CAP TODAY Pathology/Laboratory Medicine/Laboratory Management
Some gaps have occurred. “We used an absolute delta of 10. And as you start getting values that are higher and higher and higher,” Dr. Jaffe says, “that’s just too low. So we would argue for changing that, and simply say that when the value is over 100, start using a percentage criteria of, say, 20 percent.”
A second gap—“We alerted people, but obviously we didn’t educate adequately,” Dr. Jaffe says—concerned patients who present late after the onset of their MIs. “We guessed 12 hours, but that may not always be the ideal number,” Dr. Jaffe says. “They’re on the downslope of the time-concentration curve for troponin, which is much slower than the upslope.” As result, it’s easy to miss a delta.
In fact, he continues, “I have a whole collection of such cases,” where the MI was missed and elevated troponin values were instead attributed to ischemic heart disease.
Finally, Dr. Jaffe says, physicians need to be sensitized to the fact that “there are many more type two MIs now. Using sex-specific cutoffs, we’re finding more modest, or small, MIs, particularly in women.” Some, though not all, fit into the category of myocardial infarction with nonobstructed coronary arteries, a constellation of microvascular dysfunction, some of which can be due to unseen plaque ruptures. “This is an increasing commonality that we need to be careful about.”
The major lab-related issue at Mayo, says Dr. Karon, involves analytical outliers. “We have made somewhat of a hobby out of studying them,” he says. Data so far show the frequency of analytical outliers is higher with the 5th Gen than with the 4th Gen.
“The issue is you’re making decisions on much smaller changes in troponin concentration,” he says. That could give rise to analytical outliers due to fibrin strands or activated platelets, for example, or centrifugation conditions or instrument-to-instrument/platform-to-platform differences. (Mayo has a benchtop immunoassay analyzer in the stat lab and large, automated equipment in the core lab.) “You need to worry about those more, because now we’re saying a change, or at least an indeterminate delta, is 4 to 9 ng/L. So if a change is over 3 ng/L over two hours, we’re saying potentially that it’s significant. It’s called indeterminate in our practice.”
Yet he fully recognizes that this represents a very tiny change in concentration. “We’re asking this assay to do more when we’re using it for a panel and trending over time,” Dr. Karon says. “So the two issues we’ve found to be more significant, and we spend more time on, is detecting our analytical outliers and looking at how our different instruments and platforms agree with each other.”
Dr. Jaffe calls these fliers. “We had a little more than three percent nonrepeatables, which is high. These assays are highly sensitive. And consequently, little things can cause bigger signals than we’re used to seeing.”
Dr. Apple, who is familiar with Mayo’s data on outliers, says instrument/platform differences can be vexing. A three-hospital system may have three different instruments, which can present different results. And yet clinicians can’t be expected to know why. “It’s tough to educate clinicians,” Dr. Apple says, “because they don’t give a rip if it’s instrument A, B, or C. They just want a reliable number.”
Preanalytical issues like hemolysis and biotin may present challenges when reporting results, Dr. Saenger says. If the hemolysis threshold is fairly low and/or subject to visual interpretation, then hs-cTn results might be reported as falsely low or falsely high. “This would obviously affect interpretation of a single hs-cTn result,” she says, “but also could affect interpretation of serial results.” Some commercially available troponin assays are sensitive to biotin interference at fairly low biotin concentrations, Dr. Karon says, “though there is not great evidence about how often patients are presenting with biotin levels that would pose a significant risk to interpretation of troponin values.” As with hemolysis, he adds, biotin effects will change over time, potentially confounding interpretation of troponin panels.
The IFCC committee on clinical applications of cardiac biomarkers routinely updates several tables on its website (http://j.mp/2qKSK25) with analytical information related to all hs-cTn, contemporary, and POC troponin assays, Dr. Saenger says; one of these tables lists analytical specificity information for each troponin assay related to interference from hemolysis and biotin.
The various hs-cTn assays also have their own differences. Dr. Apple is principal investigator of the CONTRAST study, a comparison of the Roche T and Abbott I hs-cTn assays involving about 2,000 participants. “There are considerable differences between positives and negatives between the two assays,” he says. “There’s a subgroup of maybe 10, 15 percent of patients who don’t match up at all.” It’s unclear why.
But one thing is clear already, he says: Switching from one assay to another “will have to be looked at very carefully if that’s what you decide to do, as troponin assays are not standardized.”
He’s also finalizing another clinical study, called Scorecard, which involves 1,000 emergency department patients from three cohort sites—Scotland, Mayo, and Hennepin—and is looking at measurement of only baseline samples and 30-day outcomes. “We’re just starting to analyze the data to see what percent of patients would be able to be safely discharged based on each assay,” Dr. Apple says.
“As more people start studying troponin, we’re going to see a lot more good data coming out that will solidify within 12 months in the U.S.,” he predicts. “That literature will become more robust with evidence-based information.”
In the meantime, Drs. Apple and Jaffe helped write an article that serves as a starting point for labs making the switch to hs-cTn (Wu AHB, et al. Clin Chem. 2018;64[4]:645–655). In addition, Dr. Apple urges labs to do their due diligence when deciding on an upper reference limit (URL), since they vary based on population. He sounds a cautionary note: A URL based on a data set from a package insert may or may not take into account exclusions due to, say, surrogate biomarkers, such as increased hemoglobin A1C, elevated natriuretic peptides, or statin use, which would eliminate silent pathophysiologies and lower sex-specific URLs.
Looking back over Mayo’s experience, Dr. Karon suggests that time is the laboratory’s best friend. “Work toward implementation well ahead of when you say you’re going to bring this up,” he says. “Analytical validation took a lot longer than even I expected it would.” The protocols were developed over the fall of 2017, and the implementation group kicked into high gear in January/February 2018.It also took time to get the IT right and to train the lab staff. The clinician orders a zero- and two-hour value, and the LIS calculates whether a six-hour sample (ordered reflexively) is needed. That six-hour order appears from nowhere, or so it seemed to lab staff. “It’s probably one of the more complicated, and maybe the most complicated, sort of reflex lab protocol to support,” says Dr. Karon. “So it does take some efforts from our phlebotomists, our laboratories to work together to do this right. It’s fairly well automated now, but it can be challenging to support. There’s always weird exceptions that happen—you know, how did somebody get a six-hour sample collected before their two-hour?”
“The biggest impediment we had was trying to get our IT systems to accommodate something new,” Dr. Jaffe agrees.
IT issues and clinical buy-in are also influencing the pace of the rollout at Hennepin, Dr. Apple says. “It takes time to change order sets in LIS/Epic, and it takes time to meet with all the different clinical divisions to discuss how practice patterns may change based on a new assay.” The assay is hardly new in Hennepin’s lab. But, he says, “As much as we know about hs-cTnI, we need to gain the confidence of all the providers—clinicians, nursing, et cetera—of what the new world of a new troponin assay will entail.”
Hennepin has started working on the rollout. A three-person team that includes Dr. Apple, Dr. Saenger, and Stephen Smith, MD, an emergency medicine colleague, is leading the launch. Educational discussions are being scheduled, similar to the Mayo pathway. It will require “education, sometimes a lot of education, and certainly ongoing education,” says Dr. Saenger.
“We’re going to essentially put together a teaching deck based on our own experiences and the literature,” Dr. Apple says. The group will meet with emergency medicine physicians, cardiologists, surgeons, and hospitalists and talk about how the assay will be implemented. Even with all the experience, it’s not a matter of flipping the switch the lab’s finger has been hovering over for six years. (To be fair, even Riccardo Muti occasionally glances at his Verdi score.)
The primary teaching, says Dr. Apple, is “not to be fearful that there’s going to be a lot more positives.”
If everything goes according to plan, Hennepin will likely roll out the new assay in the first quarter of 2020, using a zero-/two-hour algorithm, “with an additional six-hour draw for patients you’re not sure about,” Dr. Apple says.
A zero-/one-hour protocol was tempting, but—like Roger Maris’ home run record—its use carries an asterisk. The algorithms for early presenters are particularly problematic, Dr. Apple says. “So we don’t want to miss an early presenter with a potential worse sensitivity, or negative predictive value. And to be honest, it’s going to be easier to draw a blood sample and not miss the time window at zero-/two-hours.”
Once hs-cTnI is in place, Dr. Apple says, “We hope to be able to discharge, conservatively, 20 percent of patients based on their zero-hour and zero-/two-hour measurements—a substantial financial savings to the hospital.”
Nevertheless, physicians are nervous, he acknowledges. “I think one fear factor with clinicians is they have [the] misconception that they’re going to have a hundred more consults a day because of additional increases in troponin,” he says. He downplays the scope while acknowledging there will indeed be at least some increase in positive findings, especially in females. Even if it’s not an indication of an MI, the elevation would be important as a marker of myocardial injury with an underlying need for risk assessment. “It’s a flag that’s being waved,” he says. Perhaps an outpatient cardiology consult is in order—a patient with a primary neurological problem, for example, and an elevated troponin is at increased risk for an adverse event. Sorting through this is going to be a learning curve for providers but significantly better for patient care, he says.
Fears about a soaring number of abnormals, especially for cardiac troponin T, has led many institutions to implement a cutoff above the 99th percentile, Dr. Saenger reports. While many clinicians think these are false-positives, she says, the hs-cTn assays actually are detecting subtle but relevant changes. “It just takes some time for them to feel comfortable interpreting these changes and determining appropriate treatment.” That being said, however, it can be difficult to identify a relevant serial change for troponin, given that each assay is different and serial changes can vary due to timing of collection protocols, early versus late patient presentation to the ED, and comorbidities, Dr. Saenger notes.
Has Dr. Apple faced pushback, even despite his years of advocacy at Hennepin? “Let’s say I have 10 cardiologists in a room,” he says. “There’s always going to be one or two who fight it.” Indeed, there are still cardiologists carrying the Lost Cause banner for CK-MB, he says. In the case of hs-cTn, however, he expects reluctance to fade in a matter of weeks once it’s implemented. “It’s awesome,” he reiterates. There may be more questions about what to do with results, but he doesn’t expect people to push back against having those results. “I think clinicians will grow to like it,” he predicts. “And there will be less worry once it’s in place six months to a year.”
The ED is excited, Dr. Apple says. So are cardiologists who understand the field of troponin. The ones who haven’t paid attention, he says, who dismiss the need for biomarkers and cling to the older version, are less thrilled.
It sounds as if he barely cares. “I embrace it,” Dr. Apple says. “It’s the next step in better patient care.”
Dr. Apple takes it one step further, in fact. “I’m hoping manufacturers will stop producing those old assays,” he says. Alluding to an old Seinfeld episode, he compares conventional and hs-cTn assays to cinnamon and chocolate babkas, respectively. “There are people out there who love cinnamon,” he acknowledges. But compared with the chocolate versions, “It is an inferior babka.”
Karen Titus is CAP TODAY contributing editor and co-managing editor.
High-sensitivity cardiac troponin in the outpatient setting
November 2019—At least for now, Mayo Clinic does not offer the panel in the outpatient setting. While there is evidence that anything above the upper reference limit, or even a rising level within the URL, on an outpatient implies greater risk, “there’s not a lot of information on what you do about it,” says Dr. Karon.
Might there be a role for individual test orders? The data to support such use is strong, Dr. Jaffe says, and single orders will likely grow. Nevertheless, next steps are unclear. “Some clinicians are reluctant to go there, because they say, ‘I don’t know what to do.’ Others say, ‘If a person has an elevated troponin, I ought to know about it.’” Perhaps that information can be useful if the patient goes to the ED, goes the thinking, or can be integrated into the patient history/physical exam to help prevent disease.
As for himself, “I use it very commonly in patients with atrial fibrillation,” Dr. Jaffe says. “I think eventually we’ll be using this a lot in the outpatient setting.”
Dr. Saenger points to cardio-oncology data on the use of high-sensitivity cardiac troponins to evaluate cardiotoxicity from chemotherapeutic agents. She also foresees a time in the (much more standardized, much more harmonized) future when hs-cTn could be used over long periods of time to monitor health, similar to glucose/HbA1c, creatinine/eGFR, or lipid testing, for example.
One of the biggest discussions going on today, says Dr. Apple, is in noncardiac surgery: Should baseline and post-op troponins be measured? “There’s a great wealth of data showing that patients are at risk—even without an MI—when they have post-op elevations,” he says. The big question, of course, is medicine’s evergreen query: What do you then do with these patients? “Because not every pathophysiology has a treatment mode.” There is, Dr. Apple says cheerfully, “only one way we’re going to learn—we kind of live it.”—Karen Titus