Derivation and external validation of new ESC 0/1-h algorithm cut-offs for high-sensitivity troponin T and I in the cancer population

Abstract Background The diagnostic performance of high-sensitivity cardiac troponin T/I (hs-cTnT/I) and the efficacy of the European Society of Cardiology (ESC) 0/1h-hs-cTnT/I-algorithms for the early diagnosis of non-ST-elevation myocardial infarction (NSTEMI) is reduced in cancer patients presenting to the emergency department (ED) with chest pain.[1] Purpose This study aimed to derive and externally validate new cut-offs for the ESC 0/1h-hs-cTnT/I-algorithms in chest pain patients with active or past cancer. Methods Patients presenting with acute chest discomfort to the ED were prospectively enrolled in an international multicenter diagnostic study. Final diagnoses were centrally adjudicated by 2 independent cardiologists applying the fourth universal definition of myocardial infarction using all available medical records, cardiac imaging, and serial hs-cTnT/I. Hs-cTn concentrations were measured at presentation and after 1 hour. New assay-specific cutoffs for the ESC 0/1h-hs-cTnT/I-algorithm were derived and externally validated in two independent cohorts. Results Among 541 and 516 eligible patients with 0h and 1h hs-cTnT and hs-cTnI values, respectively, application of the current ESC 0/1h-hs-cTnT/I algorithm triaged 317 (58.6%) and 306 (59.3%) patients to either rule-out or rule-in, leaving 224 (41.4%) and 210 (40.7%) patients in the observe-zone of the ESC 0/1h-hs-cTnT/I algorithm, respectively. The novel ESC 0/1h-hs-cTnT cut-offs ruled out NSTEMI with a cut-off <8 ng/L at presentation (if chest pain onset >3 hours) or <14 ng/L together with a 0/1-hour delta <3 ng/L, ruling-out 241 patients (44.5%) with a sensitivity and negative predictive value (NPV) of 99.2% and 99.6%, respectively. The novel ESC 0/1h-hs-cTnI cutoffs ruled out NSTEMI with a cut-off <7 ng/L at presentation (if chest pain onset >3 hours) or <10 ng/L together with a 0/1-hour delta <3 ng/L, ruling-out 281 patients (54.5%) with a sensitivity and NPV of 99.1% and 99.6, respectively. Findings were confirmed in both internal and external validation (Figure 1). A total of 167 (32.8%) patients died at 5-years, of whom 61 died of cardiovascular causes, and 30 (5.9%) patients had an MI at 5 years. 5-year all-cause mortality for the rule-out group was similar for both the current and the novel ESC 0/1-h algorithm, although the novel one identified patients in the observe zone group as having the highest mortality (Figure 2). Incidence of 5-year cardiovascular mortality or AMI provided by the current ESC 0/1h-algorithm was similar for the rule-in and observe zone groups for both hs-cTnT and hs-cTnI. Conclusions In chest pain patients with a history of cancer, new derived assay-specific cutoffs for the ESC 0/1h-hs-cTnT/I-algorithm achieved a higher proportion of patients triaged to either rule-out or rule-in without decreasing safety. The low mortality rate of the rule-out group identified by the new algorithm further reassures on the safety of the new derived cut-offs.Sensitivity analysis in the APACE cohort5-year all-cause mortality