Abstract Background The prognostic value of combining fractional flow reserve derived from coronary CT angiography (FFR-CT) with artificial intelligence enabled coronary plaque analysis (AI-CPA) has not yet been fully uncovered. Purpose The purpose of this study was to investigate how the combined information derived from FFR-CT and AI-CPA, is associated with the long-term risk of cardiovascular death (CVD) and spontaneous myocardial infarction (MI) in patients with new-onset stable angina pectoris (SAP). Methods The study cohort consisted of consecutively enrolled patients with new-onset SAP at three Danish centres participating in the Assessing Diagnostic Value of Non-invasive FFR-CT in Coronary Care registry (ADVANCE-DK). All patients (n=841) had ≥1 coronary stenosis 30% verified on CTA with subsequent successful core laboratory FFR-CT and AI-CPA analysis. Mean follow-up time was 7. 0 years (range: 6. 3-8. 2). An abnormal FFR-CT was defined as a lesion-specific (2cm distal-to-stenosis) value ≤0. 80. AI-CPA was considered abnormal when patient level total plaque burden (total plaque volume mm³/total vessel volume mm³*100) was ≥ the 50%-percentile. Patients were divided into; 1) abnormal FFR-CT + abnormal AI-CPA, 2) abnormal FFR-CT + normal AI-CPA, 3) normal FFR-CT + abnormal AI-CPA and 4) normal FFR-CT + normal AI-CPA. The endpoint was a composite of CVD or non-fatal spontaneous MI. Event data were extracted from the Western Denmark Heart Registry and electronic hospital patient files. An external independent event committee adjudicated all events. Results In 841 patients, FFR-CT was abnormal in 347 (41%) and AI-CPA in 419 (50%). Overall, two normal test results were present in 305 (36%) patients, normal FFR-CT + abnormal AI-QCPA in 189 (22%), abnormal FFR-CT + normal AI-QCPA in 117 (14%) and two abnormal test results in 230 (27%). Baseline characteristics for patients within each group are provided in Table 1. During the follow-up, 69 endpoints occurred of which 35/69 (51%) were spontaneous MIs. Risk of the endpoint was lowest in the category of normal FFR-CT + normal AI-CPA with a significant positive trend towards higher risk if abnormal FFR-CT + normal AI-CPA or normal FFR-CT + abnormal AI-CPA followed by the highest risk observed for abnormal FFR-CT + abnormal AI-CPA, Figure 1. Conclusion In patients with new-onset SAP the combined information derived from FFR-CT and AI-CPA stratified patients into low (double normal), intermediate (one normal one abnormal) and high (double abnormal) risk of adverse outcomes. These findings indicate that FFR-CT and plaque-assessment provides independent and additive long-term prognostic information in patients with intermediate range coronary stenosis.
Vestergaard et al. (Sat,) studied this question.