Development and internal validation of a prognostic model incorporating 18F-FDG PET myocardial ischemic memory imaging to predict major adverse cardiovascular events in patients with suspected unstable angina: an observational cohort study in China

18 F-fluorodeoxyglucose positron emission tomography myocardial ischemic memory imaging may serve as a novel prognostic marker for major adverse cardiovascular events in patients with suspected unstable angina. This study aimed to develop and internally validate a nomogram integrating 18 F-fluorodeoxyglucose positron emission tomography myocardial ischemic memory imaging and clinical variables for individualized prediction of the risk of major adverse cardiovascular events in this population. A total of 240 patients with suspected unstable angina who presented with chest pain, a normal initial electrocardiogram, and initially negative cardiac biomarkers were enrolled. The primary endpoint was time to the first major adverse cardiovascular event, defined as cardiovascular death, acute myocardial infarction, heart failure, readmission due to unstable angina, and stroke. Patients were randomly divided into a training cohort ( n = 168) and an internal validation cohort ( n = 72) at a 7:3 ratio. A resting fasting 18 F-fluorodeoxyglucose positron emission tomography protocol was used to identify ischemic memory-related metabolic abnormalities. Least absolute shrinkage and selection operator regression based on the Cox proportional hazards model with 10-fold cross-validation was used for predictor selection in the training cohort. Model performance was assessed using time-dependent receiver operating characteristic analysis, calibration, decision curve analysis, split-sample validation, and 1000 bootstrap resamples of the training cohort to quantify optimism. During follow-up (median 26.3 months, range 14–50 months), 50 (20.8%) major adverse cardiovascular events were recorded. The final nomogram included six predictors: male sex, previous myocardial infarction, hypertension, abdominal circumference, Global Registry of Acute Coronary Events risk score, and 18 F-fluorodeoxyglucose positivity. At 26 months, the time-dependent area under the receiver operating characteristic curve was 0.876 (95% confidence interval 0.811–0.943) in the training cohort and 0.769 (95% confidence interval 0.708–0.860) in the validation cohort. Calibration plots appeared broadly consistent with the observed risks, and decision curve analysis suggested potential net benefit within the examined threshold range. Bootstrap internal validation in the training cohort yielded an optimism-corrected 26-month area under the receiver operating characteristic curve of 0.834. In this single-center exploratory analysis, a prognostic model integrating clinical variables with 18 F-fluorodeoxyglucose positron emission tomography myocardial ischemic memory findings showed potential for estimating the risk of major adverse cardiovascular events. However, given the limited number of events and the lack of external validation, the model should be regarded as hypothesis-generating. Validation in larger prospective multicenter cohorts is required before clinical implementation.