Two large new studies have demonstrated that newer, more sensitive troponin assays may improve the early diagnosis of acute MI. The major limitation of current troponin assays is that the tests are less sensitive in the early hours of ischemia. The two studies, along with an accomapnying editorial by David Morrow, are published in the New England Journal of Medicine.
In the first study, by Reichlin et al, 718 patients with possible acute MI had cardiac troponin levels determined by a standard troponin assay and 4 sensitive assays (Abbott–Architect Troponin I, Roche High-Sensitive Troponin T, Roche Troponin I, and Siemens Troponin I Ultra). Diagnostic accuracy of the new assays was significantly improved with the new assays, especially in patients who presented within 3 hours of the onset of pain. The second study, by Keller et al, compared the Siemens Troponin I Ultra to a standard troponin T assay, and found a similar improvement in the early hours of ischemia.
The editorial by David Morrow notes that for the two studies the “principal findings are highly consistent… the accuracy of troponin for the diagnosis of myocardial infarction was improved with the sensitive assays (94 to 96%), as compared with the older assays (85 to 90%). The improved accuracy was most pronounced soon after the onset of chest-pain symptoms.” He observes, however, that these “results also confirm a trade-off of superior clinical sensitivity for diminished clinical specificity.”
Allan Jaffe provided the following detailed response to CardioBrief:
The papers by Keller and Reichlin are welcome additions to the cardiovascular literature. For too long, many have relied on older studies in defining the timing used to make the diagnosis of AMI with contemporary troponin measurements, ignoring smaller studies suggesting what is now shown in 2 large trials, that prompt early detection is not only possible but consistent. These data need to be inculcated into the thinking of groups evaluating new ways of evaluating patients with chest discomfort in the ED.
Several important caveats however are necessary to understand these data in their proper context:
1. When one uses an older less sensitive assay to define AMI as in these studies, new more sensitive methods will always increase the sensitivity of early detection. This improves both early positive predictive value and negative predictive value. However, the new assays also identify another cohort of patients who meet troponin criteria for AMI. This group will likely have less elevated values and will have lower positive and negative predictive abilities in early samples. This issue is dealt with specifically in the Keller paper in figure 2 and table 3 and still supports the value of early samples. The Reichlin paper argues that the sites involved used solid sensitive assays but the differences are so large that one has to question whether the 99th% cut off values were used at those sites and it is hard to determine how the additionally detected AMIs effected early predictive accuracies.
2. As pointed out by the accompanying editorial by Morrow, the use of more sensitive assays will increase identification of those with cardiac injury but will likely reduce the specificity for detection of AMI. Some of this reduction may be blunted by the use of a delta measurement as provided for in the Keller paper with the higher sensitivity assay which helped to identify those with acute disease. Thus, some consideration to insisting on a rising pattern to define those with acute disease might be suggested regardless of the assay being used.
3. All of the assays used in these studies with the exception of the high sensitivity Roche cTnT and cTnI assay are contemporary assays presently in use In the U.S. These studies should be used to guide their usage. High sensitivity assays are being developed. The Roche high sensitivity cTnT assay has been released everywhere but the U.S. I suspect it may identify more patients still than the other contemporary assays but such an analysis was not provided. Nonetheless, one can expect still more sensitive assays to eventually emerge all over the world. When they do, additional patients will be identified and still greater early predictive accuracy can be anticipated.
4. These support the multiple previous studies suggesting that so called early rising markers such as myoglobin and Heart fatty acid binding protein are not of use if one uses contemporary troponin assays and the recommended cut off values.
===========================================================
Don’t lose touch with CardioBrief. Click here to sign up for a daily email newsletter.
===========================================================
Click here to follow CardioBrief on Twitter and receive instant notification of new posts and links
===========================================================
Filed under: Uncategorized
