We read with interest the article by Aldrovandi et al 1 evaluating the characteristics of coronary lesions using coronary computed tomography angiography (CCTA) in patients with acute myocardial infarction (MI) and angiographically nonobstructive coronary artery disease. We congratulate the authors on their elegant study fitting comfortably with the vulnerable plaque paradigm. However, in our opinion, the main findings failed to take into account important confounders.First, the identification of an increased burden of angiographically underestimated coronary plaques by means of CCTA is to support the atherosclerotic pathophysiology of unexplained MI. However, the detection of angiographically invisible coronary plaques is a common finding and has been reported previously by intravascular ultrasound studies in asymptomatic adult populations.2,3 Thus, only when a direct comparison with regard to CCTA-detected plaque burden between patients with versus without MI and nonobstructive coronary disease is available will the statistical inference be sound enough to regard atherosclerosis as a causative or even contributing factor of angiographically silent MI. Indeed, we believe that the CCTA analysis of the 21 patients excluded from the study as a result of the absence of late enhancement at magnetic resonance would be desirable and should serve at least as a control comparison.A second aspect is the CCTA rationale for identification of vulnerable plaques that were more often distributed in infarct-related arteries (IRAs) versus non-IRAs. Nonetheless, it should be noted that the ability of 64-detector computed tomography scanners to perform precise plaque quantification is limited by its spatial resolution and reduced agreement with in vivo plaque dimensions, especially in case of mild atherosclerotic lesions. 4 Furthermore, taking into consideration a substantially higher prevalence of left anterior descending IRAs (together with the left main) compared with nonIRAs (28 versus 14 vessels) together with the estimated ratio of coronary plaques per IRAs versus non-IRAs of 3:1 (61/42=1.45 versus 40/84=0.48), we assume that there should be a significant predominance of coronary lesions located in the left anterior descending IRAs versus non-IRAs (41 versus 7 plaques)-a phenomenon that might be responsible for the reported increase in plaque area and vessel remodeling in large-caliber IRAs. Thus, additional commentary on plaque distribution with respect to both coronary vessel type and its relationship to infarct territory is mandatory. We are also curious whether the authors encountered any IRAs without coronary plaques.Finally, a previous report 5 has shown that culprit plaques in IRAs are distinctly different from nonculprit plaques in both IRAs and non-IRAs, suggesting that a vascular event in MI is determined by a vulnerable lesion rather than vulnerable IRAs. Thus, because the CCTA analysis of IRAs in the study by Aldrovandi et al 1 is derived from different plaque morphologies (both culprit and nonculprit), there is...