Cardiac troponin T (cTn) has earned a solid place as a marker of myocardial necrosis and for diagnosis and risk assessment in suspected acute coronary syndrome (ACS). Research also has shown that cTn predicts risk of heart disease and death in certain populations. Given cTn's high specificity for insults to myocardial tissue, some clinicians have proposed monitoring cTn levels after vascular surgery to better assess risk of short-term mortality. Little is known, however, about what, if any, role this analyte might have in risk prediction following non-cardiac surgery. Now, a multinational consortium of researchers has evaluated this question, and their findings are the subject of this issue of Strategies.
The Vascular Events in Noncardiac Surgery Patients Cohort Evaluation (VISION) study is an ongoing prospective cohort study that expects to recruit from eight countries 40,000 subjects undergoing noncardiac surgery. VISION is examining major complications after non-cardiac surgery, including determining whether there is a relationship between peak post-operative cTn level and 30-day mortality. Recently, VISION investigators reported findings from a phase of the trial involving 15,133 surgery patients, all 45 years or older and with at least an overnight hospital stay, who had cTn levels measured with a fourth-generation assay commonly used in the U.S. (JAMA 2012; 307:2295-2304). The study protocol called for cTn measurement 6-12 hours post-operatively and on days 1, 2, and 3 following surgery.
The researchers found peak post-operative cTn measurement during the first 3 days after surgery to be significantly associated with 30-day mortality. Risk of mortality rose with rising peak cTn measurement, so that in comparison to the reference group with peak values ≤0.01 ng/mL at the 99th percentile, patients with peak values of 0.02 ng/mL had an adjusted hazard ratio of 2.41. Adjusted hazard ratios rose with peak cTn values, so that patients with peak cTn levels of 0.03-0.29 ng/mL had an adjusted hazard ratio of 5, while those with cTn values ≥0.30 ng/mL had a hazard ratio of 10.48. Likewise, 30-day mortality rates also rose with increasing peak cTn values, ranging from 1% when peak cTn values were ≤0.01 ng/mL to 16.9% when peak cTn values were ≥0.30 ng/mL. The researchers calculated that these elevated cTn levels, present in 11.6% of subjects, explained in excess of 40% of deaths in the study population.
Corresponding author for VISION's writing group, P.J. Devereaux, MD, PhD, emphasized that some of the peak cTn values associated with increased risk currently would be considered within normal limits by both professional practice guidelines and existing standards of care. "The lab and cardiovascular communities have taken the position that they define abnormal troponin as the 99th percentile of a healthy population with a coefficient of variation of 10 percent or less," he said. "An important aspect of our study is that values strongly predicting who was going to die in the short term after surgery don't fall into those thresholds. Patients who had troponin values of 0.03 or even 0.02 ng/mL, values most cardiologists and laboratorians wouldn't consider abnormal with fourth-generation troponin T assays, were at increased risk of death." Devereaux is an associate professor of clinical epidemiology and biostatistics and a joint member in the department of cardiology at McMaster University in Hamilton, Ontario.
Devereaux also pointed out that VISION's findings relate to short-term rather than long-term mortality risk. "What was really surprising was how strong the association was and how high the death rates were in such a short period of time. At the moment, people are making a lot of focus on how high-sensitivity troponin T can predict who's going to die in the next five years. But with the VISION data, we're talking about people who are going to die within the next couple of weeks, up to one-in-six who have troponin levels of at least 0. 30 ng/mL," he said.
Both Devereaux and Amy Saenger, PhD, DABCC, suggested the findings might prompt laboratorians to review their organizations' current ordering practices around cTn. "It's now of even greater interest to me to look within my own institution and evaluate how we're using troponin today in post-operative individuals who are generally thought to have a low pre-test probability of cardiac damage, if we are at all. We know we have a high frequency of stand-alone orders for troponin T outside the emergency department, so it peaks my interest in terms of how we're using it today in non-cardiac surgery patients," commented Saenger. "I think it would be of interest to laboratorians to critically assess utilization patterns at their own institutions and determine if their physicians are even ordering troponin T outside of the emergency department, and if they are, specifically in which populations they're ordering it. They could be surprised at the frequency and physician interpretation of troponin in those clinical scenarios." Saenger, who was not involved in VISION, is director of cardiovascular laboratory medicine in the department of Laboratory Medicine and Pathology at Mayo Clinic in Rochester, Minn.
In addition to cTn perhaps not routinely being ordered in patients after non-cardiac surgery, Devereaux also suggested that cardiac events might be missed in this population because patients in the immediate post-operative period might not have classic symptoms of ACS or other heart conditions. "The vast majority of these elevated troponins would have been missed if physicians were waiting for symptoms to first order the troponins. Most patients who experience myocardial injury after surgery don't experience ischemic symptoms, either because they're receiving such effective analgesic therapy that they don't experience chest pain or they're getting suboptimal post-operative analgesic and they're so distracted by their hip or knee or bowel pain that that's all they're focused on."
Since the change in serial values is crucial in the most common use of cTn, evaluating patients with suspected ACS, Saenger wondered why the VISION authors did not report the mortality risk associated with a change value. Devereaux explained that the investigators "did look at change but it didn't end up being important. The reason I think it didn't is that we're already identifying that low levels of troponin are predicting death. So to a certain extent that's why change didn't show up being important, because these small values already predicted death," he said. This conclusion could change in the next phase of VISION, which involves a fifth-generation high-sensitivity cTn assay that has not yet been cleared by the U.S. Food and Drug Administration for commercial use in the U.S., according to Devereaux. "I suspect what we'll end up showing with the fifth-generation assay is that there are some thresholds above which there's no doubt the patient's having an event. You'd be predicting mortality but then there would also be an indeterminate range whereby change in the pre-operative value would be important in identifying who's having an event that's likely important versus not important," he added. Saenger noted that defining what that significant change is will also be critical for the high-sensitivity assays and may differ depending on the nature of the clinical post-operative situation, such as cardiac or non-cardiac surgery.
Devereaux also suggested that even without further evidence from the next phase of VISION, providers might consider changing their use of and response to elevated values of cTn in post-operative patients. "The litmus test for people to ask themselves is, if, after seeing this data, if your mother or father was about to have hip replacement or bowel surgery and had known coronary artery disease or risk factors for it, would you want your loved one's troponin measured?" he said. "Obviously we need more research to clarify how to modify that risk, but even now there are some obvious and intuitive things we can do. For example, as we've shown in prior research, at least in patients who have elevated troponins after surgery, those who end up receiving aspirin and statins have a risk-adjusted lower mortality rate. Yet a high proportion—25 percent in the case of aspirin and almost 50 percent in the case of statins—don't receive these interventions. These are simple things we can do now to mitigate risk for these patients."