Against Another Nonspecific Marker of Perfusion

To the Editor:

We read with interest the article by Frencken and colleagues and the accompanying editorial by Bonk and Meyer regarding the use of high-sensitivity troponin (hs-Tn) in pneumonia (1, 2). We applaud the authors on their publication of useful manuscripts regarding this emerging topic.

Elevated troponin levels are found in nearly half of critically ill patients, using standard troponin assays (3). Thus, it comes as little surprise that 85% of critically ill patients with pneumonia would have an elevated hs-Tn level. The challenge facing clinicians has to do with how to use these data. The pathophysiology of troponin elevation in this context is multifactorial (e.g., including inflammatory injury to myocytes, as well as myocardial oxygen supply–demand mismatch). Troponin elevations generally do not reflect acute coronary occlusion or stenosis. Rather, troponin elevation in this context functions largely as a marker of mortality (3).

These articles are important for promoting awareness of the frequency of troponin elevation in critically ill patients. All too often, such elevations are misinterpreted as evidence of coronary artery disease, leading to inappropriate use of anticoagulation and cardiac catheterization. This potential cascade of downstream testing and procedures that may result from the widespread application of hs-Tn suggests that we should exercise restraint in obtaining this test.

One conceivably rational use of troponin in the context of a severely ill patient with pneumonia could be as a disease severity marker to facilitate risk stratification. For example, patients with a troponin above a certain level are at increased risk for death, and therefore might potentially benefit from more intensive care. However, we already have validated risk-stratification tools to determine which patients require more intensive care, such as the American Thoracic Society criteria. Furthermore, Frencken found that hs-Tn was less specific as a mortality indicator compared with standard troponin assays. Thus, it is doubtful that hs-Tn could add independent and useful information beyond available risk-stratification tools.

Bonk and Meyer opined that troponin might be used as a perfusion target for resuscitation, perhaps based on the finding by Frencken and colleagues that a downward trajectory of hs-Tn was associated with lower mortality compared with persistent elevation (1). We caution against this approach for many reasons. The mechanism of elevated troponin in these patients is complex, multifactorial, and not necessarily closely related to perfusion. Furthermore, troponin can be elevated by a diverse range of pathologies (e.g., pulmonary embolism, chronic kidney disease, and heart failure) (4). With an extensive list of possible mechanisms and etiologies that may often coexist, it is unclear how this single laboratory test could specifically assess perfusion. If troponin were related to myocardial oxygen supply–demand mismatch, how would we change our approach from the default (i.e., treating the underlying cause)? And importantly, how many patients might suffer from the iatrogenic effects of additional interventions?

Wide application of hs-Tn to assess perfusion in the critically ill would be, at best, another blunt instrument among many unhelpful tools in guiding patient management. Consider the current state of assessing and treating serum lactate, widely practiced because of Surviving Sepsis Guidelines. Evidence supporting this practice is lacking, with a recent study suggesting that lactate was no more effective at gauging perfusion than capillary refill time (5). Given the current state of evidence, we advocate for targeted use of troponin testing for the evaluation and management of suspected myocardial ischemia based on history, physical exam, point-of-care echocardiography, and electrocardiogram findings only.