We retrospectively enrolled patients who underwent gated 99mTc-sestamibi myocardial perfusion imaging (MPI) and 18F-FDG cardiac PET as for myocardial viability assessment to guide the treatment strategy decision making at the department of nuclear medicine, Beijing Anzhen Hospital, affiliated with Capital Medical University from January 2019 to January 2021, and diagnosed with PMI, RMI, or CSA, respectively. Patients with MI (PMI and RMI) were examined at 30.0 (interquartile range: 25.0–40.0) days. The study protocol was approved by the ethics committee of Beijing Anzhen Hospital, affiliated with Capital Medical University (Approval No. 2017024).

PMI was defined as typical changes in biochemical markers of myocardial necrosis along with ≥ 1 of the following: ischemic symptoms, electrocardiographic changes indicative of new ischemia, development of pathological Q waves, or imaging evidence of new loss of viable myocardium or new regional wall motion abnormalities [11]. For inclusion, patients with RMI were required to have received at least two distinct diagnoses of MI within a 6-month period, defined as elevated cardiac enzymes above the diagnostic threshold, with positive high-sensitive troponin-I (hsTnI) values recorded at least once within 96h after MI [7], along with a typical clinical presentation or typical electrocardiographic changes. CSA was defined as the presence of stable anginal symptoms for ≥ 6 months with ≥ 50% luminal narrowing in ≥ 1 major coronary artery on angiography. Patients with a history of an inflammatory condition, or those taking inflammation-modulating medications within the prior 6 months, malignancy, or severe hepatic disease were excluded.

The following clinical characteristics were recorded: age, sex, body mass index, history of hypertension, diabetes mellitus, dyslipidemia, and smoking. Cardiac hsTnI, creatine kinase-MB and brain natriuretic peptide (BNP) fractions were measured.

As previously reported [12], MPI was performed 90–120 min after injection of 99mTc-sestamibi (740 MBq, Chinese Atomic Energy Institute, Beijing, China). Images were acquired for 10 min with a dual-headed Siemens Camera (Siemens Symbia Intevo 16 Systems), equipped with a multifocal (SMART ZOOM) collimator. Gated data were acquired with a 20% energy window centered over 140 keV. Images were reconstructed with the flash 3D mode and displayed as short axis and horizontal, and vertical long-axis slices.

As previously described, patient preparation followed the protocol as outlined in the 2016 American Society of Nuclear Cardiology (ASNC) guidelines [13]. After at least 12 h of fasting, the blood glucose level was controlled by oral glucose loading and, if needed, by supplemental iv insulin doses as recommended in the ASNC guidelines [13]. Blood glucose levels averaged 5.88 ± 0.67 mmol/L at the time of the injection of 18F-FDG (Chinese Atomic Energy Institute, Beijing, China). 18F-FDG cardiac PET was acquired for 10 min. Attenuation correction was performed based on CT data (120 kV, 11 mAs). Image reconstruction employed a point spread function + time of flight (TOF) algorithm (TrueX + TOF, UltraHD-PET), with 2 iterations and 21 subsets (Siemens AG, Munich, Germany).

18F-FDG metabolic activity in BM was calculated under CT-guided anatomic reference from the fifth to eighth thoracic vertebrae according to our previous investigation [14]. Besides, the average of the highest standard uptake value (SUV) was used as the mean SUV, and normalized to the right atrium. Splenic 18F-FDG metabolic activity was assessed by placing regions of interest in 3 planes (axial, sagittal, and coronal planes), SUVmax was recorded in each plane, and the splenic activity was calculated as the mean of SUVmax values of the 3 planes and normalized to the liver. Aortic 18F-FDG metabolic activity was quantified in the region of interest around each aorta on every slice of the trans-axial fusion PET/CT images. The highest SUV of the region of interest of all 3 slices within the aorta was averaged for each subject [15]. Thus, the aorta SUV was divided by the blood-pool SUV measured from the right atrium for normalization. In this way, the aortic target-to-background ratio (TBR) was calculated for each subject.

Left ventricular (LV) global functional and remodeling parameters, including LV ejection fraction (LVEF, %), end-diastolic volume (EDV, mL), and end-systolic volume (ESV, mL) were calculated by using QGS software (version 3.1, Cedars-Sinai Medical Center, Los Angeles, CA, USA), with manual correction in case of the inadequate endocardium and epicardium delineation [16]. As in our previous investigations [17, 18], myocardial perfusion and metabolic activity were assessed by two experienced physicians using the American Heart Association 17-segment and five-point scoring system. Hibernating myocardium (HM, %) was defined as a mismatch score of 1.0 or greater (perfusion score minus metabolism score ≥ 1). The infarcted tissue (scar, %) was defined as a mismatch score of less than 1.0 (perfusion score minus metabolism score < 1). One segment accounted for 6% of LV; the extent of HM and scar were calculated from the number of segments with mismatches or matches while the total perfusion deficit (TPD, %) was derived from the number of hypo-perfused segments and their deficit severity [19].

Follow-up was performed by consulting the electronic medical record system and contacting patients or their relatives by telephone. The primary endpoint was major adverse cardiac events (MACE), including all-cause death, cardiac death, MI, and readmission due to heart failure. The median follow-up time was 16.6 months (interquartile range 11.7–32.0 months).

The baseline characteristics of the patients were analyzed according to the 3 groups. Frequencies and proportions were reported for categorical variables, and either mean ± SD or median with interquartile range was reported for continuous variables based on normality of distribution. The χ2 test, Fisher exact test, 1-way ANOVA, and Kruskal Wallis test were used to compare variables among groups. Subsequent comparisons were performed by the Bonferroni post hoc test, Mann–Whitney U test, or Fisher exact test with Bonferroni-corrected P values. The 18F-FDG metabolic activity of the BM and spleen aorta were compared among the 3 groups using ANOVA with Bonferroni for multiple comparisons. Spearman correlation analysis was performed to identify the relationship between 18F-FDG metabolic activity in BM, spleen, and aorta. The cumulative incidence of MACE-free was estimated by the Kaplan–Meier curve and compared by the log-rank test. Data were analyzed using SPSS for 26.0 (SPSS, Chicago, IL). Statistical significance was defined as P < 0.05.