Clinical significance of borderline elevated troponin I levels across different assays in patients with suspected acute coronary syndrome.
Am J Cardiol. 2009 Jul 15;104(2):164-8
Authors: Zahid M, Good CB, Singla I, Sonel AF
We aimed to elucidate the clinical significance of borderline troponin I (TnI) increases and see if such increases are of similar significance across different assays. Markedly increased TnI is a strong predictor of adverse cardiac events. It is not clear if borderline increases with different commercially available TnI assays provide similar prognostic information. From January 2001 to February 2004, 3 different TnI assays (Beckman Access TnI, Beckman Access AccuTnI, and Vitros ECI Troponin) were used to evaluate the peak TnI value in 1,152 consecutive patients admitted with suspected acute coronary syndrome/non-ST-elevation myocardial infarction (MI). Recommended cutoffs were used to differentiate borderline from marked increases reported as consistent with MI. Clinical data and 30-day death/new MIs were determined by chart abstraction. Demographics and cardiac risk factors were similar for the 3 groups. Frequency of borderline TnI ranged widely among assays (18.7% to 42.1%) but was significantly less with the Vitros ECI Troponin assay (p <0.0001). Prognostic significance of borderline increased TnI values also varied greatly by assay, with borderline Beckman Access AccuTnI increases being predictive of adverse 30-day outcomes (odds ratio 4.0, 95% confidence interval 1.46 to 10.97, p = 0.007), but not with the other 2 assays. Borderline increases were significantly associated with chronic renal insufficiency (CRI; serum creatine >1.5); the relation to adverse 30-day outcomes and borderline increases persisted after correcting for CRI in a multivariate logistic regression model. In conclusion, although borderline increased TnI levels are common and significantly associated with CRI, they do not all portend the same clinical prognosis. This study highlights the need for standardization of TnI levels across different assays.
PMID: 19576340 [PubMed - in process]