High-Sensitivity Cardiac Troponin Assays: Isn't It Time for Equality?
There are abundant data that women who present with acute coronary events are less frequently diagnosed and have worse outcomes than men (1, 2 ) . Part of the problem may be that women are less apt to manifest increased biomarkers when they present and perhaps for that reason, often receive less aggressive guidelinemandated care (3, 4 ) . It may be that the reason for this is that women have lower reference values for biomarkers of cardiac injury that are rarely taken into account. This
... count. This certainly was the case for creatine kinase (CK) 3 MB, for which reference values in women were half of those in men but often were not taken into account (5 ). This appears again to be so with high-sensitivity cardiac troponin assays (6 ), for which the urge for "simplicity" may cause us to repeat this error. In addition, the pathogenesis of coronary artery disease is different in men than in women, with women having lower cardiac troponin values for any given extent of disease (7 ). Further, because women have more endothelial dysfunction and less fixed coronary disease, they are more often diagnosed with type 2 acute myocardial infarctions that are associated with lower cardiac troponin values (8 ). Given these facts, it appears at first glance axiomatic that sex-specific criteria will be necessary to detect all of the women in need of acute care. In addition, if we do not embrace this approach, the cutoff values derived from a mixed population of men and women will be lower than necessary for men, exacerbating the major issue related to high-sensitivity cardiac troponin assays, the lack of clinical specificity for acute myocardial infarction. Into this landscape come 2 recent studies reported in Clinical Chemistry (9, 10 ). Unfortunately, neither is definitive. The first, from the Thrombolysis in Myocardial Infarction (TIMI) group (9 ), evaluated only the initial samples of patients enrolled in 1 of 2 randomized therapy trials. To qualify, patients had to manifest either an increased cardiac troponin or CK MB value or electrocardiographic changes. Based on the fourth generation Roche cardiac troponin T (cTnT) assay data (an assay that is not highly sensitive), 94% of patients (3984/4215) had increased values. The Abbott hs-cTnI assay has a low, non-peerreviewed 99th percentile of 26 ng/L that when broken down by sex gives a cutoff of 16 ng/L for females and 34 ng/L for males. One would anticipate that the Abbott hs-cTnI assay would show markedly increased cTnI values in almost all of these patients from the TIMI group. Because the hs-cTnI assay was the gold standard, there were nearly 3 times more false-positive increases of cTnT than true negatives (485 compared to 144), a disturbing result. Perhaps some of the misclassifications may have resulted from the imprecision of the cTnT assay at values between 10 and 30 ng/L, as suggested in the validation paper on the hs-cTnT assay (11 ). Nonetheless, the inclusion criteria for the study and the late enrollment (mean of 13 h) made substantial increases of hs-cTnI very likely. And that is exactly what the data show. It appears that there were very few patients, regardless of sex, who had minor increases that could allow for a robust evaluation of the need for sex-specific cutoff values. There were only 40 women and 46 men with values between the overall value for all comers and the sex-specific cutoff values. Only 3 of the men who had adverse outcomes were reclassified on the basis of sex-specific values. However, there is much we do not know. Perhaps because we have only the initial hs-cTnI assay values, subsequent local laboratory cardiac troponin values made it clear that an event was occurring and treatment was then initiated, reducing subsequent events. Alternatively, perhaps these individuals did not have ischemic heart disease at all, but structural heart disease, and chronic increases were detected only with the hs-cTnI assay. Because we do not have serial samples, we cannot know whether this was or was not the case. If we suppose that to be the case, then, although cardiac troponin increases might still be of prognostic importance in the long term, they may not identify 30-day, short-term events nearly as well. Thus, this study is far from definitive in regards to whether sex-specific cutoff values will be useful. What is needed are studies similar to the one by Mills et al.