The incremental value of non-necrosis biomarkers to risk stratification in acute chest pain, with MI excluded: the first prospective long term cohort study in the high sensitive troponin era

Abstract

Abstract Background Approximately 90% of patients presenting to ED (Emergency Department) with suspected Acute Coronary Syndrome (ACS) have Myocardial Infarction (MI) excluded. Long term outcome and risk stratification for this cohort has rarely been studied. Purpose To study long term cardiovascular outcomes in patients with acute Chest Pain (CP) in whom MI is excluded, and to investigate the incremental value of biomarkers of inflammation, myocardial ischaemia and stress, for risk stratification, to that of existing methods of stratifying risk including absolute values of two high sensitive troponins below the 99th percentile. Methods A prospective observational study in a large university teaching hospital with a median follow-up of 5.8-year. Cardiovascular outcomes studied included adjudicated type 1 MI (to 4th universal definition), unstable angina requiring urgent revascularisation and cardiovascular death. Inclusion criteria included CP with features of ischaemia but with MI excluded (High sensitive cardiac troponin [hs-cTn] T ≤99th percentile). Exclusion criteria included non-cardiac chest pain, chronic atrial fibrillation, heart failure and recent ACS. The biomarkers HFABP, GDF-15, NT-proBNP, Galectin-3, HS-CRP taken on first venepuncture for high sensitive cardiac troponins were investigated using univariable and multivariable backward stepwise elimination logistic modelling, receiver operator characteristic (ROC) analysis, and Net Reclassification Indices (NRI). They were compared to absolute values of both sub-99th percentile hs-cTnT (Roche Elecsys) and hs-cTnI (Abbot Architect) both as a single entity or when the troponin values were incorporated with HEART and TIMI scores. Results 487 consented patients were enrolled. The mean age of the population was 56 years (SD 11.8), 44% were female, 16.2% were diabetic, 20.9% had a previous MI, 13.8% previous PCI and 5.1% coronary artery bypass grafting. 100% follow-up was achieved. There were 48 events in follow-up. (event rate 9.9%). Multivariable Cox Proportional Hazards analysis including hs-cTnT or hs-cTnI (below 99th percentile) demonstrated incremental improvement in risk stratification with absolute values of either hs-cTn below the 99th percentile when added to either TIMI or HEART risk score. The results were consistent with NRI analysis. Only GDF-15 demonstrated significance in multivariable analysis when analysed with the HEART score regardless of troponin assay and the TIMI score with HS-cTnI but not with HS-cTnT. Conclusion Absolute values of hs-cTnT or I, below 99th percentiles, significantly improve risk modelling for cardiovascular outcomes in acute CP patients, presenting to ED, who have MI excluded. No other biomarker retains independent risk with the exception of GDF-15. Effective dichotomisation of sub99th percentile hstnT or I, used alone or incorporated into risk scores, could allow improved targeting of resources for investigation and management for this large pool of patients.