Contrary to our hypothesis and previous study findings [19], we found that the majority of our patients (80 %), who were diagnosed with AMI and lacked visible atherosclerosis on a routine invasive coronary angiography, had totally normal coronary arteries on CTCA as well. The remaining 20 % had only minimal, one-segment, non-obstructive atherosclerosis. The absence of atherosclerosis in the large majority of patients and the limited extent of atherosclerosis found in the remaining minority might suggest that other diagnoses than acute myocardial infarction are likely for most patients in this population. Echocardiography was available in a subgroup of patients and is known to provide sensitive and objective parameters for left ventricle systolic function such as GLS and WMSI that may help in arriving at a proper diagnosis. The addition of echocardiography to the CTCA findings enabled a probable diagnosis in the majority of the patients (17/22 in the echocardiography subgroup). For example, patients with atherosclerosis on CTCA and WMA on echocardiography could likely have had an acute myocardial infarction (based on plaque rupture) in the absence of other alternative explanations for their symptoms and elevated troponin. Echocardiography can additionally identify cases of Takotsubo cardiomyopathy (5/22 in the echocardiography subgroup). Abnormal echocardiography without any coronary atherosclerosis on CTCA indicates either myocarditis or myocardial infarction with a mechanism different from plaque rupture. In case of normal findings on CTCA and echocardiography, other factors such as mild myocarditis, arrhythmia, pulmonary embolism or renal failure (exclusion criterion in this study) seems more likely to explain the elevated troponins. Most patients in the study were prescribed acetylsalicylic acid and a great proportion of them were sent home with dual antiplatelet therapy, beta-blocker and statin, thus being treated as “classical” AMI patients. We do know however, that patients with AMI and insignificant coronary atherosclerosis have a lower incidence of adverse outcomes compared with patients with significant coronary artery disease [26]. Since these patients have a lower burden of risk factors, such extensive medication may seem inappropriate and potentially harmful.

When comparing with previous studies we found that the mean age of our study population, which was 60.2 years is in accordance with previously observed age for most first time AMI [27, 28]. The prevalence of risk factors was lower from what was observed in large epidemiological studies where the prevalence of diabetes, hypertension and hyperlipidemia was 18.5 %, 39.0 % and 90 % respectively [29]. In our small study the prevalence was 0 % for diabetes, 23 % for hypertension and 17 % for hyperlipidemia supporting a non-atherosclerotic etiology or a non AMI diagnosis in the majority. Most patients were women, 77 %, (23/30), which is in accordance with previous analyses where female gender was the strongest predictor of insignificant CAD in patients with NSTEMI [26]. The most frequent cause of AMI in similar female populations with non-obstructive CAD was previously found to be plaque rupture and ulceration [30]. The small percentage of patients having non-obstructive plaques on CTCA in our study compared with other studies [19, 20] is probably due to the more stringent criteria for inclusion requiring a totally normal angiogram whereas in previous studies patients with non-obstructive atherosclerosis up to 30–50 % on ICA were included. This study shows that if a normal coronary angiogram is found after a diagnosis of AMI, the CTCA will most probably also be normal or show only minimal atherosclerosis.

The theory of rupture of a non-obstructive plaque as an explanation for this population’s clinical presentation does not seem justified in 80 % of our population which had no signs of atherosclerosis at all (either on ICA or on CTCA). This fact leaves mechanisms like vasospasm, dissection, embolism and impaired coagulation and fibrinolysis more likely to be part of the pathogenesis of AMI. Non AMI diagnoses such as myocarditis, Takotsubo cardiomyopathy, pulmonary embolism and arrhythmias are also possible causes for this type of presentation. It is reasonable to consider using an intravascular imaging method [optical coherence tomography (OCT) or IVUS] to provide information about some mechanisms of coronary damage (i.e. plaque rupture, dissection or ulceration) in the case of AMI without coronary atherosclerosis but with wall motion abnormalities suggesting myocardial injury in a coronary artery territory or angiographic evidence of plaque rupture or impaired flow. Even if these patients fulfilled current diagnostic criteria for myocardial infarction, alternative explanations for chest pain and positive biomarkers may have been overlooked. In this clinical situation with leaking biomarkers and a normal coronary angiogram, cardiac MRI is of great importance for a definite diagnosis as it can differentiate between AMI, myocarditis and Takotsubo cardiomyopathy. It has been shown earlier that up to 50 % of patients with elevated troponin and unobstructed coronary arteries had myocarditis as diagnosed by cardiac MRI [31]. In another recent study [20] where 152 patients with MINCA underwent cardiac MRI, the findings were normal in two thirds of the patients, 7 % had signs of myocarditis and about 20 % signs of myocardial necrosis. Amongst patients with normal MRI, 32 % had typical clinical signs and symptoms of Takotsubo cardiomyopathy and reversible wall motion abnormalities. Interestingly, the initial diagnosis of AMI was changed in two thirds of the patients after the MRI examination. These findings suggest that a significant part of this population actually has other diagnoses than AMI and more extensive evaluation than just coronary angiography is needed.