Introduction: The use of high-sensitive cardiac troponin (hs-cTn) assays as part of diagnostic algorithms, i.e. the European Society of Cardiology (ESC) 0/1 hour (h) and 0/3h algorithms, have become the gold standard for the diagnosis of acute myocardial infarction (AMI) in patients without ST-elevation. However, several previous investigations have revealed reduced diagnostic performance in patient populations with chronically elevated hs-cTn concentrations. According to the 4^th Universal Definition of Myocardial Infarction (UDMI), anemia may cause acute and chronic myocardial injury indicated by elevated hs-cTn concentrations. Hence, we aimed to investigate the diagnostic performance of the ESC 0/1h and 0/3h algorithms in patients presenting with anemia and suspected AMI.

Methods: We prospectively enrolled patients with symptoms suggestive of AMI and measured hs-cTnI at admission, 1 and 3h thereafter. Patients were stratified based on the hemoglobin at admission (<9; 9-12; ≥ 12 g/dl). The final diagnoses according to the 4^th UDMI were adjudicated independently by two cardiologists. Patients with ST-elevation myocardial infarction were excluded for the performance analyses of both ESC algorithms. Our primary outcomes of interest were the safety to rule-out (negative predictive value [NPV]) and the accuracy to rule-in (positive predictive value [PPV]) AMI. Patients were followed for up to 4 years for the endpoints of all-cause mortality and cardiovascular events. 

Results: We included 2,223 patients (64.1% males, age 65 [52;75]). Of those, 1,953 (87.9%) patients had a hemoglobin value ≥12g/dl, 244 (11.0%) between 9 and 12g/dl and 26 (1.2%) <9g/dl. The prevalence of AMI was similar for patients with (13.2%) and without (16.3%) anemia, defined as all patients with a hemoglobin <12g/dl (p=0.19). In patients, who were not diagnosed with AMI, the hs-cTnI concentration was significantly higher in those with anemia (p<0.001 for baseline, 1h and 3h, respectively) while there was no difference among patients with AMI (Figure 1A). Application of the ESC 0/1h algorithm revealed similar safety and accuracy to rule-out or -in AMI in patients with (NPV 100.0% [95%CI 92.6, 100.0]; PPV 50.0% [95%CI 37.8, 62.2]) and without anemia (NPV 99.4% [95%CI 98.6, 99.8]; PPV 55.9% [95%CI 50.6, 61.0]). Findings were similar for the ESC 0/3h algorithm (anemia: NPV 98.1% [95%CI 94.6, 99.6]; PPV 46.4% [95%CI 35.5, 57.6]; non-anemia: NPV 97.9 [95%CI 97.0, 98.6], PPV 58.4 [95%CI 53.2, 63.5])). The median follow-up of patients was 1.7 years. After stratification by either ESC algorithm, patients with compared to those without anemia experienced significantly worse outcome regarding all-cause death (p < 0.001; Figure 1B). In the cox-regression analysis adjusted for sex, age and troponin at baseline, anemia was an independent predictor for all-cause death (adjusted hazard ratio [adjHR] 3.3 [95%CI 2.3, 4.6]), cardiovascular death (adjHR 2.7 [95%CI 1.6, 4.7]), incidental AMI (adjHR 2.8 [95%CI 1.1, 7.1]) and rehospitalization (adjHR 1.4 [95%CI 1.1, 1.7]), but not for revascularization (adjHR 1.0 [95%CI 0.6, 1.7]).

Conclusion: The application of the ESC 0/1 and 0/3h algorithms revealed similar safety to rule-out and accuracy to rule-in AMI in patients with and without anemia. However, patients with anemia experience considerable worse outcome and might therefore benefit from additional diagnostic measures and, potentially, treatment targeting anemia and its cause.