Aiming towards evidence based interpretation of cardiac biomarkers in patients presenting with chest pain : The WESTCOR study

Abstract

Bakgrunn: Ca 10% av alle pasientar i akuttmottak blir innlagt med brystsmerter og spørsmål om akutt koronarsyndrom (AKS), men under ¼ av pasientane har AKS. Metode:WESTCOR studien er ein prospektiv observasjonsstudie som inkluderer pasientar i akuttmottak med mistanke om AKS. Ein har brukt ulike vinklingar for å identifisera kardielle endepunkt, og samanliknar endepunkta med anbefalingar frå European Society of Cardiology (ESC): 1) Utvikla 0/1 times algoritmar for eit nytt høg sensitivt troponin I assay frå Singulex (hs-cTnISgx), 2) Samanlikning av ESC 0/3 timars algoritmar (2015) og High-Sensitive Troponin i Evaluering av patientar med Akutt Koronar Syndrom (high-STEACS) algoritmar kombinert med AKS kriteria frå ESC eller elleve ulike risikoskårar, 3) Utvikling av nye hs-cTnT og hs-cTnI algoritmar som nyttar lave slutningsgrenser både for nullprøven og deltaprøvane (1 og 3 timar). Resultat: 1) Samanlikna med andre troponinmetodar, hadde nullprøven frå hs-cTnI(Sgx) signifikant høgare areal under kurven (AUC) (0.95 versus 0.91 hs-cTnTESC, P<0.001 og 0.93 for hs-cTnIESC(Abbott), P=0.004, Delong test) for å utelukka non-ST elevasjons myokardial infarkt (NSTEMI). 0/1 timars hs-cTnI(Sgx) algoritmane plasserte 92% av pasientane i enten låg eller høgrisikogruppa for NSTEMI, medan hs-cTnTESC og hs-cTnIESC(Abbott) plasserte ≤78%. 0/1 times hs-cTnI(Sgx) algoritmen gav ingen langtids prognostisk information (total mortalitet og MI). 2) Dei to 0/3 timar algoritmane kombinert med AKS kriteria plasserte 3,8-4,9% av pasientane som fekk det kombinerte endepunkt MI, total mortalitet, og revaskularisering innan 30 dagar i lågrisikogruppa. Kombinasjon med HEART score ≤3 reduserte endepunkta i lågrisikogruppa til 2.2-2.7%. 3) Den nye hs-TnT 0/1 times algoritmen hadde ein sensitivitet for det kombinerte endepunktet NSTEMI eller ustabil angina pectoris (UAP) under opphald på 95% versus 63% for ESC algoritmen (P<0.001) Evna til å senda ut pasientar med den nye hs-TnT 0/1 algorithm vart signifikant redusert, 21% versus 82 % (P<0.001). Algoritmen frå hs-cTnI(Abbott) hadde sensitiv på 87% og spesifisitet 45%. Konklusjon: Dette arbeidet viser nye måtar å vurdera og forbetra diagnostikken av brystsmerterpasientar i akuttmottak.

Background: Chest pain patients admitted to the emergency department (ED) with possible acute coronary syndrome (ACS) encompasses about 10% of the ED population, only a relatively small proportion have ACS. Method: The WESTCOR study is a prospective observational study, including patients with ACS symptoms. Different approaches for predicting major adverse cardiovascular events (MACE) are investigated and compared to European Society of Cardiology (ESC) recommendations: 1) novel 0/1 hour algorithms using a high sensitivity troponin I assay from Singulex (hs-cTnISgx) with measurable results in >99% of healthy persons. 2) 0/3 hour ESC (2015) and the High-Sensitivity Troponin in the Evaluation of patients with Acute Coronary Syndrome (high-STEACS) algorithms combined with the ACS criteria from ESC or eleven different risk scores: 3) novel hs-cTnT and hs-cTnI rule-out algorithms designed with low baseline/low (1-3) hour delta values. Results: 1) The hs-cTnI(Sgx) baseline value for rule-out of non-ST elevation myocardial infarction (NSTEMI) had significant higher Area under the Receiver Operating Characteristic (AUROC) (0.95 vs. 0.91 for hs-cTnTESC, P<0.001 and 0.93 for hs-cTnI(Abbott), P=0.004, Delong test). The 0/1-hour hs-cTnI(Sgx) algorithms allocated 92% of patients to rule-in/rule-out, while comparator algorithms only ≤ 78%. The 0/1-hour hs-cTnI(Sgx) rule-out algorithm did not provide prognostic information for combined all-cause mortality and future nonfatal MI. 2) Combining 0/3- hour troponin algorithms from ESC or high-STEACS with the ACS criteria, ruled out 3.8-4.9% of patients who experienced MI, all-cause mortality, and revascularization within 30 days. A HEART score ≤3 reduced the event rate to 2.2-2.7%. 3) The clinical sensitivity for the hs-TnT 0/1-hour rule-out algorithm for predicting combined NSTEMI or unstable angina pectoris (UAP) during index hospitalization was 95% vs. 63% for the ESC algorithm (P<0.001)). The rule-out rate for the nove algorithm was significantly reduced, 21.0% vs 82% (P<0.001). The novel algorithm for hs-cTnI(Abbot) had slightly lower sensitivity 87% but better specificity 45%. Conclusion: This work suggests several novel approaches for improving the diagnostic workup of chest pain patients in the ED.