Author + information
- L. Kristin Newby, MD, MHS∗ ( and )
- Angela Lowenstern, MD
- Division of Cardiology, Department of Medicine, and the Duke Clinical Research Institute, Duke University Medical Center, Durham, North Carolina
- ↵∗Address for correspondence:
Dr. L. Kristin Newby, Division of Cardiology, Department of Medicine, Duke Clinical Research Institute, P.O. Box 17969, Durham, North Carolina 27715-7969.
The advent of high-sensitivity cardiac troponin (hs-cTn) assays has led epidemiologists and clinicians alike to ponder the implications that the transition from contemporary fourth-generation to hs-cTn assays will have on trends in the incidence of myocardial infarction (MI), as well as resource use and costs. Will we see an “epidemic” of MI, just as epidemiological trends over the past decade in many countries, including the United States, have shown declines in MI incidence? (1). Will our cardiac catheterization labs be flooded with patients with “troponinemia”? Will our inpatient cardiology wards and cardiac intensive care units be stretched beyond capacity by an increased volume of patients with MI? Given the much-increased sensitivity of these assays for cardiac myonecrosis and their lack of etiologic specificity, these are not unreasonable questions and concerns.
Hs-cTn assays have now been used in clinical practice in many parts of the world for 8 years, and validated algorithms for hs-cTn testing are available and have been incorporated into European guideline recommendations (2). The United States, which saw the first hs-cTn assay approved by the U.S. Food and Drug Administration in January 2017, has a tremendous opportunity to learn from the experience of its colleagues in Europe, Canada, Australia, and New Zealand as it begins to transition to use of hs-cTn assays. To date, this experience suggests that MI rates and resource use may not increase as dramatically as is feared, and not nearly as dramatically as occurred with the transition from creatine kinase–based testing to conventional troponin assays—in one study, there was an increase in MI of 64% among men and 95% among women (3).
In this issue of the Journal, using nationwide health records from 2007 to 2013, Odqvist et al. (4) examined rates of MI diagnoses before and after implementation of hs-cTn assays at Swedish hospitals and trends in subsequent mortality, reinfarction, and resource use according to the type of assay (conventional or high sensitivity) used to make the MI diagnosis. As observed in the United States, there was a yearly decline in MI incidence in Sweden that was evident before the transition to hs-cTn assays began and that was sustained through and beyond it, though with modest attenuation of the decline during the transition years. This 27% overall decline in MI occurred despite a 5% overall hospital-level increase in MI diagnoses during the 90 days following implementation of hs-cTn assays compared with the 90 days before. Even with the 5% increase in MI diagnoses, the overall frequency of coronary angiography and revascularization was not significantly different between the 90 days before and 90 days immediately following hs-cTn assay implementation. These hospital-level results from Sweden are consistent with observations from the multinational APACE (Advantageous Predictors of Acute Coronary Syndrome Evaluation) study (5). Rather than considering MI diagnoses among all patients at a given hospital, the APACE study focused on patients evaluated in the emergency departments (EDs) of 3 hospitals and for longer periods of observation—2 to 4 years before and 1 to 3 years after the transition from conventional to hs-cTn assays. Despite these study design differences, there was a similar modest increase in MI diagnoses (10% vs. 14%; p < 0.001) comparing the periods before and after implementation. As in the Odqvist et al. (4) study, rates of cardiac catheterization were similar pre- and post-implementation (23% in both periods), with similar distributions of the extent of coronary disease and subsequent revascularization rates (5).
Using the Swedish nationwide health data, Odqvist et al. (4) were able to extend their observations to clinical outcomes over a mean of 3.1 ± 2.3 years among patients diagnosed with MI stratified by the assay (conventional vs. high sensitivity) used for diagnosis. There was no difference in all-cause mortality according to which assay was used, but the risk of reinfarction was 11% lower among patients diagnosed with the hs-cTn assay compared with the conventional assay. The underpinnings of this difference are not certain. There were only small, nonsignificant differences in discharge prescription of evidence-based medicines by assay type. Although there was no increase in coronary angiography and intervention at a hospital level during the immediate 90 days post-transition, considering only patients with MI, those diagnosed using the hs-cTn assay were 16% more likely to undergo coronary angiography and 13% more likely to undergo revascularization during follow-up than those diagnosed using conventional assays. It is tempting to associate the reduction in reinfarction among patients with MI diagnosed using the hs-cTn assay with increased use of angiography and revascularization, but this may be too simplistic without more details about the circumstances and timing of use and coronary anatomy. Furthermore, although there were no significant differences in discharge medication use by assay type, treatment benefit may have been greater among patients detected as having MIs by hs-cTn assays who would have been missed by the conventional assay. As Mills et al. (6) showed during the blinded phase of a transition from a third- to fourth-generation conventional troponin assay, patients for whom the MI diagnosis was missed by the less-sensitive assay were at highest risk of adverse clinical outcomes. When results of the more sensitive assay were clinically available, this group, which previously would have been missed by the less-sensitive assay, had the greatest improvement in outcomes, concurrent with increased use of both guideline-recommended medications and coronary angiography and intervention (6). Moreover, even if discharge prescription of evidence-based medications was not substantially different by type of troponin assay, it is unknown whether factors such as medication compliance, which medications were prescribed, or use of cardiac rehabilitation post-MI differed according to assay type. Finally, it is possible that secular trends in these or other factors may have also contributed to the 27% overall reduction in MI observed in Sweden from 2007 to 2013.
The Odqvist et al. (4) study raises another important consideration in the implementation of hs-cTn assays. The increased sensitivity and lower 99th percentile values will identify more individuals with abnormal troponin values, not all of which will be due to MI, leading to low positive predictive values, even when the key tenets of the universal definition of MI (clinical scenario consistent with ischemia and rising and/or falling biomarker pattern) are present (7). This potential limitation may be mitigated by capitalizing on hs-cTns as quantitative biomarkers, rather than simply considering them as “positive” or “negative” relative to the 99th percentile (8,9). Using higher threshold values combined with the amount of change between assessments, earlier rule-in with more reasonable positive predictive value can facilitate triage and subsequent evaluation (2,8,9). In the Odqvist et al. (4) study, the effect of implementation of hs-cTn assays on MI rates at individual hospitals varied widely (−37% to +93%). In addition to small numbers of MIs during the pre- and post-implementation periods at many hospitals, there were also differences among hospitals in the hs-cTn assays and diagnostic thresholds used. The investigators noted that at hospitals using higher MI thresholds for the hs-cTn assays, there was little difference in MI rates before and after implementation.
Although many concerns about hs-cTn assays focus on identification of more patients with abnormal levels, the experience from Odqvist and others outside the United States should provide some reassurance that the increase in MI diagnoses will be modest and manageable. Furthermore, the exquisite sensitivity of these assays, coupled with their enhanced precision, allows earlier MI rule-out with high negative predictive value, which holds promise to help safely decongest crowded EDs and reduce costs (8,9). In the APACE study, after implementation of hs-cTn assays, use of stress testing declined from 29% to 19% (p < 0.0001), median ED length of stay declined by 79 min (p < 0.001), and mean costs were 20% lower (p < 0.002) (5). These gains could be even more substantial in the United States, where troponin testing is more prevalent and ED length of stay for rule-out MI is generally longer (9).
The impact of hs-cTn assays on clinical care in the United States remains to be fully elucidated. Implementation will undoubtedly encounter challenges, but we should embrace the experience emerging from 8 years of use outside of the United States as a solid foundation upon which to build. Adoption (with appropriate modifications) of algorithms for more rapid MI rule-out, resulting in shorter ED stays and decreased costs will be an immediate benefit. Although the data from Odqvist and other investigators are reassuring, we will need to develop strategies to manage patients with myonecrosis from etiologies other than MI that will be detected using hs-cTn assays. Successful transition will require careful planning that engages the laboratory, emergency medicine physicians, and cardiologists in development of strategies for use of hs-cTn assays, as well as intensive education of all providers about the properties, strengths, and limitations of the assays and their appropriate use and interpretation. Finally, it will be important to monitor outcomes during and after the transition and be prepared to adapt our approaches as we learn from the accruing data. Change is often unsettling and can bring challenges, but the benefits for our patients and health care systems with the transition to hs-cTn promise to be tremendous if we embrace the change and are active participants in shaping it.
↵∗ Editorials published in the Journal of the American College of Cardiology reflect the views of the authors and do not necessarily represent the views of JACC or the American College of Cardiology.
Dr. Newby has received consulting honoraria from Roche Diagnostics, Ortho-Clinical Diagnostics, and Philips Healthcare. Dr. Lowenstern is supported by a National Institutes of Health training grant to Duke University (5 T32 HL 69749).
- 2018 American College of Cardiology Foundation
- Benjamin E.J.,
- Virani S.S.,
- Callaway C.W.,
- et al.,
- American Heart Association Council on Epidemiology and Prevention Statistics Committee and Stroke Statistics Subcommittee
- Roffi M.,
- Patrono C.,
- Collet J.P.,
- et al.,
- ESC Scientific Document Group
- Odqvist M.,
- Andersson P.-O.,
- Tygesen H.,
- Eggers K.M.,
- Holzmann M.J.
- Thygesen K.,
- Alpert J.S.,
- Jaffe A.S.,
- et al.,
- Writing Group on behalf of the Joint ESC/ACCF/AHA/WHF Task Force for the Universal Definition of Myocardial Infarction
- Twerenbold R.,
- Boeddinghaus J.,
- Nestelberger T.,
- et al.