,
- Boomsma F.,
- Verdouw P.D.,
- Duncker D.J.
- Theroux P.,
- Ross J. Jr..,
- Franklin D.,
- Covell J.W.,
- Bloor C.M.,
- Sasayama S.