IgG-bound cardiac troponin: influence on diagnosis of MI in chest pain
J Wise, S D Raudsepp, J A Lee, S Aldous, R W Troughton, P D Adamson, A M Richards, W F Peacock, J L Januzzi, J Boeddinghaus, P Lopez Ayala, C E Mueller, M Than, J W Pickering, C J PembertonAbstract
Background
A sizeable proportion of patients presenting to hospital emergency departments (ED) with suspected acute myocardial infarction (MI) have high-sensitivity cardiac troponin (hs-cTn) concentrations that do not meet European Society of Cardiology (ESC) guideline criteria for the 'rule-in' or 'rule-out' pathways and are assigned to the 'observe' pathway. From an analytical perspective, cardiac troponin–immunoglobulin G (macro-Tn) complexes can interfere with hs-cTn assay function, thus leading to erroneous positive or negative results. However, the impact of macro-Tn measurements upon hs-cTn performance within ESC guideline observe criteria is unknown.
Methods
We measured macro-Tn (TnI-IgG, novel ELISA), hs-cTnI (Abbott ARCHITECT), and hs-cTnT (Roche Gen 5) in EDTA plasma collected at ED presentation (t=0) and 2 hours later (t=2) in n=2,955 patients undergoing investigation for chest pain or suspected MI from 3 previously published studies. Statistical analyses assessed the impact of macro-Tn measurement on the proportion of patients classified in the “observe zone”. Binary logistic regression, ROC AUC, and pathway analyses were used to evaluate macro-Tn effect on low-, medium-, and high-risk categories incorporating both clinical, ECG and log transformed hs-cTn variables.
Results
Overall, there were 475 patients (22% male, median (IQR) age 71y (59-80), median (IQR) time from symptom to ED presentation 4.5h(2-11)) categorized as “observe” based on 0h hs-TnI values. This observe cohort was decreased by 10–14% (n=43–67) at 0 and 2 hours by adding a single 0h TnI-IgG cutoff of ≥2.5U for rule out for MI, while maintaining a sensitivity of 95–100%. At 100% sensitivity, specificity improved from 12% to 21% (p=0.009). Similar results were observed using 0h hs-cTnT “observe” cases (n=985; 25% male, median age 71y (59-80), median symptom to presentation of 5h (3-11)), where a single 0h TnI-IgG measurement ≥2.5U increased specificity from 7% to 15% with 99% sensitivity (p=0.020).
Conclusion
Measurement of macro-Tn in patients presenting with chest pain may enhance ESC guideline “observe-zone” risk stratification as a “rule-out” tool. Importantly, this applies to both hs-cTnI and hs-cTnT assays. These findings suggest that macro-Tn measurement may have previously unrecognized clinical utility beyond its potential to confound hs-cTn measurement itself.