https://orcid.org/0000-0002-5187-8310
Abstract
Emotional stress is associated with future cardiovascular events. However, the mechanistic linkage of brain emotional neural activity with acute plaque instability is not fully elucidated. We aimed to prospectively estimate the relationship between brain amygdalar activity (AmygA), arterial inflammation (AI), and macrophage haematopoiesis (HEMA) in acute myocardial infarction (AMI) as compared with controls.
18F-fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG-PET/CT) imaging was performed within 45 days of the index episode in 62 patients (45 with AMI, mean 60.0 years, 84.4% male; 17 controls, mean 59.6 years, 76.4% male). In 10 patients of the AMI group, serial 18F-FDG-PET/CT imaging was performed after 6 months to estimate the temporal changes. The signals were compared using a customized 3D-rendered PET reconstruction. AmygA [target-to-background ratio (TBR), mean ± standard deviation: 0.65 ± 0.05 vs. 0.60 ± 0.05; P = 0.004], carotid AI (TBR: 2.04 ± 0.39 vs. 1.81 ± 0.25; P = 0.026), and HEMA (TBR: 2.60 ± 0.38 vs. 2.22 ± 0.28; P < 0.001) were significantly higher in AMI patients compared with controls. AmygA correlated significantly with those of the carotid artery (r = 0.350; P = 0.005), aorta (r = 0.471; P < 0.001), and bone marrow (r = 0.356; P = 0.005). Psychological stress scales (PHQ-9 and PSS-10) and AmygA assessed by PET/CT imaging correlated well (P < 0.001). Six-month after AMI, AmygA, carotid AI, and HEMA decreased to a level comparable with the controls.
AmygA, AI, and HEMA were concordantly enhanced in patients with AMI, showing concurrent dynamic changes over time. These results raise the possibility that stress-associated neurobiological activity is linked with acute plaque instability via augmented macrophage activity and could be a potential therapeutic target for plaque inflammation in AMI.
See page 1896 for the editorial comment on this article (doi: )
Introduction
Chronic psychosocial stress is a known major risk factor for cardiovascular disease (CVD), making a significant contribution along with conventional risk factors.1 Amygdala is a main component of the brain’s salience network, which regulates stress perception and emotional response.2 While amygdala involves in a generalized emotional response to various types of psychosocial stress, not only limited to a specific type of stress stimuli or response,3 , 4 contemporary evidence suggest that limbic-frontal circuitry including amygdala plays a crucial role in the pathophysiology of brain–heart interconnection.5 Resting amygdalar activity (AmygA) could be quantitatively measured with high reproducibility and reliability using 18F-fluorodedoxyglucose positron emission tomography/computed tomography (18F-FDG-PET/CT),6 , 7 enabling a simultaneous estimation of arterial inflammation (AI) and haematopoietic activity (HEMA).8 Increased AmygA as assessed by 18F-FDG-PET/CT has been recently reported as an independent predictor of future major adverse cardiovascular events (MACEs) mediated by enhanced AI and up-regulated HEMA.7
Acute myocardial infarction (AMI) is a catastrophic manifestation of acute plaque instability that accompanies simultaneous enhancement of AI and HEMA.9–12 Although amygdalar activation precedes future MACE development,7 whether AmygA is indeed elevated in the phase of persistent acute plaque instability along with AI and HEMA remains unexplored in human beings. Moreover, it is controversial whether the acute plaque instability resolves dynamically over time with AmygA. Given the previous evidence linking AmygA and subsequent CVD,7 we hypothesized that emotional AmygA mechanistically associates with acute coronary plaque instability by macrophage accumulation through augmented HEMA after AMI. To answer these questions, here, we prospectively assessed AmygA, AI, and HEMA using 18F-FDG-PET/CT molecular imaging in the patients with angiographically confirmed AMI and compared those PET metabolic activities with age- and sex-matched controls without obstructive coronary artery disease (CAD). Additionally, serial 18F-FDG-PET/CT imaging was performed in a subset of patients 6 months after the initial AMI diagnosis to examine the temporal changes of the relationship between macrophage activities and emotional neural status along with resolution of clinical disease activity of CAD from AMI to stable condition.
Methods
Study design and population
This prospective cohort study comprised two types of investigations. The first was a cross-sectional study comparing AmygA, AI, and HEMA between AMI and no-CAD controls. The second was a longitudinal study investigating the temporal changes in these metabolic activities in a subset of AMI patients.
The whole study population was prospectively enrolled from the cardiovascular centre at Korea University Guro Hospital (KUGH; Seoul, Republic of Korea) between July 2017 and November 2019. The study flow chart is displayed in Figure 1. Key inclusion and exclusion criteria of the AMI cohort and control group are provided in the Supplementary material online. A subset of patients (n = 10) was randomly selected from the AMI group and underwent serial PET/CT imaging after 6 months. Control subjects without significant CAD were enrolled from a pool of clinically stable patients undergoing elective coronary angiography or coronary CT for the diagnostic evaluation of anginal chest pain depending on the preliminary age and sex distribution of the AMI cohort (matched in a 3:1 ratio; AMI/control group). The study protocol complied with the Declaration of Helsinki and was approved by the Ethics Committee and Institutional Review Board at KUGH (2018GR0101), and written informed consent was obtained from all participants.
Flow chart of the study enrolment process. The present study consisted of two types of investigations: (A) Cross-sectional analysis of acute myocardial infarction and control groups; and (B) Longitudinal analysis within the acute myocardial infarction cohort from baseline to 6-month follow-up. CAG, coronary angiography.
Patient assessment and management
Baseline demographics and underlying cardiovascular risk factors were recorded through detailed patient interviews, and laboratory data were obtained at the time of the index admission and study enrolment. Routine neuropsychological examination was conducted by an expert neuropsychiatrist (C.K.K.) during the initial patient assessment to exclude overt mental and psychological disorders. Two types of well-validated psychological questionnaires, the Perceived Stress Scale-10 (PSS-10)13 and Patient Health Questionnaire-9 (PHQ-9),14 were completed from all participants at study enrolment. Psychosocial stress scales of PSS-10 and PHQ-9 were serially assessed after 6 months with follow-up PET/CT imaging. Mini mental status examination was conducted to examine the presence of clinically meaningful cognitive disorders.15 The AMI cohort was managed in accordance with the contemporary practice guidelines of AMI.16 , 17 Those who presented with ST-elevation AMI were treated with primary percutaneous coronary intervention (PCI) if not contraindicated, and an early invasive strategy in non-ST-elevation AMI was chosen at the attending cardiologist’s discretion. Dual antiplatelet therapy (DAPT) with aspirin and P2Y12 inhibitor was prescribed in all patients after PCI. The choice of specific type and dose of statin was left to the individual physician’s discretion. The AMI cohort and control subjects received guideline-directed medical therapy for the underlying medical comorbidities including hypertension, diabetes, and dyslipidaemia.
Image acquisition and analysis
Coronary angiography and 18F-fluorodeoxyglucose positron emission tomography/computed tomography imaging
Coronary angiography was performed in all patients in the AMI cohort. Baseline diagnostic angiograms were assessed to evaluate the complexity of culprit plaque morphology and total plaque burden of the coronary vasculature using the previously validated criteria,18 , 19 including thrombolysis in myocardial infarction (TIMI) grade and Synergy between PCI with Taxus and Cardiac Surgery (SYNTAX) score. Further details of angiographic assessment are provided in the Supplementary material online.
All patients underwent brain and vascular 18F-FDG-PET/CT imaging studies within 45 days of the index episode according to previously validated methods.7 , 20 Details of the 18F-FDG-PET/CT imaging protocol are provided in the Supplementary material online. Positron emission tomography data were routinely recalculated to provide images of standardized uptake values (SUV) based on lean body weight. Using a dedicated workstation (Extended Brilliance Workspace V4.0, Philips Medical Systems), which enables analysis of multimodal fusion images, comprehensive assessment was performed to quantify 18F-FDG uptakes.
Measurement of metabolic activities of target tissues by 18F-FDG-PET/CT imaging
AmygA,7 AI,8 , 20 and HEMA7 , 8 were quantified using previously validated methods. Details are provided in the Supplementary material online. In brief, 18F-FDG tracer uptake of each target tissue was recorded as maximum and mean standardized uptake values (SUVmax and SUVmean). AmygA was corrected for background cerebral activity to provide target-to-background ratio (TBR) by dividing the amygdalar SUVs by mean temporal lobe SUV.7 , 21 The primary measurement of AmygA was the highest of each amygdalar SUVmax divided by the temporal SUVmean (Amygdala max TBRmax). The secondary measurements of AmygA were the average of each amygdalar SUVmax and the highest of each amygdalar SUVmean divided by the temporal SUVmean (Amygdala average TBRmax and max TBRmean, respectively).
Carotid SUVmax was calculated by identifying the eight consecutive axial slices showing the highest average SUVmax from both carotid arteries. Aorta SUVmax was defined as the average ascending aortic wall SUVmax. Bone marrow SUVmax was defined as the average maximum SUV of all individual thoracic vertebra. Carotid TBRmax, aorta TBRmax, and bone marrow TBRmax were calculated by dividing the SUVmax values of each target tissue by the background venous activity in the superior vena cava.
Three-dimensional reconstruction rendering process
Details are provided in the Supplementary material online and Graphical abstract.
Statistical analysis
Data are expressed as mean and standard deviation (SD) or median and interquartile range (IQR) for continuous variables, and frequency (percentage) for categorical variables. Intergroup differences were analysed by Student’s t-test or Mann–Whitney U-test for continuous variables and by Pearson’s χ2 or Fisher’s exact test for categorical variables. Temporal changes in 18F-FDG signal uptake at 6 months after the index AMI were analysed by paired t-test. The effect size was calculated using Cohen's d to evaluate the magnitude of the difference. Correlations between two relevant variables were examined by Pearson’s correlation coefficient. Multivariate linear regression was used to test association of AmygA and clinical variables. Mediation analysis was performed to estimate the direct and indirect effects of the hypothesized single-mediator pathway.22 The PROCESS macro algorithm employing an ordinary least square-based framework was used and age, sex, and study group were entered as covariates in the analysis. All analyses were two-tailed, and significant differences were defined as P-values <0.05. Sample size calculation determined that a total of 64–68 patients (matched in a 3:1 ratio = 51 patients with MI and 17 control subjects; matched in a 2.5:1 ratio = 46 patients with MI and 18 control subjects) would yield 80% power to detect an effect size of 0.80 at a significance level of 5% in the analysis of the primary measurement. The statistical analysis was performed using Statistical Package for the Social Sciences software version 20.0 (SPSS-PC Inc., Chicago, IL, USA).
Results
Baseline characteristics and angiographic findings
A total of 62 patients (45 angiographically confirmed AMI patients and 17 no-CAD controls) were prospectively enrolled. Baseline characteristics and prescribed medications are shown in Table 1. Both groups were well balanced, showing no significant differences in sex, age, or medical comorbidities at baseline. The AMI group showed significantly worse values for serum inflammatory marker, cardiac enzymes, and left ventricular ejection fraction; however, other laboratory findings did not show any significant intergroup difference at baseline. All patients in the AMI group received DAPT with aspirin and clopidogrel, and the statin prescription rate was similar between groups. The AMI group had more depressive mood as assessed by the PHQ-9, while the levels of perceived stress were similar on the PSS-10. The level of brain atrophy and cognitive impairment at baseline were similar in both groups (Table 1 and Supplementary material online, Table S1). Coronary angiographic findings of the AMI group are displayed in Supplementary material online, Table S2. Of the 45 AMI patients, 34 (75.6%) had complex plaque characteristics at the culprit lesion. Nearly half (44.4%) of the AMI patients had multi-vessel disease. The average SYNTAX score was 16.3 ± 8.4, and 26.7% (n = 12) of the AMI patients had intermediate to high (≥22) SYNTAX scores. Baseline characteristics and angiographic findings of the subgroup who underwent follow-up PET/CT imaging were mostly comparable to the remaining patients of the AMI cohort (Supplementary material online, Table S3).
Baseline characteristics and prescribed medications
| AMI (n = 45) | Control (n = 17) | P-value | |
|---|---|---|---|
| Sex (male) | 38 (84.4) | 13 (76.4) | 0.475 |
| Age (years) | 60.0 ± 9.6 | 59.6 ± 13.3 | 0.908 |
| BMI (km/m2) | 24.4 ± 2.8 | 24.0 ± 2.1 | 0.624 |
| Interval to PET/CT (days) | 21.7 ± 15.7 | 19.2 ± 12.2 | 0.551 |
| Medical history | |||
| Hypertension | 20 (44.4) | 7 (41.1) | 0.817 |
| Diabetes | 7 (15.5) | 5 (29.4) | 0.218 |
| Dyslipidaemia | 17 (37.7) | 8 (47.0) | 0.506 |
| Prior CAD | 2 (4.4) | 0 (0.0) | >0.999 |
| Smoking | 23 (51.1) | 9 (52.9) | 0.898 |
| Alcohol | 24 (53.3) | 9 (52.9) | 0.978 |
| PHQ-9 | 7.9 ± 5.0 | 4.4 ± 3.3 | 0.002 |
| PSS-10 | 16.3 ± 5.0 | 14.5 ± 3.8 | 0.182 |
| MMSE | 29.3 ± 1.3 | 28.8 ± 1.9 | 0.206 |
| Laboratory findings | |||
| Creatinine (μmol/L) | 80.4 ± 25.6 | 75.1 ± 17.7 | 0.419 |
| CrCl (mL/min) | 88.5 ± 29.5 | 90.2 ± 26.4 | 0.840 |
| HbA1c (%) | 5.8 ± 0.7 | 5.7 ± 0.6 | 0.571 |
| TC (mmol/L) | 4.84 ± 1.01 | 4.85 ± 1.11 | 0.964 |
| LDLc (mmol/L) | 2.93 ± 0.94 | 2.98 ± 1.09 | 0.859 |
| hs-CRP (mg/L) | 1.43 (0.76–3.51) | 0.65 (0.38–1.54) | 0.007 |
| Max CK-MB (ng/mL) | 219.2 (68.7–300.0) | 1.3 (0.8–1.8) | <0.001 |
| Max Troponin-T (ng/mL) | 1.53 (0.23–4.14) | 0.02 (0.01–0.03) | <0.001 |
| NT-proBNP (pg/mL) | 87.0 (37.4–398.4) | 51.7 (26.1–69.3) | 0.099 |
| LVEF (%) | 51.7 ± 6.7 | 60.3 ± 1.2 | <0.001 |
| Prescribed medications | |||
| Aspirin | 45 (100.0) | 0 (0.0) | <0.001 |
| Clopidogrel | 45 (100.0) | 2 (11.7) | <0.001 |
| Beta-blockers | 30 (66.6) | 0 (0.0) | <0.001 |
| ACE-inhibitor or ARBs | 37 (82.2) | 6 (35.2) | <0.001 |
| CCBs | 6 (13.3) | 6 (35.2) | 0.051 |
| Statins | 45 (100.0) | 16 (94.1) | 0.274 |
| Low intensity | 0 (0.0) | 5 (29.4) | |
| Moderate-high intensity | 40 (88.8) | 11 (64.7) | |
| High intensity | 5 (11.1) | 0 (0.0) |
| AMI (n = 45) | Control (n = 17) | P-value | |
|---|---|---|---|
| Sex (male) | 38 (84.4) | 13 (76.4) | 0.475 |
| Age (years) | 60.0 ± 9.6 | 59.6 ± 13.3 | 0.908 |
| BMI (km/m2) | 24.4 ± 2.8 | 24.0 ± 2.1 | 0.624 |
| Interval to PET/CT (days) | 21.7 ± 15.7 | 19.2 ± 12.2 | 0.551 |
| Medical history | |||
| Hypertension | 20 (44.4) | 7 (41.1) | 0.817 |
| Diabetes | 7 (15.5) | 5 (29.4) | 0.218 |
| Dyslipidaemia | 17 (37.7) | 8 (47.0) | 0.506 |
| Prior CAD | 2 (4.4) | 0 (0.0) | >0.999 |
| Smoking | 23 (51.1) | 9 (52.9) | 0.898 |
| Alcohol | 24 (53.3) | 9 (52.9) | 0.978 |
| PHQ-9 | 7.9 ± 5.0 | 4.4 ± 3.3 | 0.002 |
| PSS-10 | 16.3 ± 5.0 | 14.5 ± 3.8 | 0.182 |
| MMSE | 29.3 ± 1.3 | 28.8 ± 1.9 | 0.206 |
| Laboratory findings | |||
| Creatinine (μmol/L) | 80.4 ± 25.6 | 75.1 ± 17.7 | 0.419 |
| CrCl (mL/min) | 88.5 ± 29.5 | 90.2 ± 26.4 | 0.840 |
| HbA1c (%) | 5.8 ± 0.7 | 5.7 ± 0.6 | 0.571 |
| TC (mmol/L) | 4.84 ± 1.01 | 4.85 ± 1.11 | 0.964 |
| LDLc (mmol/L) | 2.93 ± 0.94 | 2.98 ± 1.09 | 0.859 |
| hs-CRP (mg/L) | 1.43 (0.76–3.51) | 0.65 (0.38–1.54) | 0.007 |
| Max CK-MB (ng/mL) | 219.2 (68.7–300.0) | 1.3 (0.8–1.8) | <0.001 |
| Max Troponin-T (ng/mL) | 1.53 (0.23–4.14) | 0.02 (0.01–0.03) | <0.001 |
| NT-proBNP (pg/mL) | 87.0 (37.4–398.4) | 51.7 (26.1–69.3) | 0.099 |
| LVEF (%) | 51.7 ± 6.7 | 60.3 ± 1.2 | <0.001 |
| Prescribed medications | |||
| Aspirin | 45 (100.0) | 0 (0.0) | <0.001 |
| Clopidogrel | 45 (100.0) | 2 (11.7) | <0.001 |
| Beta-blockers | 30 (66.6) | 0 (0.0) | <0.001 |
| ACE-inhibitor or ARBs | 37 (82.2) | 6 (35.2) | <0.001 |
| CCBs | 6 (13.3) | 6 (35.2) | 0.051 |
| Statins | 45 (100.0) | 16 (94.1) | 0.274 |
| Low intensity | 0 (0.0) | 5 (29.4) | |
| Moderate-high intensity | 40 (88.8) | 11 (64.7) | |
| High intensity | 5 (11.1) | 0 (0.0) |
Data are expressed as n (%), mean ± standard deviation or median (interquartile range). ACE, angiotensin-converting enzyme; ARB, angiotensin receptor blocker; CCB, calcium channel blocker; CK-MB, creatinine kinase-MB; CrCl, creatinine clearance; HbA1c, glycated haemoglobin; hs-CRP, high-sensitivity C-reactive protein; LDLc, low-density lipoprotein cholesterol; LVEF, left ventricular ejection fraction; MMSE, mini mental status examination; NT-proBNP, N-terminal pro-B-type natriuretic peptide; TC, total cholesterol.
Baseline characteristics and prescribed medications
| AMI (n = 45) | Control (n = 17) | P-value | |
|---|---|---|---|
| Sex (male) | 38 (84.4) | 13 (76.4) | 0.475 |
| Age (years) | 60.0 ± 9.6 | 59.6 ± 13.3 | 0.908 |
| BMI (km/m2) | 24.4 ± 2.8 | 24.0 ± 2.1 | 0.624 |
| Interval to PET/CT (days) | 21.7 ± 15.7 | 19.2 ± 12.2 | 0.551 |
| Medical history | |||
| Hypertension | 20 (44.4) | 7 (41.1) | 0.817 |
| Diabetes | 7 (15.5) | 5 (29.4) | 0.218 |
| Dyslipidaemia | 17 (37.7) | 8 (47.0) | 0.506 |
| Prior CAD | 2 (4.4) | 0 (0.0) | >0.999 |
| Smoking | 23 (51.1) | 9 (52.9) | 0.898 |
| Alcohol | 24 (53.3) | 9 (52.9) | 0.978 |
| PHQ-9 | 7.9 ± 5.0 | 4.4 ± 3.3 | 0.002 |
| PSS-10 | 16.3 ± 5.0 | 14.5 ± 3.8 | 0.182 |
| MMSE | 29.3 ± 1.3 | 28.8 ± 1.9 | 0.206 |
| Laboratory findings | |||
| Creatinine (μmol/L) | 80.4 ± 25.6 | 75.1 ± 17.7 | 0.419 |
| CrCl (mL/min) | 88.5 ± 29.5 | 90.2 ± 26.4 | 0.840 |
| HbA1c (%) | 5.8 ± 0.7 | 5.7 ± 0.6 | 0.571 |
| TC (mmol/L) | 4.84 ± 1.01 | 4.85 ± 1.11 | 0.964 |
| LDLc (mmol/L) | 2.93 ± 0.94 | 2.98 ± 1.09 | 0.859 |
| hs-CRP (mg/L) | 1.43 (0.76–3.51) | 0.65 (0.38–1.54) | 0.007 |
| Max CK-MB (ng/mL) | 219.2 (68.7–300.0) | 1.3 (0.8–1.8) | <0.001 |
| Max Troponin-T (ng/mL) | 1.53 (0.23–4.14) | 0.02 (0.01–0.03) | <0.001 |
| NT-proBNP (pg/mL) | 87.0 (37.4–398.4) | 51.7 (26.1–69.3) | 0.099 |
| LVEF (%) | 51.7 ± 6.7 | 60.3 ± 1.2 | <0.001 |
| Prescribed medications | |||
| Aspirin | 45 (100.0) | 0 (0.0) | <0.001 |
| Clopidogrel | 45 (100.0) | 2 (11.7) | <0.001 |
| Beta-blockers | 30 (66.6) | 0 (0.0) | <0.001 |
| ACE-inhibitor or ARBs | 37 (82.2) | 6 (35.2) | <0.001 |
| CCBs | 6 (13.3) | 6 (35.2) | 0.051 |
| Statins | 45 (100.0) | 16 (94.1) | 0.274 |
| Low intensity | 0 (0.0) | 5 (29.4) | |
| Moderate-high intensity | 40 (88.8) | 11 (64.7) | |
| High intensity | 5 (11.1) | 0 (0.0) |
| AMI (n = 45) | Control (n = 17) | P-value | |
|---|---|---|---|
| Sex (male) | 38 (84.4) | 13 (76.4) | 0.475 |
| Age (years) | 60.0 ± 9.6 | 59.6 ± 13.3 | 0.908 |
| BMI (km/m2) | 24.4 ± 2.8 | 24.0 ± 2.1 | 0.624 |
| Interval to PET/CT (days) | 21.7 ± 15.7 | 19.2 ± 12.2 | 0.551 |
| Medical history | |||
| Hypertension | 20 (44.4) | 7 (41.1) | 0.817 |
| Diabetes | 7 (15.5) | 5 (29.4) | 0.218 |
| Dyslipidaemia | 17 (37.7) | 8 (47.0) | 0.506 |
| Prior CAD | 2 (4.4) | 0 (0.0) | >0.999 |
| Smoking | 23 (51.1) | 9 (52.9) | 0.898 |
| Alcohol | 24 (53.3) | 9 (52.9) | 0.978 |
| PHQ-9 | 7.9 ± 5.0 | 4.4 ± 3.3 | 0.002 |
| PSS-10 | 16.3 ± 5.0 | 14.5 ± 3.8 | 0.182 |
| MMSE | 29.3 ± 1.3 | 28.8 ± 1.9 | 0.206 |
| Laboratory findings | |||
| Creatinine (μmol/L) | 80.4 ± 25.6 | 75.1 ± 17.7 | 0.419 |
| CrCl (mL/min) | 88.5 ± 29.5 | 90.2 ± 26.4 | 0.840 |
| HbA1c (%) | 5.8 ± 0.7 | 5.7 ± 0.6 | 0.571 |
| TC (mmol/L) | 4.84 ± 1.01 | 4.85 ± 1.11 | 0.964 |
| LDLc (mmol/L) | 2.93 ± 0.94 | 2.98 ± 1.09 | 0.859 |
| hs-CRP (mg/L) | 1.43 (0.76–3.51) | 0.65 (0.38–1.54) | 0.007 |
| Max CK-MB (ng/mL) | 219.2 (68.7–300.0) | 1.3 (0.8–1.8) | <0.001 |
| Max Troponin-T (ng/mL) | 1.53 (0.23–4.14) | 0.02 (0.01–0.03) | <0.001 |
| NT-proBNP (pg/mL) | 87.0 (37.4–398.4) | 51.7 (26.1–69.3) | 0.099 |
| LVEF (%) | 51.7 ± 6.7 | 60.3 ± 1.2 | <0.001 |
| Prescribed medications | |||
| Aspirin | 45 (100.0) | 0 (0.0) | <0.001 |
| Clopidogrel | 45 (100.0) | 2 (11.7) | <0.001 |
| Beta-blockers | 30 (66.6) | 0 (0.0) | <0.001 |
| ACE-inhibitor or ARBs | 37 (82.2) | 6 (35.2) | <0.001 |
| CCBs | 6 (13.3) | 6 (35.2) | 0.051 |
| Statins | 45 (100.0) | 16 (94.1) | 0.274 |
| Low intensity | 0 (0.0) | 5 (29.4) | |
| Moderate-high intensity | 40 (88.8) | 11 (64.7) | |
| High intensity | 5 (11.1) | 0 (0.0) |
Data are expressed as n (%), mean ± standard deviation or median (interquartile range). ACE, angiotensin-converting enzyme; ARB, angiotensin receptor blocker; CCB, calcium channel blocker; CK-MB, creatinine kinase-MB; CrCl, creatinine clearance; HbA1c, glycated haemoglobin; hs-CRP, high-sensitivity C-reactive protein; LDLc, low-density lipoprotein cholesterol; LVEF, left ventricular ejection fraction; MMSE, mini mental status examination; NT-proBNP, N-terminal pro-B-type natriuretic peptide; TC, total cholesterol.
Comparison between acute myocardial infarction and no-coronary artery disease controls
In the cross-sectional study, the average duration from the date of the index episode to PET/CT imaging was 21.0 ± 14.8 days. Graphical abstract and Supplementary material online, Video S1 shows the customized three-dimensional (3D)-rendered reconstruction images enabling comprehensive visualization of the active signals from the target areas of interest such as amygdada, vasculatures, and bone marrow. Figures 2 and 3 show a quantitative comparison of conventional PET/CT images that highlights the characteristic 18F-FDG uptakes at the target areas between different clinical settings. AmygA (max TBRmax: 0.65 ± 0.05 vs. 0.60 ± 0.05; P = 0.004; Figure 3A), carotid AI (TBRmax 2.04 ± 0.39 vs. 1.81 ± 0.25; P = 0.026; Figure 3B), and HEMA (TBRmax 2.60 ± 0.38 vs. 2.22 ± 0.28; P < 0.001; Figure 3D) were significantly higher in AMI patients compared with no-CAD controls. Aortic AI (TBRmax 2.14 ± 0.34 vs. 1.99 ± 0.24; P = 0.115; Figure 3C) was also higher in the AMI group, though the difference was not statistically significant. Detailed information of each metabolic activity is displayed in Supplementary material online, Table S4. Consistently, the secondary measurements of AmygA (average TBRmax and max TBRmean) showed a similar pattern to those of the primary measurement. Although individual SUVs were similar between groups, each TBR corrected by background signal activity showed significant difference (Supplementary material online, Table S4). The 18F-FDG uptakes at the control brain region involved in the limbic-frontal circuitry showed no significant difference between the groups (Supplementary material online, Table S4). A clear demonstration of concurrently enhanced AmygA, AI, and HEMA at the early stage of AMI suggests a possible association of the neural activities and acute plaque instability through enhanced macrophage haematopoiesis. Furthermore, time-dependent dynamic changes of these metabolic activities provide evidence to support this biological interconnection.
Conventional positron emission tomography/computed tomography images presenting representative 18F-FDG uptake patterns of the amygdala, carotid artery, aorta, and bone marrow. Axial view of amygdala (first line) and carotid artery (second line), coronal view of aorta (third line), and sagittal view of thoracic spinal bone marrow (fourth line) are displayed. The solid and dashed circular region indicates the amygdala and carotid artery, respectively. The arrow indicates the aortic wall. The white scale bar denotes 5 cm.
Cross-sectional and longitudinal comparison of signal activities at the amygdala, carotid artery, aorta, and bone marrow in acute myocardial infarction. Panels (A–D) compare the signals of the target tissues between acute myocardial infarction patients and controls. Panels (E–H) show temporal changes in these metabolic activities 6 months after the acute myocardial infarction episode.
Temporal changes in metabolic activities at 6 months after the index acute myocardial infarction
In the longitudinal study, a serial PET/CT assessment was performed in a subset of 10 AMI patients, in whom the average interval between the baseline and follow-up imaging was 180.5 ± 5.0 days. All patients were clinically stable without experiencing any significant MACEs during follow-up. Psychosocial scales of PHQ-9 and PSS-10 numerically decreased over time; however, the difference did not show statistical significance. Intriguingly, AmygA (max TBRmax: baseline 0.66 ± 0.05 vs. follow-up 0.59 ± 0.04; P = 0.002; Figure 3E), carotid AI (TBRmax: baseline 2.43 ± 0.48 vs. follow-up 2.07 ± 0.33; P = 0.010; Figure 3F), and HEMA (TBRmax: baseline 2.64 ± 0.38 vs. follow-up 2.35 ± 0.21; P = 0.034; Figure 3H) decreased significantly 6 months after the index AMI episodes. Although not statistically significant, Aortic AI (TBRmax: baseline 2.33 ± 0.42 vs. follow-up 2.20 ± 0.21; P = 0.349; Figure 3G) also decreased at 6 months following AMI. Statistical serial comparison of the Aortic AI may have been underpowered due to a small sample size. Supplementary material online, Table S5 shows detailed information of the temporal changes in metabolic activities and psychological scales of AMI patients.
Association of positron emission tomography metabolic activities with psychological scales and laboratory measurements
AmygA correlated significantly with carotid AI (r = 0.350; P = 0.005; Figure 4A), aortic AI (r = 0.471; P < 0.001; Figure 4B), and HEMA (r = 0.356; P = 0.005; Figure 4C) in the overall subjects. Psychological measurements of PHQ-9 (r = 0.598; P < 0.001; Figure 4D) and PSS-10 (r = 0.507; P < 0.001; Figure 4E) also correlated well with AmygA assessed by PET/CT imaging. Both psychological measurements of PHQ-9 (P < 0.001) and PSS-10 (P = 0.001) remained as an independent predictor of AmygA after adjustment of potential confounders (Supplementary material online, Table S6). Correlation between the PET metabolic parameters and laboratory biomarkers are displayed in Supplementary material online, Table S7. High-sensitivity C-reactive protein at baseline did not show any significant correlation with AmygA (r = 0.064, P = 0.619). AmygA correlated positively with max creatinine kinase-MB (CK-MB) (r = 0.331, P = 0.008), max troponin-T (r = 0.258, P = 0.043), and N-terminal pro-B-type natriuretic peptide (NT-proBNP) (r = 0.368, P = 0.003). Haematopoiesis correlated positively with max CK-MB (r = 0.263, P = 0.039) and negatively with left ventricular ejection fraction (r = −0.346, P = 0.006). These findings suggest that PET metabolic activities of amygdala and bone marrow associates with the extent of myocardial damage. After adjustment of potential confounders, NT-proBNP (P = 0.016) and max CK-MB (P = 0.003) remained as an independent predictor of AmygA (Supplementary material online, Table S8). Further details about the association between PET metabolic activities and laboratory measurements are provided in Supplementary material online, Tables S6–S12. Mediation analysis of the hypothesized pathway from AmygA to AI demonstrated HEMA as a significant mediator for this relationship (standardized β 0.099, 95% CI 0.007–0.220, P = 0.014), accounting for 22% of the total effect (Figure 5). Mediation effect of HEMA was more evident in the pathway ending with Aortic AI compared with that ending with Carotid AI (Supplementary material online, Figure S1). Further details about the mediation analysis results are provided in Supplementary material online.
Correlation between amygdalar activity, arterial inflammation, haematopoiesis, and psychosocial stress. Panel (A) shows a significant correlation between amygdalar activity, arterial inflammation, and haematopoiesis. Panel (B) shows a significant correlation between amygdalar activity and psychological parameters.
Hypothesized pathway linking amygdalar activity to arterial inflammation. A single-mediator analysis demonstrated that haematopoietic activity was a significant mediator of the association between amygdalar activity and arterial inflammation (accounting for 22% of the total effect). All metabolic activities are measured as SUVmax of the target tissues of interest. Arterial inflammation was defined as the positron emission tomography signal uptakes in the aortic wall. Analysis included age, sex, and study groups as covariates. Standardized regression coefficients (standardized β) are displayed. c, total effect of amygdalar activity on arterial inflammation; cʹ, residual direct effect independent of mediation pathway; cʹʹ, indirect effect mediated by haematopoietic activity.
Angiographic sub-analysis within the acute myocardial infarction group
AmygA was significantly higher in those who represented complex coronary plaque at the culprit lesion (max TBRmax: 0.66 ± 0.05 vs. 0.60 ± 0.04; P < 0.001; Figure 6A), higher SYNTAX score (≥22; max TBRmax: 0.68 ± 0.02 vs. 0.63 ± 0.06; P = 0.004; Figure 6B), and significant flow limitation (TIMI <3; max TBRmax: 0.66 ± 0.05 vs. 0.61 ± 0.06; P = 0.008; Figure 6C). These findings suggest that AmygA closely associates with loco-regional disease activities at the level of coronary vasculature. In contrast, carotid AI, aortic AI, and HEMA merely showed significant differences according to the presence of the aforementioned complex angiographic characteristics (Supplementary material online, Figure S2). Detailed information of the sub-analysis results based on angiographic findings is provided in Supplementary material online, Table S13. Haematopoiesis was unexpectedly higher in patients without significant flow limitation (TIMI 3) and this was suggested to be influenced by longer pre-hospital ischaemic time (5.42 ± 2.75 vs. 2.79 ± 1.75 h; P < 0.001; Supplementary material online, Table S14). Although the severity of AMI could potentially affect AmygA and culprit lesion morphology, angiographic culprit lesion complexity still remained as an independent predictor of AmygA after adjusting for the extent of myocardial injury (Table 2).
Level of amygdalar activation according to the presence of angiographic complex lesion characteristics. Amygdalar activity was significantly higher in acute myocardial infarction patients presenting angiographic complex culprit plaque morphology (A), intermediate to high SYNTAX score (≥22) (B), and significant flow limitation (TIMI 0-2) (C). SYNTAX, Synergy between PCI with Taxus and Cardiac Surgery; TIMI, thrombolysis in myocardial infarction.
Angiographic sub-analysis: prediction of amygdalar activity by complex lesion characteristics
| B | SE | Standardized β | t | P-value | |
|---|---|---|---|---|---|
| Lesion complexity | |||||
| Complex lesion (unadjusted) | 0.063 | 0.016 | 0.508 | 3.865 | <0.001 |
| Complex lesion (adjusteda) | 0.060 | 0.017 | 0.488 | 3.485 | 0.001 |
| SYNTAX score | |||||
| SYNTAX ≥22 (unadjusted) | 0.051 | 0.017 | 0.424 | 3.074 | 0.004 |
| SYNTAX ≥22 (adjusteda) | 0.056 | 0.017 | 0.464 | 3.261 | 0.002 |
| Flow limitation | |||||
| TIMI 0-2 (unadjusted) | 0.047 | 0.017 | 0.390 | 2.780 | 0.008 |
| TIMI 0-2 (adjusteda) | 0.040 | 0.019 | 0.334 | 2.138 | 0.039 |
| B | SE | Standardized β | t | P-value | |
|---|---|---|---|---|---|
| Lesion complexity | |||||
| Complex lesion (unadjusted) | 0.063 | 0.016 | 0.508 | 3.865 | <0.001 |
| Complex lesion (adjusteda) | 0.060 | 0.017 | 0.488 | 3.485 | 0.001 |
| SYNTAX score | |||||
| SYNTAX ≥22 (unadjusted) | 0.051 | 0.017 | 0.424 | 3.074 | 0.004 |
| SYNTAX ≥22 (adjusteda) | 0.056 | 0.017 | 0.464 | 3.261 | 0.002 |
| Flow limitation | |||||
| TIMI 0-2 (unadjusted) | 0.047 | 0.017 | 0.390 | 2.780 | 0.008 |
| TIMI 0-2 (adjusteda) | 0.040 | 0.019 | 0.334 | 2.138 | 0.039 |
Adjusted for sex, age, Max CK-MB, Max troponin-T, NT-proBNP. Amygdalar activity was defined using Amygdala max TBRmax. Culprit lesion was considered complex if there were two or more of following angiographic morphologic features: thrombus, ulceration, plaque irregularity, and impaired flow. CK-MB, creatinine kinase-MB; Max, maximum; NT-proBNP, N-terminal pro-B-type natriuretic peptide; SYNTAX, Synergy between PCI with Taxus and Cardiac Surgery; TIMI, thrombolysis in myocardial infarction.
Angiographic sub-analysis: prediction of amygdalar activity by complex lesion characteristics
| B | SE | Standardized β | t | P-value | |
|---|---|---|---|---|---|
| Lesion complexity | |||||
| Complex lesion (unadjusted) | 0.063 | 0.016 | 0.508 | 3.865 | <0.001 |
| Complex lesion (adjusteda) | 0.060 | 0.017 | 0.488 | 3.485 | 0.001 |
| SYNTAX score | |||||
| SYNTAX ≥22 (unadjusted) | 0.051 | 0.017 | 0.424 | 3.074 | 0.004 |
| SYNTAX ≥22 (adjusteda) | 0.056 | 0.017 | 0.464 | 3.261 | 0.002 |
| Flow limitation | |||||
| TIMI 0-2 (unadjusted) | 0.047 | 0.017 | 0.390 | 2.780 | 0.008 |
| TIMI 0-2 (adjusteda) | 0.040 | 0.019 | 0.334 | 2.138 | 0.039 |
| B | SE | Standardized β | t | P-value | |
|---|---|---|---|---|---|
| Lesion complexity | |||||
| Complex lesion (unadjusted) | 0.063 | 0.016 | 0.508 | 3.865 | <0.001 |
| Complex lesion (adjusteda) | 0.060 | 0.017 | 0.488 | 3.485 | 0.001 |
| SYNTAX score | |||||
| SYNTAX ≥22 (unadjusted) | 0.051 | 0.017 | 0.424 | 3.074 | 0.004 |
| SYNTAX ≥22 (adjusteda) | 0.056 | 0.017 | 0.464 | 3.261 | 0.002 |
| Flow limitation | |||||
| TIMI 0-2 (unadjusted) | 0.047 | 0.017 | 0.390 | 2.780 | 0.008 |
| TIMI 0-2 (adjusteda) | 0.040 | 0.019 | 0.334 | 2.138 | 0.039 |
Adjusted for sex, age, Max CK-MB, Max troponin-T, NT-proBNP. Amygdalar activity was defined using Amygdala max TBRmax. Culprit lesion was considered complex if there were two or more of following angiographic morphologic features: thrombus, ulceration, plaque irregularity, and impaired flow. CK-MB, creatinine kinase-MB; Max, maximum; NT-proBNP, N-terminal pro-B-type natriuretic peptide; SYNTAX, Synergy between PCI with Taxus and Cardiac Surgery; TIMI, thrombolysis in myocardial infarction.
Discussion
To the best of our knowledge, this is the first prospective comparison to demonstrate a mechanistic link between brain emotional neural activity, arterial atherosclerotic inflammation, and macrophage haematopoiesis as assessed by 18F-FDG-PET/CT imaging in subjects with manifest CAD particularly during the phase of persistent acute plaque instability. The present study revealed several important findings. The brain emotional amygdalar activity increased concomitantly with AI via augmented haematopoietic macrophage activity in the patients with AMI. The brain AmygA correlated well with carotid and aortic AI, HEMA, and also with psychological measurements of chronic emotional stress and current depressive symptoms. Intriguingly, AmygA, AI, and HEMA showed concurrent dynamic changes over time with resolution of acute plaque instability from AMI to chronic coronary syndrome, featuring a decrease of the PET signals 6 months after the index AMI episode. Finally, AmygA was markedly enhanced particularly in AMI patients with complex angiographic features representing higher-risk plaques with a greater total atheroma burden. Taken all together, these findings suggest that brain emotional neural activity is mechanistically linked with acute plaque instability via macrophage haematopoiesis after AMI. Macrophage haematopoiesis was considered to play a critical role by mediating the association between brain emotional neural activity and AI. Considering the time-dependent dynamic changes in plaque inflammation depending on the emotional brain activity, AmygA is considered as a possible modifiable factor with consequent stabilization of the inflamed high-risk plaques. Our customized 3D volumetric rendered PET/CT imaging is expected to provide more comprehensive assessment to support the present findings and deepened our insights into the brain–haematopoiesis–vascular axis in the field of CVD.23 , 24
Earlier landmark studies demonstrated the association between AmygA and future CVD events in clinically stable circumstances from retrospective cohort studies.7 , 25 One prospective study was conducted in chronic inflammatory disease from a cohort of psoriasis patients.26 In current prospective comparative study, enhanced brain emotional neural activity with heightened macrophage activities of both vascular beds and bone marrow in AMI patients provides a robust clinical evidence to support the brain–haematopoiesis–vascular axis concept in the acute plaque instability,9 , 23 , 24 and expands the recently proposed connections between AmygA, AI, and HEMA7 to AMI-related situations. Furthermore, the present data showing dynamic changes of the PET signals could advance our knowledge about the association of AmygA with the disease activity of atherosclerotic CVD.
Previous preclinical studies postulated the presence of multi-organ network linking brain neural activity, macrophage haematopoiesis, and AI in a murine model of AMI.9 , 10 Enhanced sympathetic nervous signalling after AMI accelerated haematopoietic stem cell activities in the bone marrow niche, which further promoted the recruitment of circulating monocytes into the atherosclerotic plaque to increase vulnerability.9 , 10 Current study suggests further exploration of this multi-organ network in AMI subjects for the brain emotional activity assessed by 18F-FDG-PET/CT imaging. Graphical abstract shows a schematic conceptualization describing the interconnections between up-regulation of brain emotional neural activity vs. AI and macrophage haematopoiesis in human subjects of AMI, and its time-dependent dynamic course followed by a successful resolution of acute plaque instability.
AmygA was reported to be stable over time under a steady clinical environment.6 On the other hand, notably, our AmygA, AI, and HEMA changed over time along with the resolution of unstable acute plaque conditions, reflecting a concurrent decrease of these activities. Stabilized atherosclerotic disease activity and improved current emotional status over time were suggested as key contributing factor for the dynamic change of AmygA. Indeed, persisting plaque instability after AMI was related to higher emotional stress and enhanced representative neurobiological activity, AmygA, and that alleviation of emotional stress went along with decreasing HEMA and, mediated by this decrease, also decreased AI. Increased AmygA is evidenced as a predisposing factor for future CVD events along with social environmental factors.25 Amygdala involves in a generalized emotional response to various psychosocial stress,4 therefore, active modulation of this brain emotional stress response could potentially provide cardiovascular benefits. Increased AmygA at baseline and its dynamic change over time closely associates with more current measures of emotional status, therefore, emphasizes a relative importance of modulating the current emotional distress after AMI. Indeed, the benefit of routine screening on psychosocial stress after AMI is controversial,27 , 28 but active therapeutic intervention on emotional stress and depressive mood in AMI survivors have shown potential benefits on long-term mortality.29–31 Considering the potential crosstalk between AmygA and acute plaque instability, deliberate alleviation of current emotional distress affecting AmygA via physio-psychological intervention could add a potential therapeutic implication on the top of the standard secondary preventive measures, providing a more effective transition from high-risk plaque to more stable one.
In the angiographic assessment, current AMI patients with complex lesion morphology showing a higher SYNTAX score exhibited a stronger brain emotional AmygA. As the angiographic complex morphology of the plaques with a higher SYNTAX score represents a culprit plaque instability,19 , 32 the association of AmygA with acute plaque instability could be highlighted in the loco-regional level of the coronary vasculature as well as systemic disease activity. Interestingly, the angiographic characteristics of unstable culprit plaque were more clearly delineated by AmygA rather than AI and HEMA. Furthermore, while more severe AMI could potentially result in greater AmygA via increased emotional stress, angiographic culprit plaque complexity was independently associated with AmygA even after adjusting for the extent of myocardial injury. Future studies with intravascular integrative imaging techniques to delineate microstructural plaque instability are required to elucidate whether AmygA modification could influence the culprit plaque instability, or vice versa, in coronary arteries.33–35
To overcome the inherent limitation of single-slice axial imaging, we performed a customized 3D-rendered reconstruction from PET/CT imaging. This new technology clearly displayed the distribution of AmygA encompassing the whole range of the active signals, which cannot be represented on the single-slice conventional images (Figure 2), and firmly supported the key findings of the present study (Graphical abstract).
Our study has several limitations. The present study was inconclusive regarding a direct causal relationship between the assessed metabolic activities, even though it is challenging to interrogate causality using clinical serial PET/CT imaging in a cohort of AMI patients. While mediation analysis in current study provided a possible causal inference, interconnection between the PET metabolic activities could be interpreted in the viewpoint of bi-directional association. Despite of these intrinsic limitations, the biological interconnection in our study could expand the understanding of the integrative biology connecting neurobiological activity with atherosclerotic inflammation after AMI. Furthermore, our prospective longitudinal analysis supports the robust temporal linkage between multi-organ systems including brain neural activity following AMI. As the present study was conducted in a single centre with a relatively small sample size, large-scaled prospective longitudinal studies enrolling at-risk subjects of CVD are needed to establish the causal relationship. The well-designed preclinical studies are also expected to add detailed mechanistic insights into the underlying pathophysiology of this multi-organ linkage. Based on the potential sexual difference in stress susceptibility after AMI,36 future studies with balanced sexual distribution are needed to assess sex-specific differences in AmygA since the present study population consisted mostly of men (78.4%). The PET imaging co-registered with brain CT has limited spatial resolution for detailed volumetric assessment. Future studies should incorporate brain magnetic resonance imaging as standard brain imaging, enabling more sophisticated metabolic assessment of brain territories. Despite the robust association between AmygA and angiographic high-risk plaque characteristics, the plaque characterization solely based on angiographic assessment should be considered as exploratory, and thus, future studies using comprehensive intravascular imaging are required. The drawback of 18F-FDG-PET/CT imaging is limited specificity to target macrophage activity of the plaque, particularly in the coronary beds. Novel PET imaging agents are expected to overcome this technical difficulty in the future. The time-interval between AMI and PET/CT imaging was relatively longer in our study. While average 3-week time-interval fairly avoids various stressful factors of the peri-procedural period to influence the PET signal activities, optimal imaging time needs to be further determined.
Nonetheless, the several findings of our study outweigh these limitations by providing novel insight into the biological mechanism linking AmygA, AI, and HEMA with acute plaque instability in human beings. Future studies should be required to further elucidate the causal interconnection and prognostic impacts regarding whether modulation of brain neural activity has the potential to lead to a significant resolution of plaque inflammation affecting CVD outcomes. In addition, AmygA as assessed by multimodal PET imaging could provide collateral information with distinct clinical values and prognostic means, therefore, should comprise an area of active research involving various clinical fields.
Conclusion
The current findings illuminate a novel extension of biological interconnection linking AmygA, AI, and HEMA in human AMI subjects by demonstrating enhanced metabolic activities at baseline followed by concurrent dynamic changes over time. The mechanistic linkage between multi-organ systems suggests that brain emotional neural activity is closely linked with acute plaque instability via augmented haematopoietic macrophage activity. While further interventional studies are needed, the present findings highlight that stress-associated neurobiological activity is a potential therapeutic target for promoting a successful regression of acute plaque instability.
Supplementary material
Supplementary material is available at European Heart Journal online.
Data availability
The corresponding author has full access to the study data, and anonymized data will be available upon request from a qualified researcher.
Acknowledgements
We would like to thank Professor Soon Young Hwang, Ph.D. (Department of Biostatistics, Korea University College of Medicine) for statistical analysis. The authors thank Editage for English language editing.
Funding
The Korean Society of Interventional Cardiology (2017-2 to E.J.P.). This work was supported by the National Research Foundation of Korea (NRF) funded by the Korean Government (NRF-2018R1A2B3002001 to J.W.K.; NRF-2019M3A9E2066882 to J.W.K. and H.Y.). This work was supported by the Korea Medical Device Development Fund Grant funded by the Korean Government (202011C12-02 to J.W.K and H.Y.).
Conflict of interest: none declared.
‡ Twitter handle: Brain emotional neural activity mechanistically linked with acute plaque instability via augmented macrophage activity in human AMI subjects.
References
Author notes
Dong Oh Kang, Jae Seon Eo and Eun Jin Park authors contributed equally as first authors.
