Fred S. Apple
Case 1
Following an episode of shortness of breath and fainting, a 79-year-old woman is driven to the emergency department by her family. She has a history of rheumatoid arthritis and coronary artery disease, with limited physical activity. Her initial electrocardiogram (ECG) showed mild, nonspecific changes, including T waves. To assist in her differential, serial cardiac troponin (cTn) values were obtained. While the clinician did not expect the patient to have a myocardial infarction (MI), her substantial increase in cTnI (based on a first-generation assay that is no longer on the marketplace) was at odds with (a) neither a rising or falling pattern found on subsequent values and (b) normal and unchanging total creatine kinase (CK) and creatine kinase-MB (CK-MB) values. Following clinician contact with the laboratory, reanalysis of the specimens using a second-generation cTnI assay by the same manufacturer and a third-generation cTnT assay demonstrated no detectable cTn, and the laboratory results were corrected in the medical record. Follow-up studies by the laboratory revealed the presence of a heterophile-interfering antibody, which once removed (Scantibody tubes) resulted in normal cTnI values by the original assay.
Case 2
The patient presents with a chief complaint of "I have a pain in my chest that hurts very bad." He is a diabetic with a history of atypical chest pain over the past 3 months. He now presents with ischemic symptoms, chest pain, shoulder pain, aching jaw, and nausea. His ECG demonstrates an ST-segment elevation acute MI. His initial cTnT value is increased above the 99th percentile reference cutoff (>0.01 ng/mL) at 0.013 ng/mL and increases to 0.073 ng/mL over the next 4 hours. He is immediately transferred, following medical therapy, to the catherization laboratory, where a stent is successfully placed.
However, the laboratory findings after the initial rising cTnT over the first two samples (0h and 4h), followed by a subsequent decrease on the 8h sample, were quite perplexing to both the attending cardiologist and the pathologist in the laboratory, since the patient was diagnosed with an acute, evolving ST-segment elevation MI. At 12 hours, the cTnT value again demonstrated an increasing value. An astute laboratory medicine resident reviewing the case recollected a paper that demonstrated the potential of lower cTnT results in heparin-plasma specimens (green top tubes) versus serum (red top tubes). Further investigation did reveal that the 0h and 4h specimens were serum and that the 8h and 12h specimens were heparin plasma, and both plasma samples had analytically false low values. When waste serum specimens, drawn for other chemistries, were located in the laboratory refrigerator for the same 8h and 12h draw times and reanalyzed, both showed substantially higher and rising cTnT concentrations at 0.133 and 0.763 ng/mL as expected. Since July 2006, the cTnT assay by Roche, now a fourth-generation assay, has been reformulated and currently does not show any significant difference between serum and heparin plasma, allowing for a laboratory to use mix and matched specimen types. However, as cTnI assays have also been shown to demonstrate either a constant or random lower heparin-plasma cTnT lower bias, it is recommended to use only one specimen type for an individual patient when ruling in or out an acute MI.
Case 3
A 64-year-old male is found unresponsive at home by his wife while he was sitting and apparently watching a football game on a Sunday afternoon at 2:30 p.m. 911 was called after she was unable to arose him. Emergency medical services arrive within 15 minutes and upon examination his ECG demonstrates an ST-segment depression and T waves. He is transported on 100% oxygen to the hospital, and is now awake but disoriented, but complaining of severe chest pain and left shoulder pain. His ECG remains unchanged and his 0h presenting cTnI concentration performed at the bedside using a point-of-care (POC) assay (15-min turnaround time) is within normal limits: less than the 99th percentile cutoff of 0.04 ng/mL. During the course of the patient's treatment in the emergency department, a second POC cTnI at 3 hours is also normal. However, a call from the central laboratory at this time reports that the initial plasma sample (0h) when reanalyzed per protocol in the central laboratory reflects an increased value of 0.07 ng/mL (central laboratory 99th percentile cutoff 0.025 ng/mL). Based on this discrepant finding, the patient is immediately transferred to a telemetry unit and the diagnosis of a non-ST-segment elevation MI is made. Further investigation of two additional serial cTnI samples shows a rising pattern by both the POC and central laboratory assays, but reveals that the POC assay's poor low-end analytical sensitivity was not able to detect the early increase in cTnI until 8h versus 0h for the second-generation central laboratory cTnI assay. Further it was found that there was a poor correlation between the two different assays. This case demonstrates the importance of understanding the limitations of first-generation versus second-generation assays, irrespective of whether they are POC or central laboratory platforms. The first-generation assays are not as analytically sensitive nor as precise as the newer generation troponin assays. Therefore, different clinical impressions based on the troponin assay used can confuse the clinician caring for a patient. One needs to know the assay and understand that not all assays are created equal.
Discussion of Cases 1, 2, and 3
A European Society of Cardiology/American College of Cardiology (ESC/ACC) consensus conference along with the AHA (American Heart Association) /ACC guidelines for differentiating acute MI and unstable angina codified the role of cTn monitoring by advocating that (a) the diagnosis of MI and (b) establishing a high-risk profile (evidence of myocardial injury) are based on increases of cTnI or cTnT in the appropriate clinical setting. These guidelines are also supported by parallel statements by the IFCC Committee on Standardization of Markers of Cardiac Damage (C-SMCD). The guidelines recognized the reality that neither the clinical presentation nor the ECG had adequate clinical sensitivity and specificity for detecting MI without the use of biomarkers. The guidelines do not suggest that all increases of these biomarkers should elicit a diagnosis of acute MI or high-risk profile-only those associated with the appropriate clinical and ECG findings. When cTn increases not due to acute ischemia, the clinician is obligated to search for another etiology for the elevation (see Chapter 8). Overall, the goal of both laboratorians and clinicians is to establish acceptable, uniform criteria for all cTn assays so that they can be objectively evaluated for their analytical qualities and clinical performance].
The first investigators to develop an assay (radioimmunoassay) to measure cTn using polyclonal anti-cTnI antibodies were Cummins et al.. While the assay showed approximately 2% cross-reactivity with skeletal TnI, it still had excellent clinical specificity for cardiac muscle injury. However, the assay was never developed for commercial use. The first monoclonal, anti-cTnI antibody-based immunoassay was described by Bodor et al.. This assay has <0.1% cross-reactivity with skeletal TnI, but it was not suited for clinical use because of the lengthy assay time. Over the past 15 years, numerous manufacturers have described the development of monoclonal antibody-based diagnostic immunoassays for the measurement of cTnI and cTnT in serum. Assay times range from 5 to 30 minutes. Table 1.1 shows that over a dozen assays have been cleared by the Food and Drug Administration (FDA) for patient testing within the United States on central laboratory and POC-testing platforms. In addition to these quantitative assays, several assays have been FDA cleared for the qualitative determination of cTnI and cTnT. Over 50% of the assays are newer second-, third-, or fourth-generation assays that have improved low-end analytical sensitivity, without analytical interferences that have plagued first-generation assays.
Two major hurdles are present that limit the ease for switching from one cTnI assay to another. Assay concentrations fail to agree because (1) there is currently no primary reference cTnI material available for manufacturers to use for standardizing their assays and (2) different epitopes are recognized by the different antibodies used on individual platforms. An effort has been under way for the past 3 years by the AACC Subcommittee on Standardization of Cardiac Troponin I to prepare a primary reference material. In collaboration with the National Institute for Standards and Technology (NIST), a reference material, a cTnT-cTnI-cTnC ternary complex, has been identified (SRM 2921). Working with NIST and the in vitro diagnostic industries, preliminary round-robin studies have demonstrated that while standardization of assays remains elusive, harmonization of cTnI concentrations by different assays has been narrowed from a 20-fold difference to a 2- to 3-fold difference.
cTnI is present in the circulation in three forms: (1) free, (2) bound as a two-unit binary complex (cTnI-cTnC), and (3) bound as a three-unit ternary complex (cTnT-cTnI-cTnC). In addition, there are potentially several additional forms that also exist for these three forms, representing N- and C-terminal degradation forms, oxidation and reduction forms, and phosphorylated forms. Therefore, different assays do not produce equivalent concentration results, and comparisons of absolute cTnI and cTnT concentrations in clinical studies cannot and should not be made because not all assays measure the different forms with equal molarity (Case 3). Comparisons between assay systems must view changes as relative to each assay's respective upper reference limit. Users must understand the analytical characteristics of each troponin assay prior to clinical implementation.
There is only one cTnT assay in the marketplace, currently a fourth generation, due to intellectual property rights owned by Roche. Several adaptations of the cTnT immunoassay kit marketed by Roche Diagnostics (Indianapolis, IN) have been described. Two monoclonal anti-cTnT antibodies are used in the second- through fourth-generation assays. Skeletal muscle TnT is no longer a potential interferent, as was found in the first-generation ELISA cTnT assay. In contrast to cTnI, no standardization bias exists for cTnT because the same antibodies (M11, M7) are used in both the central laboratory and POC quantitative and POC qualitative assay systems. The fourth-generation assay is no longer prone to interference due to heparin, as found in green top sample collection tubes, which previously was shown to cause assay-decreased cTnT and assay-dependent cTnI values when compared to serum.
Surveys on cTn use have been carried out, but the data in the peer-reviewed literature are minimal. The distribution of cTn assays used as reported over the past several years by the College of American Pathologists surveys accounted for approximately 85% of cTnI assays (11 vendors) and 15% cTnT assays (1 vendor). Approximately 10-15% of all users utilize POC-testing assays.
In 2001, the IFCC C-SMCD established recommended quality specifications for cTn assays. The objectives were intended for use by the manufacturers of commercial assays and by clinical laboratories utilizing troponin assays. The overall goal was to attempt to establish uniform criteria in order that all assays could objectively be evaluated for their analytical qualities and clinical performance. Both analytical and preanalytical factors were addressed as shown in Table 1.2. First, an adequate description of the analytical principles, method design, and assay components needs to be made. This includes the following recommendations. Antibody specificity as to what epitope locations are identified needs to be delineated. Epitopes located on the stable part of the cTnI molecule should be a priority. Further, assays need to clarify whether different cTn forms (i.e., binary versus ternary complex) are recognized in an equimolar fashion by the antibodies used in the assay. Specific relative responses need to be described for the following cTnI forms: free cTnI, the cTnI-cTnC binary complex, the cTnT-cTnI-cTnC ternary complex, and oxidized, reduced, and phosphorylated isoforms of the three cTnI forms. Further, the effects of different anticoagulants on binding of cTnI need to be addressed (Case 2). Second, the source of material used to calibrate cTn assays, specifically for cTnI, should be reported. Currently, a cTnI standardization subcommittee of the AACC is recommending the use of SRD 2921 as a primary reference material that will assist in at least harmonizing cTnI concentrations across different assays, providing traceability. Because antibody differences will always be present in different assays, complete standardization will never be possible for cTnI. For cTnT however, as there is only one assay manufacturer (Roche Diagnostics), standardizing between assay generations has been consistent. Third, assays need to describe minimal detection limits and total imprecision at the 99th percentile reference cutoff, as well as potential interferent, such as rheumatoid factors, heterophile antibodies, human antimouse antibodies (Case 1). Preanalytical factors that should be described include effect of storage time and temperature, effect of glass versus plastic tubes and gel separator tubes, and influence of anticoagulants and whole blood measurements. As more assay systems are devised for POC testing, the same rigors applied to the central laboratory methodologies need to be adhered to by the POC-testing systems.
While clinicians and laboratorians continue to publish guidelines supporting TATs of <60 minutes for cardiac biomarkers, the largest TAT study published to date has demonstrated that TAT expectations are not being met in a large proportion of hospitals. A CAP Q-probe survey study of 7020 cTn and 4368 CK-MB determinations in 159 hospitals demonstrated that the median and 90th percentile TAT for troponin and CK-MB were as follows: 74.5, 129, 82, 131 minutes, respectively. Less than 25% of the hospitals were able to meet the <60-minute TAT, representing the biomarker order-to-report time. Unfortunately, a separate subanalysis of just POC-testing systems was not reported. However, preliminary data have shown that implementation of POC cTn testing can decrease TATs to <30 minutes in cardiology critical care and short-stay units. These data highlight the continued need for laboratory services and health-care providers to work together to develop better processes to meet a <60-minute TAT as requested by physicians.
(Continues...)
Excerpted from Markers in Cardiology Copyright © 2007 by Jesse E. Adams. Excerpted by permission.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.