Accelerated ^99mTc-sestamibi clearance associated with mitochondrial dysfunction and regional left ventricular dysfunction in reperfused myocardium in patients with acute coronary syndrome

ACS patients were prospectively recruited from the Department of Cardiovascular Medicine at the Hokkaido University Hospital from August 2006 to February 2012. We enrolled patients diagnosed with ACS [17] who had revascularization immediately after admission to the hospital [mean age 69.2 ± 8.7 years, 10 males (56 %)]. ACS included unstable angina, non-ST-elevated myocardial infarction (NSTEMI), and ST-elevated myocardial infarction (STEMI) [18]. Exclusion criteria were (1) patients with prior myocardial infarction, (2) patients who were younger than 20 years old, and (3) patients whose condition was unstable. The study was approved by the Hokkaido University Graduate School of Medicine Human Research Ethics Board. Written informed consent was obtained from all patients.

Within 20 days of the onset of ACS, patients had rest and delayed rest MIBI single-photon emission computed tomography (SPECT), rest ¹¹C-acetate PET, and echocardiography at rest. These three imaging data acquisitions were performed within 7 days. The interval between nuclear imaging and echocardiography was 1.7 ± 2.4 days (Fig. 1). Follow-up echocardiography was performed 6 months after the onset of ACS (Fig. 1).

MIBI SPECT myocardial perfusion imaging was performed at rest. Six hundred megabecquerels (MBq) of MIBI (FUJIFILM RI Pharma, Tokyo, Japan) was intravenously administered at rest. Standard data acquisition was performed 30 min after MIBI administration [19], and an additional rest SPECT data acquisition was performed 180 min after MIBI administration [5].

All images were obtained using a dual-detector gamma camera (Millennium MG, General Electric, Elgems, Tirat Carmel, Israel) equipped with a parallel hole, low-energy, high-resolution collimator. Energy discrimination was provided by a 20 % window centered at 140 keV. Thirty-two images were obtained over a 180° arc. Each image was acquired over 30 s. The data were stored on a 64 × 64 matrix. A series of 6.78-mm-thick contiguous trans-axial images were reconstructed with a filtered back-projection algorithm without attenuation correction. These trans-axial images were then reoriented in the short axis, vertical long axis, and horizontal long axis of the left ventricle.

The LV wall was divided into 16 segments based on the American Society of Echocardiography (ASE) recommendations [20]. Each basal and mid-ventricular region was divided into six segments, and the apical region was divided into four segments. The border between the anteroseptal and anterior segments was considered to be at the anterior insertion of the right ventricular wall into the left ventricle. Also, the inferior insertion of the right ventricular wall into the left ventricle defined the border between the inferoseptal and inferior segments [20]. The association between the myocardial segments and the three major coronary arteries was also defined based on the ASE recommendations [20]. The left anterior descending artery (LAD) region included six segments, the left circumflex artery (LCx) region included five segments, and the right coronary artery (RCA) region included five segments. Two nuclear cardiologists independently performed visual evaluation of myocardial perfusion imaging. The observers performed image evaluations blinded to the patients’ clinical information and other imaging data. Discordant findings were resolved by a third observer. A standard five-point visual scoring system was used for evaluating regional myocardial MIBI uptake: 0 normal perfusion, 1 mild reduction, 2 moderate reduction, 3 severe reduction, and 4 absent uptake [19, 21]. Accelerated MIBI myocardial clearance was defined as an increase of 1 or more in the segmental defect score at the additional delayed rest image obtained at the 3-h mark after MIBI administration as compared with the early rest image [5]. MIBI redistribution was defined as a decrease of 1 or more in the segmental defect score at the delayed rest image as compared with the early rest image in segments with reduced uptake at initial rest imaging [22]. The percentage peak uptake was also analyzed in each LV segment using the Heart Function View software (Nihon Medi-Physics Co., Ltd., Tokyo, Japan).

Echocardiography was performed early after reperfusion therapy and at 6 months after the onset of ACS (Fig. 1). All echocardiographic examinations were performed by experienced echocardiographers blinded to the clinical information, PET image findings, and SPECT image findings. We used a commercially available ultrasound system (Sonos 5500, Philips Medical Systems, Andover, MA, USA) equipped with a broadband harmonic phased-array transducer (S4 probe). LV wall motion was evaluated using the 16-segment model based on the ASE recommendations [20]. LV wall motion was evaluated using a five-point scoring system according to the ASE guidelines: 1 normokinesis, 2 mild hypokinesis, 3 severe hypokinesis, 4 akinesis, and 5 dyskinesis [20]. Six months after the onset of ACS, we evaluated the regional wall motion as a follow-up study (Fig. 1). A decrease of more than 1 in the LV wall motion score at follow-up compared with the initial study was defined as regional wall motion improvement [23]. Left ventricular hypertrophy (LVH) was defined as interventricular septal wall thickness or posterior wall thickness of 11 mm or more [20].

Patients were instructed to fast for at least 6 h prior to ¹¹C-acetate PET study. Patients were positioned with the heart centered in the field of view in a whole-body PET scanner (ECAT HR+, Siemens/CTI Knoxville, TN, USA) [9, 24]. A dynamic PET acquisition was initiated (10 × 10, 2 × 30, 5 × 100, 3 × 180, and 2 × 300 s) [10, 25] just after intravenous administration of 740 MBq of ¹¹C-acetate. Blood pressure, heart rate, and electrocardiography were monitored during the PET scans.

PET image data were analyzed using an image analysis package (Dr. View; Asahi Kasei, Tokyo, Japan) and a special dedicated in-house software called the Hokkaido Quantitative Tool (HOQUTO) [9, 26]. The images were iteratively reconstructed and resliced along the short axis [27]. Based on the ASE recommendations [20], regions of interest were defined for each of the 16 segments.

Regional oxidative metabolism was determined from the mono-exponential function (k _mono) fit to the linear portion of the semilogarithmic plot. The mono-exponential fit began at the point where the blood pool was stable (usually 2 to 4 min after injection) as previously described [8, 24, 28]. All data were analyzed by nuclear cardiologists blinded to clinical information and other imaging data.

Continuous variables were expressed as mean plus standard deviation. Categorical variables were described as number and percentage. The Wilcoxon signed-rank test was performed between the initial and follow-up echocardiographic measurements. We evaluated the difference in oxidative metabolism (k _mono) among the three types of segments, namely those in group N, group AC, and group F, using a linear mixed effects model. Random effects were defined as subject, and fixed effects were defined as segment type. Then, we analyzed the differences among the three groups using multi-group comparison. Logistic regression analysis was performed to analyze the relationship between the time of ACS onset to reperfusion and the presence or absence of accelerated MIBI clearance. All statistical analyses were performed using R version 3.0.2 (The R Foundation for Statistical Computing, Vienna, Austria).

We enrolled 18 ACS patients who had immediate revascularization for the culprit vessel. No patient had a diagnosis of hypertrophic cardiomyopathy. Nine patients were revascularized for LAD. Two patients were revascularized for the LCx artery and the remaining seven patients for the RCA (Table 1). All patients were revascularized using a coronary artery stent and obtained TIMI flow grade 3 by the end of percutaneous coronary intervention (PCI). The required time from ACS onset to revascularization was 5.4 ± 6.8 h. Peak creatine kinase (CK) level was 1697.6 ± 1522.5 IU/L (range 112–5056 IU/L, Table 2). One patient had already undergone PCI for the LAD artery due to angina pectoris prior to ACS onset. However, this patient’s culprit region at the ACS event was the RCA. Therefore, we included this patient in the present study. Detailed information related to the ACS events of each patient is provided in Table 2.

Initial echocardiography study was performed 7.4 ± 2.8 days after ACS onset (range 2–12 days). LVH was observed in eight patients based on the ASE guidelines criteria [20]. At the initial study, group AC (n = 14) showed higher wall motion scores than group N (n = 64) (2.29 ± 0.99 vs 1.47 ± 0.62, p < 0.01). Group F also showed higher wall motion scores than group N (n = 17) (2.41 ± 1.06 vs 1.47 ± 0.62, p < 0.01).

MIBI SPECT was performed 8.9 ± 3.7 days after ACS onset (range 3–20 days). Accelerated MIBI clearance was observed in 10 out of 18 patients (56 %). The time from ACS onset to revascularization was not associated with the presence or absence of accelerated MIBI clearance (p = 0.52). There was no significant correlation between the time from revascularization to SPECT and numbers of segments (R = 0.23, p = 0.36). In addition, the time from ACS onset to initial MIBI SPECT imaging was not associated with the presence or absence of accelerated MIBI clearance (p = 0.20). Among a total 288 LV segments, we evaluated 99 segments related to revascularized coronary arteries. Four segments showed MIBI redistribution in a delayed rest image, and these segments were excluded from the analysis [29]. Among the remaining 95 segments, 64 were defined as having normal myocardial perfusion at early and delayed rest images (group N). Fourteen segments showed accelerated MIBI clearance (group AC), and 17 segments showed fixed perfusion defect (group F).

In group AC, the defect score significantly increased at delayed rest images as compared with early rest images (p < 0.001) (Table 3). Group N and group F showed similar defect scores at early and delayed rest images (group N p = 0.32 and group F p = 0.33) (Table 3). The percentage peak uptake was analyzed in 17 patients. Reduction of percentage peak uptake tended to be higher in group AC than in group N (Tables 4 and 5).

¹¹C-acetate PET was performed 8.8 ± 2.9 days after ACS onset (range 3–13 days). Group AC showed a lower oxidative metabolism (k _mono) than group N (0.041 ± 0.009 vs 0.049 ± 0.010, p = 0.02) (Figs. 2 and 3). Group F also showed a lower oxidative metabolism (k _mono) than group N (0.039 ± 0.012 vs 0.049 ± 0.010, p = 0.01). However, there was no difference in oxidative metabolism (k _mono) between group AC and group F (p = 0.99).

At the follow-up (mean follow-up 7.8 ± 4.5 months), three patients had in-stent restenosis in the reperfused coronary artery region as shown through coronary angiography. One patient did not have follow-up echocardiography. Therefore, these four patients were excluded from the follow-up echocardiographic study, and we evaluated regional wall motion changes between the initial and follow-up study in 14 patients. Among 224 segments in these 14 patients, 74 segments were related to revascularized coronary arteries, and we evaluated these segments as a follow-up study. Group N showed improved LV wall motion score compared with that at its initial study (n = 47) (1.47 ± 0.62 to 1.25 ± 0.53, p = 0.002) (Fig. 4). Group AC also showed improved LV wall motion score compared with that at its initial study (n = 10) (2.29 ± 0.99 to 2.00 ± 0.82, p = 0.04). In contrast, the regional wall motion score in group F did not change compared with that at its initial evaluation (n = 17) (2.41 ± 1.06 to 2.47 ± 1.12, p = 0.82).

Approximately 90 % of MIBI accumulates on the mitochondrial membrane in relation to its electrical gradient [3, 30]. In experimental studies, loss of mitochondrial membrane potential was associated with a decrease in MIBI uptake [3]. Myocardial cell injury caused accelerated MIBI clearance in ischemic myocardium in a canine model [31, 32]. In addition, in hypertrophic cardiomyopathy, there was an association between accelerated MIBI clearance and the change of mitochondrial structure as observed in histological examinations [16, 33]. Hayashi et al. reported that accelerated MIBI clearance was correlated with the severity of degeneration in the mitochondria in dilated cardiomyopathy [16]. In the current study, segments with accelerated MIBI clearance showed decreased myocardial oxidative metabolism in patients with ACS. Therefore, reduced oxidative metabolism may reflect mitochondrial dysfunction in ischemic myocardium as well as in either hypertrophic or dilated cardiomyopathy. These data appear to corroborate those from previous experimental studies, and the current data may therefore clarify the possible mechanisms of accelerated MIBI clearance in patients with ACS.

In the current study, accelerated MIBI clearance was observed in ten patients (55 %), and segmental analysis showed that 14 of the 99 segments (14 %) exhibited accelerated MIBI clearance. Takeishi et al. reported that 15 of the 22 patients (68 %) with acute myocardial infarction showed accelerated MIBI clearance [5]. In their study, these patients had percutaneous transmural coronary angioplasty and thrombolytic therapy. The frequency of accelerated MIBI in the current study was lower than in Takeishi’s study. In their study, all patients who underwent angioplasty obtained TIMI grades 2 or 3 by the end of the revascularization procedure. The difference in the frequency of accelerated MIBI between the current study and Takeishi’s study may be due to differences in coronary intervention approach. In the current study, all patients underwent stent placement as the ACS treatment. In addition, all patients had obtained TIMI grade 3 by the end of the revascularization procedure. Given recent developments in treatment approaches, ACS treatments may have improved in comparison with those used in previous studies. Therefore, newer revascularization approaches may result in less myocardial damage and a lower frequency of accelerated MIBI clearance. Even with the lower frequency of accelerated MIBI clearance, the main finding of the current study agreed with the findings from the previous study, and the current study adds new pathophysiological insights over the previous study.

Fujiwara et al. reported that the regions of accelerated MIBI clearance were closely correlated with those showing reduced uptake of ¹²³iodine-beta-methyl-iodophenyl-pentadecanoic acid (BMIPP) [34]. BMIPP defect is associated with abnormal myocardial fatty acid metabolism [35]. Their data indicate that accelerated MIBI clearance may be associated with myocardial metabolic dysfunction in ACS. Although they did not evaluate oxidative metabolism, their data may support the current findings.

In the current study, reduction of percentage peak uptake in the segments of accelerated sestamibi clearance showed a trend of being higher than that in normal perfusion segments. However, this was not significant. Since tracer decay might have impacts on the percentage peak uptake, this may have had an influence on the relative uptake analysis.

The regions with accelerated MIBI clearance showed LV wall motion improvement at follow-up after revascularization. Using low-dose dobutamine stress echocardiography, Fujiwara et al. evaluated the association between accelerated MIBI clearance and regional wall motion after ACS (within 7 days of admission) [6]. Segments with accelerated MIBI clearance showed better functional recovery during low-dose dobutamine administration than those with fixed MIBI defects. Their study revealed accelerated MIBI clearance associated with dysfunctional but viable myocardium early after ACS. The current study further added to the new insight that accelerated MIBI clearance in segments was related to regional wall motion recovery at follow-up. Thus, the current study provided insights into the association between accelerated MIBI clearance and regional LV functional recovery in addition to those provided by previous studies [5, 6].

Myocardial oxidative metabolism in the segments with accelerated MIBI clearance decreased as the number of segments with fixed perfusion abnormality did. Segments with accelerated MIBI clearance showed improved LV wall motion. However, the LV wall motion in segments with fixed perfusion abnormality remained unchanged at follow-up. Thus, based on the current findings, it may be difficult to predict regional wall motion recovery using ¹¹C-acetate PET data. The effectiveness of ¹¹C-acetate PET to predict myocardial functional recovery has not been fully recognized [36‐40]. Hicks et al. reported that oxidative metabolism did not depend on myocardial perfusion. They also reported that oxidative metabolism varied in accordance with myocardial conditions such as ischemia or infarction [37]. They concluded that predicting functional recovery through the evaluation of oxidative metabolism alone was difficult. Our results suggest that segments with impaired oxidative metabolism were associated with myocardial injury. However, this finding was not sufficient to predict myocardial functional recovery based on the current data. Therefore, further studies are required.

Evaluation of accelerated MIBI clearance may provide additional information in a clinical setting. Our study may indicate that segments with accelerated MIBI clearance represent myocardium damage as a result of mitochondrial dysfunction.

Our study had some limitations. First, histological findings for the regions showing accelerated MIBI clearance were not evaluated. Rather than performing biopsy sampling, we evaluated myocardial oxidative metabolism as a marker of mitochondrial dysfunction using ¹¹C-acetate PET. In experimental studies, the clearance of ¹¹C-acetate from myocardium is associated with mitochondrial function [15]. Therefore, the current data support our hypothesis. Second, myocardial oxidative metabolism was not evaluated at follow-up. A previous study reported an improvement in oxidative metabolism in the reperfused myocardium [37]. Examining this information with regard to physiological changes to ACS after reperfusion therapy might be an important next step. Third, for evaluation purposes, we did not separate patients with ACS into groups based on the following pathological conditions: unstable angina, NSTEMI, and STEMI. These three types of pathological conditions may exhibit related pathology in terms of myocardial damage. Finally, our study involved a small study population. The current protocol included two sestamibi SPECT data acquisitions, echocardiography, and one ¹¹C-acetate PET study. ¹¹C-acetate PET is usually applied for specific pathophysiological studies in a limited number of research facilities [25, 41], and it would be difficult to apply this comprehensive study protocol to a large number of subjects. In addition, the sample size of the present study was small but similar to that of previous studies [39, 41, 42]. Despite the small sample size, with careful preparation, we showed the pathophysiological mechanism of accelerated sestamibi washout. While a small sample size may have had a minimal impact on the current data, we definitely need further study using a larger study population to confirm the efficacy of evaluating MIBI clearance. The time from revascularization to SPECT showed some variability. However, there was no significant correlation between the time from revascularization to SPECT and the numbers of accelerated sestamibi clearance segments. Therefore, time from revascularization to SPECT might not have had an impact on the numbers of segments with accelerated sestamibi clearance.