Letter by Stiermaier et al Regarding Article, "Dynamic Edematous Response of the Human Heart to Myocardial Infarction: Implications for Assessing Myocardial Area at Risk and Salvage"

Thomas Stiermaier, Jan Bogaert, Ingo Eitel


We have read with great interest the article by Fernández-Jiménez et al1 regarding the dynamic edematous response to myocardial infarction (MI). In this study, the authors expand their previous preclinical findings and report a bimodal edema pattern in 16 patients with anterior ST-segment elevation MI. An initial wave of edema directly after reperfusion almost resolved at 24 hours and was followed by a deferred wave peaking between days 4 and 7 postreperfusion.1 These findings provide valuable insight into the dynamic process and pathophysiological understanding of ischemia and reperfusion injury. In addition, the implications of these results for clinical research are of major interest. Cardiac magnetic resonance (CMR) parameters of myocardial injury are appealing end points in clinical studies of MI because they reduce the required sample size. Consequently, quantification of the edematous area as a marker of myocardium at risk, mostly using T2-weighted short-tau, triple inversion-recovery sequences, and calculation of myocardial salvage were performed in numerous previous trials.2–4 Based on the assumption of a stable edematous reaction, scans were performed on various time-points within the first week after MI. Post-hoc analyses of these studies did not reveal a dynamic course of the extent of myocardial edema.2,3 In contrast to the present study these investigations did not comprise serial CMR measurements in the same patients, but single scans at different times after MI. However, serial CMR scans ≈9 hours, 3 days, and 10 days after MI were performed in another previous clinical study which did not find a bimodal edema pattern either.5

So what are potential explanations for these divergent findings? Fernández-Jiménez et al1 designed a specific study to evaluate edema which included only patients with anterior MI and totally occluded coronary arteries whereas the other study cohorts were more heterogenous with various infarct regions and residual coronary flow. Of note, also in the highly selected patient population in the present study, a bimodal edema was not observed in all participants.1 About one-quarter showed only a slightly decreased edematous reaction at 24 hours and edema increased in 1 patient (online-only Data Supplement Figures III and IV from Fernández-Jiménez et al1). Therefore, additional factors which remain to be elucidated might significantly impact the dynamics of edema following MI. Furthermore, CMR scans are hardly performed within 3 days of MI in clinical routine given the recommended continuous monitoring for at least 48 hours. Therefore, the early edema peak is not reflected in these patient cohorts. In fact, the vast majority of CMR scans in previous clinical studies were performed exactly in the timeframe recommended by the authors, between days 4 and 7, with only ≈10% being performed prior or thereafter.2 Moreover, in our personal experience of analyzing several large-scale multicenter CMR studies, we did not observe negative salvage values, which supports the validity of previously reported data, in addition to, randomized study designs and similar timing of CMR imaging in treatment groups. Nevertheless, we agree with the authors that standardized image acquisition protocols and more robust sequences to delineate myocardial edema are required to enhance the quality and validity of future MI trials.
Original languageEnglish
Issue number16
Pages (from-to)1752-1753
Number of pages2
Publication statusPublished - 17.04.2018

Research Areas and Centers

  • Academic Focus: Center for Brain, Behavior and Metabolism (CBBM)


Dive into the research topics of 'Letter by Stiermaier et al Regarding Article, "Dynamic Edematous Response of the Human Heart to Myocardial Infarction: Implications for Assessing Myocardial Area at Risk and Salvage"'. Together they form a unique fingerprint.

Cite this