Microvascular Resistance Predicts Myocardial Salvage and Infarct Characteristics in ST‐Elevation Myocardial Infarction

Background The pathophysiology of myocardial injury and repair in patients with ST‐elevation myocardial infarction is incompletely understood. We investigated the relationships among culprit artery microvascular resistance, myocardial salvage, and ventricular function. Methods and Results The index of microvascular resistance (IMR) was measured by means of a pressure‐ and temperature‐sensitive coronary guidewire in 108 patients with ST‐elevation myocardial infarction (83% male) at the end of primary percutaneous coronary intervention. Paired cardiac MRI (cardiac magnetic resonance) scans were performed early (2 days; n=108) and late (3 months; n=96) after myocardial infarction. T2‐weighted‐ and late gadolinium–enhanced cardiac magnetic resonance delineated the ischemic area at risk and infarct size, respectively. Myocardial salvage was calculated by subtracting infarct size from area at risk. Univariable and multivariable models were constructed to determine the impact of IMR on cardiac magnetic resonance–derived surrogate outcomes. The median (interquartile range) IMR was 28 (17–42) mm Hg/s. The median (interquartile range) area at risk was 32% (24%–41%) of left ventricular mass, and the myocardial salvage index was 21% (11%–43%). IMR was a significant multivariable predictor of early myocardial salvage, with a multiplicative effect of 0.87 (95% confidence interval 0.82 to 0.92) per 20% increase in IMR; P<0.001. In patients with anterior myocardial infarction, IMR was a multivariable predictor of early and late myocardial salvage, with multiplicative effects of 0.82 (95% confidence interval 0.75 to 0.90; P<0.001) and 0.92 (95% confidence interval 0.88 to 0.96; P<0.001), respectively. IMR also predicted the presence and extent of microvascular obstruction and myocardial hemorrhage. Conclusion Microvascular resistance measured during primary percutaneous coronary intervention significantly predicts myocardial salvage, infarct characteristics, and left ventricular ejection fraction in patients with ST‐elevation myocardial infarction. (J Am Heart Assoc. 2012;1:e002246 doi: 10.1161/JAHA.112.002246)

prognosis. 2 Despite successful primary percutaneous coronary intervention (PCI), at least one third of patients subsequently develop microvascular obstruction that portends an adverse prognosis. 2,3 Ex vivo studies in experimental animals have suggested that maintained microvascular patency in ischemic myocardium limits infarct size. 4 However, in vivo human studies are lacking because it has not been possible to measure coronary microvascular function directly. Standard clinical methods for assessing the efficacy of reperfusion, such as the electrocardiogram 5 and angiographic perfusion grade, 6 are indirect and have limited diagnostic accuracy for culprit artery microvascular function. Microvascular damage revealed noninvasively by contrast-enhanced cardiac magnetic resonance (CMR) imaging is an independent predictor of prognosis after acute myocardial infarction (MI). 3 However, because of issues around feasibility and cost, CMR is not normally performed early after MI. When CMR is available, it is usually performed up to 7 days after hospital admission in medically stabilized patients. Therefore, a direct quantitative measure of microvascular function during primary PCI would be an important advance for very early risk stratification in routine clinical practice.
The index of microvascular resistance (IMR) is a simple coronary guidewire-based method for assessing coronary microvascular function, [7][8][9][10][11] including in patients with STelevation MI (STEMI). 10,11 IMR measured at the end of primary PCI is an independent predictor of left ventricular ejection fraction as measured by echocardiography 10 or CMR 11 3 months after PCI. These observations suggest that IMR can be used to study the pathophysiology of microvascular function in STEMI and that IMR could have utility for very early prognostication.
Given that myocardial salvage can be measured with T 2weighted CMR, 2,12-16 we aimed to investigate the ability of acutely measured IMR to predict subsequent myocardial salvage and left ventricular remodeling in a cohort of patients with STEMI. Our first hypothesis was that microvascular resistance measured by IMR after culprit artery reperfusion is a determinant of subsequent myocardial salvage. Our second hypothesis was that IMR is a predictor of left ventricular function during follow-up. Our final hypothesis was that IMR would predict the occurrence and severity of infarct characteristics, such as myocardial hemorrhage and microvascular obstruction, which have prognostic importance. 2,3 Methods

Patient Population and STEMI Management
Consecutive acute STEMI patients undergoing primary PCI at a regional cardiac center were screened. The inclusion criteria were an indication for primary PCI for acute STEMI due to a history of symptoms consistent with acute myocardial ischemia and with supporting changes on the electrocardiogram (ie, STsegment elevation or new left bundle-branch block). 17 Exclusion criteria represented standard contraindications to CMR, including an estimated glomerular filtration rate <30 mL/min per 1.73 m 2 . The project was approved by the West of Scotland Research Ethics Committee.
Acute STEMI management followed contemporary guidelines. 17 Aspiration thrombectomy, direct stenting, antithrombotic drugs, and other therapies were administered according to clinical judgment.

IMR Measurement After Coronary Reperfusion
A coronary pressure-and temperature-sensitive guidewire (St Jude Medical, Uppsala, Sweden) was used as the primary guidewire. The guidewire was calibrated outside the body, equalized with aortic pressure at the ostium of the guide catheter, and then advanced to the distal third of the culprit artery. IMR is defined as the distal coronary pressure multi-plied by the mean transit time of a 3-mL bolus of saline at room temperature during maximal coronary hyperemia, measured simultaneously (mm Hg · s, or units). [7][8][9][10][11] Hyperemia was induced by 140 μ/kg per minute of intravenous adenosine preceded by a 2-mL intracoronary bolus of 200 μg nitrate. The mean aortic and distal coronary pressures were recorded during maximal hyperemia. Previous studies in patients with stable coronary disease have established that IMR is repeatable and independent of hemodynamic variations, including heart rate, blood pressure, and myocardial contractility. 8,9 In our study, the repeatability of IMR was assessed by duplicate measurements 5 minutes apart in a subset of 12 consecutive patients.

CMR Acquisition and Analyses
CMR was performed approximately 2 days after MI and 3 months after MI on a Siemens MAGNETOM Avanto (Erlangen, Germany) 1.5-Tesla scanner with a 12-element phased-array cardiac surface coil. The imaging protocol included cine MRI with steady-state free precession, edema imaging with brightblood T 2 -weighted CMR, 12,18,19 and delayed-enhancement phase-sensitive inversion-recovery pulse sequences. 20 The bright-blood T 2 -weighted ACUT2E method (Acquisition for Cardiac Unified T 2 Edema) 18,19 incorporates elements of the bright-blood 21 and turbo spin-echo methods. 22 Typical imaging parameters were as follows: acquisition time 7 to 12 seconds, matrix 192×192, flip angle 180 • , echo time 1.69 ms, and bandwidth 789 Hz/pixel. Twenty-nine coherent spin echoes (echo train length) were obtained per heartbeat, and the time interval (echo spacing) between the 180 • inversion pulses was 3.4 ms. Data were acquired every second RR interval. The voxel size was 1.9×1.9×6 mm 3 . Microvascular obstruction was defined as a central dark zone on early delayed-enhancement imaging 1, 3, 5, and 7 minutes after contrast injection and within an area of late gadolinium enhancement (LGE). MI was imaged by using a segmented phase-sensitive inversion-recovery turbo fast lowangle shot 18

MR Image Analyses
The images were analyzed on a Siemens workstation by 2 cardiologists (A.R.P., C.B.) with 3 and 5 years' CMR experience, respectively. 12,19,23 These cardiologists have considerable CMR experience in patients with acute MI, including protocols with bright-blood T 2 -weighted CMR. A third CMR trained-cardiologist (N.T.) provided an independent opinion when disagreement occurred, and agreement was established by consensus. All of the cardiologists were blinded to the PCI and IMR results. Left ventricular dimensions, volumes, and ejection fractions were quantified with computer-assisted planimetry.
Reference ranges used in the laboratory were 105 to 215 g for left ventricular mass in men, 70 to 170 g for left ventricular mass in women, 77 to 195 mL for left ventricular end-diastolic volume in men, 52 to 141 mL for left ventricular end-diastolic volume in women, 19 to 72 mL for left ventricular end-systolic volume in men, and 13 to 51 mL for left ventricular end-systolic volume in women.

Myocardial Edema and Standardized Measurements of T 2 -Weighted Area at Risk
Quantitative assessments were performed by 2 independent observers, and their results were averaged. The approach to image analysis, including interobserver agreement, has been reported previously. 12,19 The jeopardized area at risk on each axial image was defined as the percentage of left ventricular area delineated by the hyperintense zone on T 2 -weighted images. 12,19 Infarct Definition and Size The presence of acute infarction was established with CMR on the basis of abnormalities in cine wall motion, rest first-pass myocardial perfusion, and LGE imaging. In addition, supporting changes on the electrocardiogram and coronary angiogram were required. Acute infarction was considered present only if LGE was confirmed on both the short-and long-axis acquisitions. The myocardial mass of LGE (in grams) was quantified by a semiautomatic detection method that used a signal intensity threshold of >2 standard deviations (SDs) above a remote reference region. 11,15,19 Infarct regions with evidence of microvascular obstruction were included within the infarct area. Microvascular obstruction was classified as relevant (central dark zone with a subendocardial or intramural distribution) or nonrelevant (dots or nil) and also was expressed as a percentage of total left ventricular mass.

Myocardial Salvage
Myocardial salvage was calculated by subtraction of percent infarct size from percent area at risk. 12 The myocardial salvage index was calculated by dividing the myocardial salvage area by the initial area at risk.

Definition and Detection of Hemorrhage on T2-Weighted CMR
Myocardial hemorrhage revealed by bright-blood T 2 -weighted CMR was defined as a confluent dark zone with a mean signal intensity <2 SDs of the mean signal intensity of the surrounding affected brighter area. 23 Hemorrhage was expressed as a percentage of total left ventricular mass.

Angiographic Analyses
All angiograms were analyzed by 2 interventional cardiologists independent of the CMR analyses. The Thrombolysis In Myocardial Infarction (TIMI) flow classification 24 and corrected TIMI frame count 25 were used to grade culprit artery flow at initial angiography and at the end of the PCI. The TIMI myocardial perfusion grade was used to assess myocardial perfusion. 26

Statistical Analyses
The primary end point was myocardial salvage index. The sample size estimate was based on the adjusted effect (multiple regression) for the relationship between IMR measured acutely and myocardial salvage index as revealed by contrastenhanced CMR, after adjustment for prespecified known prognostic clinical confounders. In multiple-regression models for myocardial salvage index, 80 subjects would provide ≥80% power to detect a correlation of 0.3 with a type 1 error rate of 0.05. With allowance for occasional problems with data acquisition, the projected sample size was ≥100 subjects. All subjects with available data were used in models fitted for CMR parameters at initial and follow-up scans. Normality was assessed by using the Shapiro-Wilk test. Means and SDs were used to summarize normally distributed data, geometric means and SDs to describe skewed data, percentages to summarize distributions, and Pearson coefficients (r) to describe correlations between log IMR and CMR parameters. Normal distributions were achieved where necessary by logarithmic transformation. All tests were 2 tailed, with a significance level of 0.05.
Between-group comparisons of continuous variables used Student t tests or Mann-Whitney tests, whereas differences in proportions were assessed with either a Fisher exact test or a χ 2 test with continuity correction, as appropriate. Randomeffects models were used to compute interrater reliability measures (intraclass correlation coefficient [ICC]) for CMR parameters measured independently by 2 observers in 20 randomly selected patients. Univariable and multivariable regression models were constructed to determine the prognostic significance of IMR with CMR parameters: myocardial salvage index, infarct size, microvascular obstruction, hemorrhage, area at risk, and left ventricular ejection fraction and volumes. Further subgroup analysis explored patients with and without a culprit left anterior descending coronary artery at initial angiography and patients with and without an occluded culprit coronary artery. We used a backward variable selection process with the cut point of 0.1 to select variables for the final multivariable models. IMR was specified in each multivariable model, but because renal function and serum troponin are not available at the time of primary PCI when IMR is measured, these variables were not included. Multivariable models adjusted for variables in stepwise selection models. For all baseline CMR outcomes, adjustment was made for the following variables: previous MI, sex, pain-to-balloon time, door-to-balloon time, thrombectomy, catheterization laboratory pulse pressure, TIMI flow, corrected TIMI frame count, and percent of ST elevation still present after PCI. For all follow-up CMR outcomes, we adjusted for baseline value of the CMR parameter, age, door-to-balloon time, glycoprotein IIb/IIIa, TIMI flow, and percent of ST elevation still present after PCI. IMR associations were described by an effect estimate (either additive or multiplicative) on the continuous CMR parameters per 20% increase in IMR, with 95% confidence intervals (CIs) and P values. All statistical analyses were performed in SAS (SAS Institute, Cary, NC).

Patient Characteristics
One hundred eight patients with acute STEMI treated by primary PCI were enrolled between October 1, 2009, and July 8, 2010 (Table 1). Forty-seven patients (44%) had anterior STEMI, and 61 patients (56%) had nonanterior MI (P=0.1). IMR was measured at the end of primary PCI in all patients. Intravenous adenosine was well tolerated in all patients, and there were no adverse events. All patients were alive after 3 months' followup.
The Figure 1 illustrates observations in a patient with inferolateral STEMI, including the IMR value measured at the end of primary PCI and subsequent myocardial salvage by CMR 2 days later.

CMR Findings
One hundred eight patients underwent CMR an average of 2 days after primary PCI. Ninety-six patients (89%) returned for a Data are given as mean±SD, n (%), or * median (interquartile range).
CMR was performed during the index hospitalization (n=108) and 3 months afterward (n=96). The clinical characteristics of the patients with paired CMR scans at baseline and at follow-up (n=96) were similar to those of patients with a single scan at baseline (n=108).
follow-up scan ≈3 months later ( Table 2). The baseline characteristics of the patients who did or did not attend the 3-month follow-up MRI scan were similar. In an assessment of interobserver agreement, 20 patients were randomly selected for independent analysis by 2 observers.

Microvascular Resistance and CMR Findings Early (2 Days) After MI
The patients were divided into quartiles according to IMR.
With higher values of IMR, the area at risk and initial infarct size increased and left ventricular ejection fraction decreased Data are given as n (%) or mean±SD. * Values are geometric mean±SD.
The CMR findings in patients with paired scans (n=96) were similar to those of patients with a single scan at baseline (n=108). (

Univariable and Multivariable Relationships Between IMR and Infarct Characteristics
IMR was a univariable predictor of area at risk and early infarct size, myocardial salvage, microvascular obstruction, hemorrhage, and left ventricular ejection fraction. IMR was a univariable predictor of infarct size and left ventricular ejection fraction at 3 months (Table 4). At 2 days, multivariable models confirmed significant predictive effects of IMR on initial injury characteristics, including area at risk, infarct size, myocardial salvage, microvascular obstruction, and myocardial hemorrhage (Table 5). At 3 months, after adjustment for values at 2 days, IMR continued to show a significant predictive effect on infarct size. In separate linear models of IMR and initial infarct size, microvascular obstruction, and hemorrhage, and after adjustment for left ventricular mass, IMR was more closely associated with infarct size (data not shown) than with either hemorrhage or microvascular obstruction (P<0.001 for both comparisons).

Patients With Anterior or Nonanterior Infarcts
In a comparison of patients with anterior and nonanterior infarctions (Table 6), IMR was a multivariable predictor of early  infarct characteristics, including infarct size, myocardial salvage, microvascular obstruction, and myocardial hemorrhage, in both groups. In patients with anterior MI, at 3 months, after adjustment for initial CMR findings, IMR predicted infarct size, myocardial salvage, and left ventricular ejection fraction (Table 6).

Patients With Occluded or Patent Culprit Arteries at Presentation
In a comparison of patients with an occluded culprit artery (TIMI flow grade 0/1) at initial presentation to those with a patent culprit artery (TIMI flow grade 2/3) (Table 7), IMR was a multivariable predictor of early infarct characteristics, including infarct size, microvascular obstruction, and hemorrhage, in both groups. At 3 months, IMR predicted myocardial salvage in patients with an occluded artery but not after adjustment for the initial extent of salvage (data not shown).

Microvascular Resistance and Left Ventricular Remodeling
Multivariable models with left ventricular end-diastolic and end-systolic volumes at 3 months showed no association with IMR after adjustment for initial volumes (Tables 5, 6, and 7).

Discussion
Our results indicate that after primary PCI for STEMI, microvascular resistance assessed by IMR is a significant predictor of myocardial salvage. Second, microvascular resistance was associated significantly with left ventricular ejection fraction in all patients initially and in patients with anterior MI after 3 months' follow-up. Finally, in addition to predicting myocardial salvage, microvascular resistance was a significant predictor of several infarct characteristics, including infarct size, microvascular obstruction, and myocardial hemorrhage. Importantly, IMR provided additive predictive value to established measures of coronary/myocardial reperfusion, including time to reperfusion, coronary flow grades, and electrocardiogram changes. IMR adds early prognostic information at the key time of emergency reperfusion therapy when CMR is not possible, and, potentially, IMR represents a measurable therapeutic target after coronary reperfusion in STEMI. Because microvascular resistance significantly predicted initial myocardial salvage, we conclude that microvascular patency is a key pathophysiological determinant of the injury/repair response in the infarcted human heart. This observation is a new insight into the pathophysiology of MI and is consistent with previous experimental observations in large animal models. 4 Independent of other early predictors of myocardial salvage, such as time to reperfusion and initial infarct size, our findings suggest that microvascular function influences the balance between recovery of viable injured tissue and progressive infarction. The pathophysiology of microvascular injury in STEMI is complex and encompasses embolization of proximal thrombus and in situ microvascular thrombosis, as well as extravascular pathologies such as edema and hemorrhage. 27 We also have demonstrated for the first time that microvascular resistance shortly after coronary reperfusion is associated with subsequent evidence of myocardial hemorrhage as revealed by CMR, consistent with the notion that hemorrhage is an adverse complication of more severe MI. Our study extends the results of Fearon et al, 10 who found that IMR predicted left ventricular wall motion score at 3 months after MI in 28 STEMI patients, as well as our own data in a prior study of a separate cohort of patients in whom CMR was used to assess ventricular function. 11 Our present study confirms these observations in a much larger cohort but also demonstrates that microvascular resistance predicts infarct characteristics, including initial microvascular obstruction and infarct size, especially in patients with anterior MI or an occluded artery.
Intuitively, it makes sense that IMR did not predict final left ventricular ejection fraction in all patients because its predictive value is most likely to be discriminative in patients with large infarctions. For example, a small MI might have little impact on systolic function (eg, occlusion of a left ventricular branch of the right coronary artery) even if microvascular injury is pronounced within distribution of the culprit artery. One other explanation could be that because the sample size is relatively small, there might have been insufficient power to detect a true inverse association between IMR and ejection fraction. We also found that IMR was more closely associated with infarct size than with microvascular injury (ie, microvascular obstruction or myocardial hemorrhage). Because IMR was performed before CMR, we cannot say whether infarct burden is a determinant of microvascular dysfunction or vice versa. However, IMR also correlated with the extent of microvascular obstruction and hemorrhage, which points to the pathological importance of the extent and not just the presence of these abnormalities. 3 The accuracy and precision of IMR has been established in patients with stable coronary artery disease, but its performance in STEMI has not been fully explored. 9 We found that IMR was highly repeatable when measured twice 5 minutes apart at the end of primary PCI.
We have shown that IMR is predictive of myocardial hemorrhage as revealed by bright-blood T 2 -weighted CMR. 23 In the current study, IMR was nearly 2-fold higher in patients with myocardial hemorrhage than in patients without hemorrhage. Myocardial hemorrhage could represent an unintended consequence of evidence-based antithrombotic and antiplatelet therapies, a manifestation of more severe MI (as suggested by our data), or both. Our other main finding is that early myocardial salvage is a significant predictor of subsequent changes in left ventricular function. This result complements the cohort study by Dall'Armellina et al, 15 who performed serial CMR scans in 30 STEMI patients over 6 months of follow-up and found that regression in infarct size during follow-up predicted the recovery in wall motion. Taken together, both studies indicate that the propensity for improvement in the initial extent of injury (ie, myocardial repair and salvage) is a determinant of future changes in left ventricular systolic function. On the basis of our findings, microvascular patency might permit recovery of salvageable myocardium, whereas microvascular injury might predispose to limited salvage and impaired systolic function in the longer term.
Our results raise the possibility that microvascular injury might be a measurable therapeutic target after coronary reperfusion in STEMI. For this to be the case, IMR should identify patients who might benefit from a specific treatment. Ideally, the treatment should be given at the time of primary PCI, leading to a measurable reduction in IMR or an improvement in prognostically important surrogate measures of microvascular injury. [4][5][6] Results from preliminary therapeutic trials during primary PCI for STEMI support the notion that microvascular injury is a therapeutic target. [28][29][30] Limitations Microvascular patency is only one of several characteristics that influence myocardial salvage and systolic function after MI, and our multivariable models account for only part of this variability, as reflected by the R 2 values. We could have used a larger signal intensity difference between remote and affected myocardium on LGE CMR to delineate infarct area, but we believe that the main conclusions in our analysis remain valid. Although IMR a multivariable predictor of left ventricular volumes after 3 months, this was not the case after adjustment for volumes measured early after MI. The predictive value of IMR for adverse remodeling needs further prospective study in a larger cohort.

Conclusions
Microvascular function is a determinant of myocardial salvage in patients with reperfused STEMI and predicts changes in left ventricular function and remodeling. IMR seems to be most informative in patients with an anterior MI or occluded culprit artery.