Background— It is known that a significant number of patients experiencing an acute myocardial infarction have normal coronary arteries or nonsignificant coronary disease at coronary angiography (CA). Computed tomography coronary angiography (CTCA) can identify the presence of plaques, even in the absence of significant coronary stenosis. This study evaluated the role of 64-slice CTCA in detecting and characterizing coronary atherosclerosis in these patients. Methods and Results— Consecutive patients with documented acute myocardial infarction but without significant coronary stenosis at CA underwent late gadolinium-enhanced magnetic resonance and CTCA. Only the 50 patients with an area of myocardial infarction identified by late gadolinium-enhanced magnetic resonance were included in the study. All of the coronary segments were assessed for the presence of plaques. CTCA identified 101 plaques against the 41 identified by CA: 61 (60.4%) located in infarct-related arteries (IRAs) and 40 (39.6%) in non-IRAs. In the IRAs, 22 plaques were noncalcified, 17 mixed, and 22 calcified; in the non-IRAs, 5 plaques were noncalcified, 8 mixed, and 27 calcified ( P =0.005). Mean plaque area was greater in the IRAs than in the non-IRAs (6.1±5.4 mm 2 versus 4.2±2.1 mm 2 ; P =0.03); there was no significant difference in mean percentage stenosis (33.5%±14.6 versus 31.7%±12.2; P =0.59), but the mean remodeling index was significantly different (1.25±0.41 versus 1.08±0.21; P =0.01). Conclusions— CTCA detects coronary plaques in nonstenotic coronary arteries that are underestimated by CA, and identifies a different distribution of plaque types in IRAs and non-IRAs. It may therefore be valuable for diagnosing coronary atherosclerosis in acute myocardial infarction patients without significant coronary stenosis.

The antiplatelet drug clopidogrel is a commonly prescribed therapy in patients with acute coronary syndrome. However, its clinical efficacy is hampered by a wide inter-patient response variability, with over 30% of patients treated with this drug experiencing an inadequate antiplatelet response. There are growing evidences that clopidogrel response variability is associated with cytochrome P450 (CYP) enzyme genetic polymorphisms, primarily CYP2C19 which is responsible for the conversion of clopidogrel into its active metabolite. All of the CYP2C19 polymorphism data suggest that carriers of allele *2 or *17 are at greater risk of ischemic or bleeding events, particularly in patients with acute coronary syndrome undergoing percutaneous coronary intervention. Yet, CYP2C19 status explains only 12% of clopidogrel response variability, indicating that genetic variants other than CYP2C19 might be important. Clopidogrel undergoes intestinal efflux via P-glycoprotein, encoded by the ABCB1 gene. The C3435T polymorphism in this gene affects the bioavailability of clopidogrel, however, its effects on clinical outcomes are inconclusive. Similarly, a polymorphism in the gene encoding PON1, a rate-limiting enzyme for clopidogrel bioactivation, also affects the response to clopidogrel. Among nongenetic factors, an adverse drug interaction between proton pump inhibitors and clopidogrel is often reported, but evidence is inconclusive. A genetic test to identify potential responders to clopidogrel might be useful. However, the use of such tests is currently limited because they focus mainly on CYP2C19 loss-of-function alleles, and there is no empirical evidence yet for genotype-guided clopidogrel therapy.

We thank Opolski et al for their comments and appreciate their interest in our study.1 They raise several points, which we address below. As the authors note, coronary atherosclerosis is a common finding also in asymptomatic patients; indeed, we think that, in the special setting of patients with acute myocardial infarction proven by cardiac magnetic resonance imaging and without significant coronary stenosis, coronary atherosclerosis is often underestimated because of an apparently normal coronary angiogram. As a consequence, the diagnosis may be challenged, and patients may not receive an adequate secondary prevention therapy.The mechanisms of myocardial infarction in this setting are certainly multiple and are not related exclusively to atherosclerosis, but we think that the different distributions of coronary plaque patterns between infarct-related arteries (IRAs) and non-IRAs support that coronary atherosclerosis may be the most likely mechanism in a number of patients.As the authors suggest, a control group could be useful to support the data, but we do not think that patients excluded from the study could represent an appropriate control group; they cannot be classified clearly, because they are symptomatic patients with high troponin but unclear diagnosis.The 64-slice computed tomography technology has a limited spatial resolution compared with intravascular ultrasound or optical coherence tomography, but several studies have shown that plaque area measurements are accurate compared with intravascular ultrasound.2 As the authors correctly note, in our study, coronary plaques were located more frequently in the left anterior descending coronary artery, but we do not think that this finding is responsible for the larger plaque areas in IRAs. In fact, it is well known that coronary plaques cluster in the proximal segments, 3 but coronary arteries caliper in the proximal segments do not differ significantly, except for the left main vessel. Therefore, the higher plaque area and remodeling index are not related exclusively to location. In our analysis, we took into account only coronary arteries with coronary plaques.Previous studies have shown that coronary computed tomography is a reliable noninvasive method to analyze coronary plaque composition, but it is not able to visualize plaque disruption. Therefore, it cannot identify culprit plaques in acute coronary syndrome. Taking into account this limitation, our aim was to investigate whether coronary plaques located on IRAs showed a different morphological pattern compared with non-IRA plaques. We think that this type of analysis can provide some important clues. In fact, our data highlight an increased burden of plaques with vulnerable features in IRAs. Of course, the analysis includes both culprit and nonculprit plaques located on IRAs, but in our opinion, this fact reinforces the results because the presence of nonculprit plaques probably mitigates the differences between IRA and non-IRA plaques.