Microvascular Function in Viable Myocardium After Chronic Infarction Does Not Influence Fractional Flow Reserve Measurements

Study Population

Patients were included in the study if they fulfilled the following criteria: a documented single myocardial infarction more than 2 mo before the study, a nonoccluded infarct-related artery with a proximal culprit lesion, and at least one normal non–infarct-related coronary artery. Patients whose infarct-related or normal non–infarct-related artery was a nondominant right coronary artery were not included. PET and FFR measurements were performed on the same day in 22 patients; in the remaining 5 patients, the interval between the 2 procedures was less than 1 wk. For all patients, recent information was available from dobutamine stress echocardiography or ^99mTc-sestamibi perfusion scintigraphy on the extent and location of the infarcted area and whether it was ischemic. The study was approved by the medical ethical committee of the VU University Medical Centre. Written informed consent was obtained from all patients.

Myocardial Perfusion Scintigraphy

SPECT was performed according to a 2-d stress/rest protocol. Exercise or adenosine was used for the stress images. ^99mTc-labeled sestamibi was injected at maximal exercise or after intravenous infusion of adenosine (0.14 mg/kg per minute). SPECT was performed using a 2-head γ-camera equipped with low-energy high-resolution collimators. The images were acquired using a 360° circular orbit and were analyzed using a 13-segment model (11). Stress and rest segments were semiquantitatively scored as normal (grade 0) or as having a mild, moderate, or severe (grade 3) perfusion defect. Perfusion defects were allocated to the territory of a coronary artery. Defects in the anterior and anteroseptal region were allocated to the left anterior descending coronary artery, defects in the posterolateral wall to the left circumflex coronary artery, and inferior and basal inferoseptal defects to the right coronary artery. Apical defects were considered to be in the region of the left anterior descending coronary artery, unless the defect extended to the posterolateral (left circumflex coronary artery) or inferior (right coronary artery) wall. Perfusion defects were considered to be reversible and positive for the presence of ischemia when the rest perfusion score improved by one grade or more over the stress score. Segments with irreversible perfusion defects or normal perfusion were considered negative for the presence of ischemia. The scintigrams were scored by 2 experienced cardiologists; in cases of disagreement, a third cardiologist scored the images to achieve a majority decision.

Dobutamine Stress Echocardiography

An intravenous infusion of dobutamine was started at a rate of 10 μg/kg per minute and was increased by 10 μg/kg per minute every 3 min until either wall motion abnormalities were observed or a maximal rate of 40 μg/kg per minute was reached. In patients who did not reach 90% of their age-adjusted maximal heart rate and who had no objective signs of ischemia, 0.25 mg of atropine was given every minute up to a maximum of 1.0 mg while the dobutamine infusion was continued. Endpoints for stopping the infusion were the same as mentioned in the guidelines (12). Two-dimensional echocardiography was performed, obtaining parasternal long- and short-axis views and apical 4- and 2-chamber views. Imaging was performed throughout the study and during recovery until resolution of new wall motion abnormalities. Online digital images in quad screen format were analyzed for the presence, extent, severity, and location of segmental wall motion abnormalities. Myocardial contractile function was graded as normal, hypokinetic, akinetic, or dyskinetic in each segment. An echocardiographic stress test was considered positive when new or worsening stress-induced wall motion abnormalities were observed. The standard algorithm was used to assign ventricular segments to coronary territories: left anterior descending coronary artery (basal and midanteroseptal segments; basal, mid, and apical anterior segments; and mid and apical septal and apical lateral segments), left circumflex coronary artery (basal and mid lateral segments and basal and mid posterior segments), and right coronary artery (basal, mid-, and apical inferior segments and basal septal segments) (13). The dobutamine stress echocardiograms were scored by 2 experienced cardiologists; in cases of disagreement, a third cardiologist scored the images to achieve a majority decision.

PET Study

A short rectilinear transmission scan was performed to ascertain that the heart was positioned in the center of the view of the scanner. Then, a complete transmission scan was obtained to correct the subsequent emission scans for tissue attenuation. After the transmission scan, myocardial perfusion was measured using a dynamic H₂¹⁵O scan sequence. This measurement was taken at baseline and after hyperemia, induced by intravenous administration of adenosine (140 μg/kg/min). Finally, a static C¹⁵O scan was obtained to image the blood volume. Anatomic tissue images were generated by subtracting the blood pool from the transmission images. Transaxial anatomic tissue images of the left ventricle were reoriented according to the anatomic axis of the heart and displayed as short-axis slices. The same reslicing parameters were applied to the dynamic H₂¹⁵O images. Regions of interest were defined manually on the transaxial images at the basal, midventricular, and apical levels. Each basal and midventricular slice was divided in 6 equidistant sectors starting from the posterior insertion of the right ventricular free wall into the left ventricle. Corresponding regions of interest from a variable number of slices were grouped in each patient to generate 13 volumes of interest (6 basal, 6 midventricular, and 1 apical). Additional regions of interest were defined in the left and right ventricular chambers. The latter set was projected on the dynamic H₂¹⁵O images to generate image-derived input functions. The standard single-tissue-compartment model was used together with these input functions to determine myocardial blood flow (mL/min/mL of perfusable tissue), including intrinsic corrections for spillover from both left and right ventricles (14).

The original 13 regions of interest were regrouped into 5 larger segments (anteroseptum, inferoseptum, anterior, lateral, and inferoposterior walls). The apex and the anterior and anteroseptal walls were allocated to the left anterior descending coronary artery, the lateral wall to the left circumflex coronary artery, and the inferoseptum and inferoposterior wall to the right coronary artery. The RFR was defined as the ratio of myocardial perfusion in the infarcted area to perfusion in the contralateral normally perfused area, at hyperemia.

Quantitative Angiography and FFR Measurement

A 6- or 7-French guiding catheter was positioned in the coronary ostium, and the aortic pressure was measured. After intracoronary administration of 0.2 mg of isosorbide dinitrate, an angiogram of the infarct-related artery was made in 2 orthogonal projections. Nitrates were given to maximize epicardial dilatation. In this way, flow-mediated vasodilatation, which may affect the FFR measurement, was prevented. In addition, nitrates prevent coronary spasm from being induced by the presence of the wire in the coronary stenosis. Quantitative coronary angiography was performed off-line using the CAAS II system (Pie Medical Data). A pressure wire (WaveWire; Volcano Therapeutics) was advanced to the tip of the guiding catheter. At that time, both pressures were verified to be identical. Then, the wire was positioned distal to the coronary stenosis and the distal pressure was recorded. Maximal hyperemia was induced by intracoronary administration of 40 μg of adenosine. FFR was calculated as the ratio of distal pressure to aortic pressure at maximal hyperemia. Afterward, FFR was also determined in the contralateral reference coronary artery.

Statistics

All data are presented as mean ± SD. Differences in continuous variables between groups were compared using the Student t test. The relationship between FFR and RFR was assessed using linear regression analysis. A value of P less than 0.05 was considered significant. Receiver operator characteristics were calculated to determine the best FFR cutoff value to predict ischemia in the infarcted area by noninvasive testing. The sensitivity, specificity, and FFR cutoff to determine ischemia were calculated.

Rationale for Measuring RFR with H₂¹⁵O PET in Chronic Myocardial Infarction

H₂¹⁵O PET provides quantitative information on regional flow in the myocardium (15,16). In patients with chronic infarction, the amount of viable myocardium is decreased and so also is maximal achievable blood flow to the infarcted area, even in the absence of an infarct-related artery stenosis. Because the H₂¹⁵O PET technique expresses flow per milliliter of perfusable (viable) tissue only (and not per milliliter of total tissue in the infarcted area), the decreased amount of viable myocardium within the infarcted area is accounted for and, thus, comparison of blood flow between infarcted and reference myocardial areas is possible.

Correlation Between FFR and RFR in Chronic Myocardial Infarction

In patients without prior myocardial infarction, FFR measurements have been validated through comparison with RFR (17). FFR equals RFR if 3 conditions are fulfilled. The first of these is that significant narrowing of the reference coronary artery be absent; therefore, in the present study, FFR in the reference coronary artery was measured. The second condition is that FFR and RFR measurements be performed under similar hemodynamic conditions; in the present study, the rate–pressure products at maximal hyperemia during FFR measurement and during PET did not differ significantly. The third condition is that microvascular resistance be identical in the stenotic and contralateral areas during maximal hyperemia. This condition does not mean that microvascular resistance had to be normal in the reference and infarcted areas. Coronary flow reserve and maximal myocardial blood flow have been shown to be significantly higher in healthy volunteers than in myocardial regions perfused by normal coronary arteries in patients with remote areas of myocardial ischemia (18). In concordance with the data of Uren et al., maximal myocardial blood flow data measured in volunteers at our institution (4.31 ± 0.90 mL/min/mL) were also significantly higher than the reference area flow data in the present study (19). The implication is that microvascular resistance in the reference area in the present study was higher than truly normal resistance. However, maximal myocardial blood flow data and absolute flow reserve in the reference area (3.3 ± 1.2) were comparable between the present study and other H₂¹⁵O PET studies on patients with coronary artery disease (18,20–22). This comparability suggests that the degree of microvascular functional impairment in the present study was similar to that found by other studies.

In the acute and subacute phases of a myocardial infarction, microvascular resistance in the infarcted area can be increased (23–25). There are no data on microcirculatory resistance in chronically infarcted areas in patients. If microcirculatory function in a chronically infarcted area is similar to that in a reference area, FFR in an infarct-related artery should be identical to RFR. If microvascular resistance in a chronically infarcted area remains greater than that in a reference area, FFR will be higher than RFR. In this situation, FFR will overestimate RFR proportionally by the degree to which microvascular resistance is higher in the infarcted than the reference area. In the present study, RFR was measured with H₂¹⁵O PET. In infarcted areas, H₂¹⁵O-PET provides perfusion data only on tissue that is capable of exchanging water rapidly—a property of viable myocardium (7–9). Myocardial blood flow in nonperfusable (scar) tissue is negligible. The fact that in the present study the regression line of the relationship between FFR and RFR approached the line of identity (Fig. 1A), and that the mean difference in the Bland–Altman plot was not significantly different from zero (Fig. 1B), indicates that microvascular resistance did not significantly differ between the infarcted and reference areas.

The FFR Cutoff After Myocardial Infarction

In the present study, the optimal cutoff of FFR to detect ischemia in the infarcted area was 0.79, resulting in a sensitivity of 81% and a specificity of 82%. An FFR cutoff of 0.75 had a sensitivity of 69% and a specificity of 82%. These findings are in line with those of previous studies, performed during the subacute phase of an infarction. De Bruyne et al. found in 57 patients with a recent myocardial infarction (average time, 20 d) that the best cutoff of FFR was 0.78 (26). In another study, FFR measurements were compared with SPECT 4 d after myocardial infarction. The optimal FFR for discriminating inducible ischemia also was 0.78 (27). Usui et al. compared the reliability of FFR in infarct-related and non–infarct-related artery stenoses. They found a similar sensitivity and specificity in patients with and without prior infarction at an FFR cutoff of 0.75 (28). In all 3 studies, the ejection fraction was rather well preserved (58%, 53%, and 56%, respectively). The present study extends the validity of the 0.75 FFR cutoff to patients with chronic infarction and considerably diminished left ventricular function.

Limitations

PET and pressure measurements were not performed simultaneously. However, no difference in rate–pressure products between the 2 procedures was observed at hyperemia. Furthermore, hyperemia during PET and pressure measurements was induced by intravenous and intracoronary adenosine administration, respectively. Studies have shown, however, that both intracoronary and intravenous administration of adenosine can achieve maximal hyperemia (29). In the postinfarction setting, the accuracy of noninvasive stress testing is decreased, possibly influencing the FFR cutoff and its sensitivity and specificity. The FFR cutoff in the chronic phase of myocardial infarction should be confirmed in a larger study population.