Study Urges Higher Cutoff for Troponin Elevation in Diagnosing MI

Download this article's Factoid in PDF (& PPT for Gold Subscribers)

A new study is raising questions about how to best diagnose periprocedural myocardial infarction (MI) after elective stent placement. To detect event frequencies and predict 1-year mortality risk at similar levels, the threshold for troponin elevation should be much higher than for creatine kinase-MB (CK-MB), according to a paper published online February 27, 2012, ahead of print in the Archives of Internal Medicine.

Currently, the universal definition of MI sets the bar at 3 times the 99th percentile of the upper reference limit, or upper limit of normal (ULN) for both biomarkers.

Donald E. Cutlip, MD, of Beth Israel Deaconess Medical Center (Boston, MA), and colleagues studied 4,930 patients who underwent elective coronary stenting between July 2004 and September 2007 as part of the Evaluation of Drug-Eluting Stents and Ischemic Events Registry (EVENT). Study protocol mandated routine biomarker assessment before and after treatment, though individual centers used a variety of assays and did not standardize their results via a core lab.

Different Cutoffs, Disparate Outcomes

When the cutoff was 3 times the ULN, substantially more patients had periprocedural MI by the troponin vs. CK-MB criteria. Mortality risk was increased at 1 year with both biomarkers using this threshold, but the risk was greater for CK-MB than for troponin. But when the cutoff for troponin was 20 times the ULN, the biomarker detected a frequency of periprocedural MI and a 1-year mortality risk that matched those seen using the standard CK-MB criteria (table 1).

Table 1. Effect of Biomarker Thresholds


Prevalence of Periprocedural MI

Adjusted HR (95% CI) for 1-Year Mortality

3 x ULN
Troponin (I or T)


2.5 (1.5-4.1)
1.7 (1.1-2.5)

20 x ULN
Troponin (I or T)


2.6 (1.6-4.3)

When the 3 times ULN definitions were used, 74.3% of those with troponin MI lacked a similar rise in CK-MB. Conversely, 13.7% of those with an MI based on CK-MB values did not meet the same cutoff for troponin. Only 6.2% of patients surpassed the 3 times ULN cutoff for both biomarkers.

Not a Choice Between Biomarkers

“The present results indicate that either CK-MB or troponin may be used as a predictor of increased late mortality, but the degree of elevation that is discriminatory for a similar absolute increase in mortality differs substantially for the 2 biomarkers,” the study authors conclude. “Given the move toward troponin and away from CK-MB as a standard method by most clinical laboratories and the continued importance of a universal definition, it is critical that revisions of these criteria incorporate the differences between these biomarkers in diagnosis and clinical implications.”

But an editorial accompanying the paper by Joseph S. Alpert, MD, of the University of Arizona (Tucson, AZ), and Allan S. Jaffe, MD, of the Mayo Clinic (Rochester, MN), took away a different message. “The authors imply that CK-MB is more useful because of the higher hazard ratio,” they say. “We disagree.”

For one, the study’s follow-up duration “may have been inadequate to determine potential negative effects of smaller amounts of myocardial injury detected by troponin,” Drs. Alpert and Jaffe assert. An even larger problem, they add, is that the investigators “assumed that baseline troponin values were normal before the procedures.”

Until efforts are made to ensure that baseline troponin levels are within the 99th percentile of the ULN, “studies such as this one and others that ignore this critical issue will continue to confuse the field,” they conclude.

Missing the Message

In a telephone interview with TCTMD, Dr. Cutlip expressed frustration with the editorial. “They sort of missed our message, unfortunately,” he said, noting that both biomarkers are useful and the question is only a matter of the proper thresholds.

Stephen G. Ellis, MD, of the Cleveland Clinic (Cleveland, OH), agreed that the paper’s conclusion is “quite correct,” in not supporting 1 biomarker over the other but rather saying that a higher cutoff is needed.

“The key thing here to stress is that 3 times CK-MB and 3 times troponin after PCI have very different prognostic meanings, hopefully to be reflected in new universal definition guidelines due out next year,” he told TCTMD in an e-mail communication, adding that the “[National Cardiovascular Disease Registry] and other reporting bodies should translate this into their reports ASAP.”

Although Dr. Cutlip did not know the details of the new definition, he said physicians are curious to see how it shapes up. “What I gather from the commentary is that the threshold [for troponin] is going to increase, and that they’re going to require some more clinical evidence of an event [beyond] PCI. I’m not sure this is helpful because most of these patients don’t have clinical symptoms and I think it is still important that we capture those events that might be significant.”

More Relevant to Trial Design Than Patient Care

When considering trial design, Dr. Cutlip said, “[m]y concern is that [the 3 times ULN troponin threshold] doesn’t allow us to discriminate between devices or between therapies that have different effects,” he noted.

Similarly, Dr. Ellis said that “low-level troponin elevation less than 20 times does not have much adverse prognostic importance in this setting and shouldn't be used to define MI after PCI. I think most contemporary DES trials are still using CK-MB.”

Though he agreed with the paper’s overall conclusion, Sorin J. Brener, MD, of Weill Cornell Medical College (New York, NY), told TCTMD in a telephone interview, “The problem is that this 20 was kind of picked out of a hat [based on where the mortality curve started to straighten]. It makes sense, but it’s still an arbitrary cutoff point as is the 3 times for CK-MB, which was picked 20 years ago and become the standard.”

Dr. Brener reported that, in terms of clinical practice, routine biomarker testing has a class 2a recommendation. “I think it is good practice, . . .  but it’s not mandatory,” he said, adding that it is also unclear what the consequences of an enzyme elevation should be in asymptomatic patients. “It’s just nice to know. . . . Some people would say, you should know because you should do more aggressive risk factor modification. But in fact you should do [this] in everybody.”

Study Details

In the EVENT registry, periprocedural MI was diagnosed according to peak biomarker levels measured between 6 and 24 hours after PCI.


1. Novack V, Pencina M, Cohen DJ, et al. Troponin criteria for myocardial infarction after percutaneous coronary intervention. Arch Intern Med. 2012;Epub ahead of print.

2. Alpert JS, Jaffe AS. Interpreting biomarkers during percutaneous coronary intervention: The need to reevaluate our approach. Arch Intern Med. 2012;Epub ahead of print.



Related Stories:

  • Funding for EVENT and the current analysis was provided by grants from Millennium Pharmaceuticals and Schering-Plough.
  • Dr. Cutlip reports serving as a consultant to and receiving academic salary support from the Harvard Clinical Research Institute as well as receiving research support paid to his institution from Medtronic.
  • Drs. Alpert, Jaffe, and Brener report no relevant conflicts of interest.
  • Dr. Ellis reports serving as a consultant to Abbott Vascular and Boston Scientific.