Prognostic value of [68Ga]Ga-DOTA-FAPI-04 PET/MR for major adverse cardiovascular events after acute ST-elevation myocardial infarction
Y Wang, X Yan, W Quan, R ZhangAbstract
Background
The fibroblast activation protein inhibitor-04 (FAPI-04) specifically binds to the FAP of activated myocardial fibroblasts, which makes [68Ga]Ga-DOTA-FAPI-04 positron emission tomography (PET)/magnetic resonance (MR) a new potential imaging technique for the evaluation of myocardial fibrosis. However, the long-term prognostic value of FAPI PET during the early phase of STEMI remains unclear.
Aims
This study aimed to evaluate the potential value of [68Ga]Ga-DOTA-FAPI-04 PET/MR in predicting major adverse cardiovascular events (MACE) in patients with STEMI.
Methods
A total of 164 patients with STEMI were enrolled. All patients underwent [68Ga]Ga-DOTA-FAPI-04 PET/MR imaging within 45 days post-infarction. The primary endpoint was MACE, defined as a composite of cardiac death, heart failure, stroke, and reinfarction. Cox proportional hazards regression and two distinct machine learning algorithms were used for feature selection.
Results
Among the enrolled cohort (median age 65 years, 86.6% male), 12 patients (7.3%) experienced at least one MACE over a median follow-up of 17.8 (IQR: 12.3–25.7) months, comprising a total of 14 adverse events (2 cardiac deaths, 4 heart failures, 5 strokes, and 3 reinfarctions). [68Ga]Ga-DOTA-FAPI-04 uptake volume (UV), but not maximum and mean target-to-background ratio (TBRmax and TBRmean), was significantly correlated with left ventricular end-systolic volume (r=0.684), end-diastolic volume (r=0.586), and ejection fraction (r=-0.506) (all P<0.001). FAPI UV was a significant predictor of MACE (HR 1.01 per 1-cm³, 95% CI 1.01-1.02,P<0.001), with the hazard increasing by 76% per 1-SD (+63.4 cm³; HR 1.76, 95% CI 1.23-2.52, P=0.002) and 56% per 50-cm³ increment (HR 1.56, 95% CI 1.17-2.08, P=0.002). All PET/MR parameters were further ranked based on their predictive importance for adverse outcomes, using two distinct machine learning (ML) algorithms (random forest and gradient boosting; Figure 1). In both Cox regression and ML analyses, FAPI UV emerged as the most important predictor of MACE. Compared to MR imaging and cardiac function parameters, UV demonstrated better predictive ability (AUC = 0.870, P < 0.001) for MACE, with sensitivity of 66.7% and specificity of 98.1%. Finally, scatter visualization of an LVEF-UV quadrant (Figure 2A) revealed the highest MACE incidence (10.8%) in the high-risk zone (LVEF<50% and UV≥ median). Conversely, patients with a low UV burden showed consistently low event rates (~6%) regardless of LVEF. A logistic regression-based heatmap (Figure 2B) further confirmed that predicted MACE probabilities escalated predominantly along the UV axis rather than the LVEF axis.
Conclusion
[68Ga]Ga-DOTA-FAPI-04 PET/MR is an effective tool to non-invasively quantify myocardial fibroblast activation, and FAPI UV assessed in the early post-infarction period demonstrates potential predictive value for adverse outcomes.Figure 1.Importance rankingFigure 2.Risk mapping