To identify robust circulating predictors for incident atrial fibrillation (AF) using classical regressions and machine learning (ML) techniques within a broad spectrum of candidate variables. In pooled European community cohorts (n = 42 280 individuals), 14 routinely available biomarkers mirroring distinct pathophysiological pathways including lipids, inflammation, renal, and myocardium-specific markers (N-terminal pro B-type natriuretic peptide [NT-proBNP], high-sensitivity troponin I [hsTnI]) were examined in relation to incident AF using Cox regressions and distinct ML methods. Of 42 280 individuals (21 843 women [51.7%]; median [interquartile range, IQR] age, 52.2 [42.7, 62.0] years), 1496 (3.5%) developed AF during a median follow-up time of 5.7 years. In multivariable-adjusted Cox-regression analysis, NT-proBNP was the strongest circulating predictor of incident AF [hazard ratio (HR) per standard deviation (SD), 1.93 (95% CI, 1.82–2.04); P < 0.001]. Further, hsTnI [HR per SD, 1.18 (95% CI, 1.13–1.22); P < 0.001], cystatin C [HR per SD, 1.16 (95% CI, 1.10–1.23); P < 0.001], and C-reactive protein [HR per SD, 1.08 (95% CI, 1.02–1.14); P = 0.012] correlated positively with incident AF. Applying various ML techniques, a high inter-method consistency of selected candidate variables was observed. NT-proBNP was identified as the blood-based marker with the highest predictive value for incident AF. Relevant clinical predictors were age, the use of antihypertensive medication, and body mass index. Using different variable selection procedures including ML methods, NT-proBNP consistently remained the strongest blood-based predictor of incident AF and ranked before classical cardiovascular risk factors. The clinical benefit of these findings for identifying at-risk individuals for targeted AF screening needs to be elucidated and tested prospectively. AMDMS, Averaged minimal depth of a maximal subtree; BMI, body mass index; BP, blood pressure; CRP, C-reactive protein; HDL, high-density lipoprotein; HF, heart failure; LASSO, Least absolute shrinkage and selection operator; LDL, low-density lipoprotein; LOD, limit of detection; MI, myocardial infarction; NT-proBNP, N-terminal pro B-type natriuretic peptide; RSF, Random survival forest; VIMP, variable importance.

High-sensitivity cardiac troponin (hs-cTn) assays are used for detection of myocardial infarction (MI). 99^th percentiles show wide inter-assay variation. Use of sex-specific cutoffs is recommended as definitory cutoff for MI. We compared diagnostic performance and prognostic value of sex-specific 99^th percentiles of four hs-cTn assays in patients with suspected MI. Concentrations of four hs-cTn assays were measured at presentation and after 3 hours in patients with suspected MI. Final diagnoses were adjudicated according to 4^th UDMI. Unisex and sex-specific 99^th percentiles were evaluated as diagnostic cutoffs following the ESC 0/3h algorithm. These cutoffs were used in Cox-regression analyses to investigate the association with a composite endpoint of MI, revascularization, cardiac rehospitalization and death. Non-ST-elevation MI was diagnosed in 368 of 2,718 patients. Applying the unisex 99^th percentile, Elecsys hs-cTnT provided highest negative predictive value (NPV) of 99.7 and a positive predictive value (PPV) of 75.9. The analyzed hs-cTnI assays showed slightly lower NPVs and comparable PPVs (Architect [NPV 98.0, PPV 71.4]; Atellica [NPV 97.7, PPV of 76.1]; Pathfast [NPV 97.7, PPV of 66.6]). Application of sex-specific 99^th percentiles did not significantly affect diagnostic performance. Concentrations above 99^th percentile were independent predictors for impaired long-term outcome (hazard ratios 1.2 - 1.5, p<0.001). We describe a good diagnostic accuracy of four hs-cTn assay using the assay-specific 99^th percentile for detection of MI. Application of sex-specific 99^th percentiles did neither affect diagnostic performance nor prognostic value significantly. Finally, values above the 99^th percentile were associated with poor long-term outcome.

Background: The accurate identification of patients with high cardiovascular risk in suspected myocardial infarction (MI) is an unmet clinical need. Therefore, we sought to investigate the prognostic utility of a multi-biomarker panel with 29 different biomarkers in in 748 consecutive patients with symptoms indicative of MI using a machine learning-based approach. Methods: Incident major cardiovascular events (MACE) were documented within 1 year after the index admission. The selection of the best multi-biomarker model was performed using the least absolute shrinkage and selection operator (LASSO). The independent and additive utility of selected biomarkers was compared to a clinical reference model and the Global Registry of Acute Coronary Events (GRACE) Score, respectively. Findings were validated using internal cross-validation. Results: Median age of the study population was 64 years. At 1 year of follow-up, 160 cases of incident MACE were documented. 16 of the investigated 29 biomarkers were significantly associated with 1-year MACE. Three biomarkers including NT-proBNP (HR per SD 1.24), Apolipoprotein A-I (Apo A-I; HR per SD 0.98) and kidney injury molecule-1 (KIM-1; HR per SD 1.06) were identified as independent predictors of 1-year MACE. Although the discriminative ability of the selected multi-biomarker model was rather moderate, the addition of these biomarkers to the clinical reference model and the GRACE score improved model performances markedly (∆C-index 0.047 and 0.04, respectively). Conclusion: NT-proBNP, Apo A-I and KIM-1 emerged as strongest independent predictors of 1-year MACE in patients with suspected MI. Their integration into clinical risk prediction models may improve personalized risk stratification. Graphical Prognostic utility of a multi-biomarker approach in suspected myocardial infarction. In a cohort of 748 patients with symptoms indicative of myocardial infarction (MI) to the emergency department, we measured a 29-biomarker panel and performed regressions, machine learning (ML)-based variable selection and discriminative/reclassification analyses. We identified three biomarkers as top predictors for 1-year major adverse cardiovascular events (MACE). Their integration into a clinical risk prediction model and the Global Registry of Acute Coronary Events (GRACE) Score allowed for marked improvement in discrimination and reclassification for 1-year MACE. Apo apolipoprotein; CRP C-reactive protein; CRS clinical risk score; ECG electrocardiogram; EN-RAGE extracellular newly identified receptor for advanced glycation end-products binding protein; FABP fatty acid–binding protein; GS Grace Score; hs-cTnI high-sensitivity cardiac troponin I; KIM-1 kidney injury molecule–1; LASSO least absolute shrinkage and selection operator; MACE major adverse cardiovascular events; MI myocardial infarction; NRI net reclassification improvement; NT-proBNP N-terminal prohormone of brain natriuretic peptide.

Background: Current guidelines recommend 0/1 h algorithms using high-sensitivity cardiac troponin (hs-cTn) for fast diagnosis of myocardial infarction (MI). Yet, for some assays, existing data is limited. We aimed to evaluate the diagnostic performance and the prognostic value of a rapid 0/1 h algorithm for the Access hs-cTnI assay. Methods: In consecutive patients presenting with suspected MI, we measured concentrations of Access hs-cTnI at presentation and after 1 hour. Final diagnosis was adjudicated independently by 2 cardiologists. Parameters for diagnostic performance were calculated, applying the recently derived European Society of Cardiology (ESC) 0/1 h algorithm for Access hs-cTnI. Additionally, we assessed the prognostic utility of Access hs-cTnI for the composite end point of all-cause mortality and incident MI at 3 years. Results: In 1879 patients, 257 non-ST-elevation MIs occurred. Application of the 0/1 h algorithm classified 44.5% as rule-out, 20.3% as rule-in, and triaged 35.1% to the observe group. High rule-out safety was confirmed with a sensitivity of 97.7% (95% CI, 95.0%–99.1%) and a negative predictive value of 99.3% (95% CI, 98.4%–99.7%). Rule-in capacity was moderate with a specificity of 88.0% (95% CI, 86.3%–89.6%) and a positive predictive value of 50.8% (95% CI, 45.7%–55.9%). After exclusion of patients with ST-elevation MI the results showed strong prognostic value, even after adjustment for cardiovascular risk factors and comorbidities, with adjusted hazard ratios of 2.51 (95% CI, 1.56–4.04) in the observe and 3.55 (95% CI, 2.18–5.79) in the rule-in group for the composite end point of all-cause mortality and incident MI at 3 years, compared to ruled-out patients. Conclusion: The ESC 0/1 h algorithm for Access hs-cTnI allows safe and efficient triage of patients with suspected MI and has strong prognostic utility up to 3 years after the initial evaluation.

Background: High‐sensitivity cardiac troponin (hs‐cTn)‐based diagnostic algorithms are recommended for the management of patients with suspected myocardial infarction (MI) without ST elevation. Although mirroring different phases of myocardial injury, falling and rising troponin patterns (FPs and RPs, respectively) are equally considered by most algorithms. We aimed to compare the performance of diagnostic protocols for RPs and FPs, separately. Methods and Results: We pooled 2 prospective cohorts of patients with suspected MI and stratified patients to stable, FP, and RP during serial sampling separately for hs‐cTnI and hs‐cTnT and applied the European Society of Cardiology 0/1‐ and 0/3‐hour algorithms comparing the positive predictive values to rule in MI. Overall, 3523 patients were included in the hs‐cTnI study population. The positive predictive value for patients with an FP was significantly reduced compared with patients with an RP (0/1‐hour: FP, 53.3% [95% CI, 45.0–61.4] versus RP, 76.9 [95% CI, 71.6–81.7]; 0/3‐hour: FP, 56.9% [95% CI, 42.2–70.7] versus RP, 78.1% [95% CI, 74.0–81.8]). The proportion of patients in the observe zone was larger in the FP using 0/1‐hour (31.3% versus 55.8%) and 0/3‐hour (14.6% versus 38.6%) algorithms. Alternative cutoffs did not improve algorithm performances. Compared with stable hs‐cTn, the risk for death or MI was highest in those with an FP (adjusted hazard ratio [HR], hs‐cTnI 2.3 [95% CI, 1.7–3.2]; RP adjusted HR, hs‐cTnI 1.8 [95% CI, 1.4–2.4]). Findings were similar for hs‐cTnT tested in 3647 patients overall. Conclusions: The positive predictive value to rule in MI by the European Society of Cardiology 0/1‐ and 0/3‐hour algorithms is significantly lower in patients with FP than RP. These are at highest risk for incident death or MI. REGISTRATION: URL: https://www.clinicaltrials.gov ; Unique identifiers: NCT02355457, NCT03227159.