| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letters |
1 Department of Pathology and Laboratory Medicine, Hartford Hospital, Hartford, CT 06102
2 Department of Laboratory Medicine and Pathology, Hennepin County Medical Center, University of Minnesota, School of Medicine, Minneapolis, MN 55415
aAuthor for correspondence. E-mail awu{at}harthosp.org.
To the Editor:
The letter by Tate et al. attempts to resolve the issue of which cutoff concentration for troponin should be used in the diagnosis of myocardial infarction (MI) and risk assessment of acute coronary syndrome (ACS) patients. One of the underlying goals of the European Society of Cardiology (ESC)/American College of Cardiology (ACC) (1) and American Heart Association (AHA)/ACC guidelines (2) as well as the IFCC (3) recommendations for establishing a MI cutoff at a concentration with a 10% CV was to challenge manufacturers of cardiac troponin assays to improve the low-end analytical characteristics at the 99th percentile reference limit, to better identify ACS patients who are at higher risk of short- and long-term cardiac events. Our profession demands evidence-based studies to validate each troponin assay individually, addressing 99th percentile reference limits with appropriately powered numbers for gender and race, with MI diagnostic findings and risk-stratification information based on these reference limits. Clearly, the growing evidence-based literature supports the notion that any measurable troponin in ACS patients has pathologic significance for risk assessment independent of an assay’s analytical precision.
At present we suggest that both the laboratory medicine and clinical communities rally behind the ESC/ACC and AHA/ACC guidelines, which are supported by the IFCC Committee on Standardization of Markers of Cardiac Damage (CSMCD), which proposed the troponin 99th percentile reference limit for MI detection if a 10% CV can be attained. For assays for which the imprecision at the 99th percentile is >10%, we continue to support the lowest concentration that produces a 10% CV as the diagnostic cutoff (4)(5).
For the purposes of continuing debate, Tate et al. suggest the use of a mathematical model for calculation of the minimum detection limit for cardiac troponin assays and use of this value as the diagnostic cutoff. This model computes a limit from the mean plus 3 SD above zero. Use of this criterion for the cutoff will produce values that are necessarily below the 10% CV cutpoint. Although the IFCC CSMCD has also recommended this analytical approach for determining the detection limit for troponin assays, the CSMCD endorsed a higher cutpoint (i.e., 10% CV) as the decision limit (4). Tate et al. justified the lower cutoff by citing risk stratification studies such as Morrow et al. (6) in the TACTICS-TIMI Trial, where additional risk stratification was demonstrated for troponin concentrations between the detection limit and the 10% CV cutpoint. Similarly, in the FRISC II trial, additional risk stratification data were produced when the cutpoint for cTnT was lowered from the 10% CV value of 0.03 µg/L to 0.01 µg/L (7). Despite the increased number of patients identified at high risk by use of this lower cutoff, the authors of this last study nevertheless went against their own evidence and recommended use of the higher 10% CV cutpoint.
Although the data from these clinical trials are compelling, the use of a biomarker for acute MI (AMI) diagnosis is different from its use for risk stratification. Although all patients with confirmed AMI will require immediate medical attention, those patients in whom an AMI was ruled out but who are identified as having high future risk can be managed on a more elective basis. It is also important to consider differences in the prevalence and pretest likelihood of patients who present acutely with a suspicion of AMI from those with confirmed ischemic disease where future risks are being assessed. The TACTICS-TIMI Trial enrolled 2220 patients, all with ACS (6). The composite adverse event rate (death, MI, or rehospitalization for ACS) for the enrolled patients was 9.2% at 30 days and 17.7% at 6 months. If the purpose of measuring troponin is only to risk-stratify patients with ACS for adverse events, consideration should be given to lowering the troponin cutpoint below the 10% CV value, as suggested by Tate et al. However, these low troponin cutoffs are not likely appropriate for the primary diagnosis of AMI or acute ischemia in a cohort of patients with chest pain, where the prevalence of disease is 15–25% (8)(9)(10).
Tate et al. indicated that use of the mean + 3 SD method might produce one false positive among 100 positive test results. This might be an acceptable cutoff in a confirmed ACS population in which 10% of the individuals develop an adverse event at 4 weeks. However, in a chest pain population, the prevalence of a individual having ACS and developing an adverse event is much lower. Consider, for example, the case of testing a population of chest pain patients who have an ACS prevalence of 20% and total population rate of adverse events of 2% (i.e., 10% of the 20% with ACS). If 10 000 patients were screened by a troponin assay with a low cutoff that yields a 1% false-positive rate, 9800 would be negative for ACS, with 9702 having a true-negative troponin result and 98 a false-positive result (1%). The remaining 200 patients would have ACS, and presumably all of these would have a positive troponin. Given these assumptions, the false-positive rate is roughly one-third of all patients with a positive troponin (98 of 298). This high fraction of false positives for this population would be inappropriate given the potential for adverse reactions produced by the therapy, as stated by Tate et al. In actual clinical practice, however, one would also factor in the clinical history, type and severity of chest pain, electrocardiographic changes, and other measures, which would reduce the number of patients undergoing unnecessary therapy.
The principal question then becomes, "Should laboratories have one troponin cutoff for AMI diagnosis and a second, lower troponin cutoff for risk stratification of patients with confirmed ACS?" It is our feeling that administration of antithrombotic and antiplatelet therapies on the basis of a very liberal low biomarker cutoff concentration alone in a low-risk population is difficult to justify. On the other hand, a patient with independent evidence of ischemia, e.g., ST-segment depression on the electrocardiogram, might benefit from therapy even if the troponin is below the AMI cutoff but above the risk stratification cutoff. Although we do not endorse two cutoffs for cardiac troponin, clinicians should be aware of the clinical evidence that any increase in troponin predicts increased cardiovascular risk (11).
In summary, we commend the efforts of Tate et al. We believe, however, that this will only lead to further confusion, as their formula is not evidence-based with clinical trial support. One must also not lose sight of the continued need to monitor biomarkers over time, to continue to assess serial patterns documenting an increasing (or decreasing) cardiac troponin pattern after presentation. The impact of lower cutoff values on society, epidemiologic studies, clinical trials, and routine clinical practice are and will continue to be critically important and must not be taken lightly. We sense that there will be more evidence in support of the 99th percentile and 10% CV cutoffs. The 10% CV cutoff can be determined in each laboratory by performing day-to-day precision studies using serum or plasma samples containing low concentrations of troponin. Many manufacturers are now listing the 10% CV cutoff value in their package inserts.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
