The unique value of cardiovascular magnetic resonance in patients with suspected acute coronary syndrome and culprit-free coronary angiograms

Coronary artery disease (CAD) and especially acute coronary syndrome (ACS) are among the leading causes of mortality and morbidity. Detailed guidelines are available for the management and treatment of both acute myocardial infarction (AMI) with and without ST-segment elevations [1, 2]. In addition to several clinical, laboratory and electrocardiographic (ECG) parameters, coronary angiography is one of the basic examinations performed on these patients and in most cases, it provides invaluable information when deciding on the subsequent therapy. However, in several trials, coronary angiography without any visible stenoses/occlusions was found in 1–12% of all patients with ACS [3‐9]. In contrast to precise recommendations for the management of proven AMI, there are no guidelines for these culprit-free patients. This means that while the management and treatment of ACS with a clear culprit lesion are nearly identical in all centres, a very different approach is applied to a significant portion of other patients. The methods for clarifying diagnoses and the use of imaging modalities vary in different hospitals. Cardiovascular magnetic resonance (CMR) is a great example of a very useful tool that is used in the vast percentage of cases. While some centres are aware of the great contribution of CMR and its irreplaceability in many cases, other teams use it only rarely.

CMR is a robust non-invasive diagnostic tool in cardiology. It is a precise and highly reproducible technique used to assess ventricular volumes, masses and function. CMR can define cardiac anatomy and structure, quantify myocardial perfusion and measure blood flow. CMR images are acquired without using ionizing radiation or iodinated or radioactive contract agents. Oedema visualization and the late gadolinium enhancement (LGE) technique provide the unique opportunity to assess myocardial tissue in vivo – it enables tissue characterization in ischemic and non-ischemic cardiomyopathies and other cardiac diseases. All these advantages can help to refine the diagnostics of patients with unclear diagnoses [10].

The purpose of this study was to investigate the diagnostic value of CMR in a cohort of patients with suspected ACS and unobstructed coronary arteries.

The study included a total of 136 patients: 79 (58%) males and 57 (42%) females. The baseline characteristics of the patients are shown in Table 1 - the mean age of the entire population was 48.7 years. One-third of them suffered from arterial hypertension, 8.5% diabetes mellitus, 17% dyslipidaemia, and 28.5% of them were former or current smokers.

All patients were initially admitted to the coronary unit or the cathlab for chest pain with a suspicion of ACS. The final diagnoses were made after the completion of all examinations, including CMR. CMR was successfully performed in all 136 patients with no adverse events (Table 2). Wall motion abnormalities were found in 62 (47.7%) patients, LV EF was 57.6 ± 12.1%, EDV 130 ± 47 ml, signs of myocardial oedema were diagnosed in 75 (59.1%) patients and myocardial LGE in 94 (72.3%) patients. In 21 (15%) patients, LGE was subendocardial or transmural, while in 73 (54%) patients, LGE was subepicardial, mid-wall or patchy. Twenty-nine patients had pericardial effusion and/or late enhancement of the pericardium. The majority of the patients suffered from inflammatory diseases – myocarditis (n = 52, 38%) or perimyocarditis (n = 24, 18%) (Fig. 1). No patents had pericarditis alone without any myocardial involvement. Angiographically unrecognised AMI was seen on CMR in 25 patients (18%) (Fig. 2). Takotsubo CMP was diagnosed in 21 patients (15%) as a typical apical ballooning (n = 13, 10%) or as the less frequent mid-ventricular form (n = 8, 6%). Other CMPs were detected for the first time in 4 patients – 2 (1%) of them had dilated CMP and 2 (1%) had hypertrophic CMP. In 2 patients (1%), the chest pain was concluded to most probably be vertebrogenous, with an unclear reason for Troponin elevation. Despite maximum efforts, the final diagnosis remained unclear in 8 (6%) patients.

For a detailed analysis, 130 patients were included; patients with infrequent diagnoses (dilated CMP, hypertrophic CMP, and vertebrogenous disorders) were excluded from the comprehensive analysis. In line with the expectations, patients with inflammatory diseases were younger than others and had higher values of C-reactive protein. Takotsubo CMP was found more frequently in women. LV EF measured close to admission was lowest in the Takotsubo group, followed by rapid recovery (lower values in earlier vetriculographic and echocardiographic measurements compared to later CMR).

From the whole cohort, 121 (89%) patients had at least one pathology on CMR, while 15 (11%) patients had a completely normal CMR – 8 patients had unclear diagnoses, 2 had vertebral chest pain, 1 had Takotsubo CMP, 1 had histologically proven myocarditis, and 3 had a final diagnosis of more probable small AMI caused by a coronary artery spasm or small distal embolization despite the absence of a visible oedema or a scar on CMR.

All included patients had uncertain diagnoses before the CMR exam, as this was one of the inclusion criteria. On the other hand, all patients were suspected of having some type of pre-CMR diagnosis, but usually there was insufficient data for making such a diagnosis, and it was often poorly defined and chaotic. In 48 (35%) patients, CMR confirmed this previous suspicion and contributed to the diagnostic conclusion. The incremental contribution of CMR for determining the final diagnosis was crucial in 74 (57%) of the patients, where the diagnosis would have remained unclear or incorrect without using CMR - the greatest benefit of CMR was found in patients with unrecognized AMI (76% of them were diagnosed mainly by CMR) and inflammatory diseases (64% of the conclusions were reached only through the use of CMR). Only 11 (8%) CMR examinations did not help at all and had no influence on either the diagnosis or treatment (Table 3).

Our study, which evaluated the utility of CMR in a cohort of patients with suspected ACS and unobstructed coronary disease, highlights several important findings. First, it demonstrated that CMR has a high incremental diagnostic value - in more than one-half of the patients, the diagnosis would have remained unclear or incorrect without the use of CMR. Secondly, the study confirmed the results from other trials that the vast majority of patients admitted to hospital with a suspicion of ACS suffered from an inflammatory disease, which should have been the main differential diagnosis. Third, our study showed the importance of CMR in cases of AMI in patients in combination with the absence of coronary artery occlusions or stenoses.

Differently than the other trials describing CMR examination in similar cohorts of patients, our work calculated the incremental contribution of CMR for determining the final diagnosis. In more than one-half of the cases (57%), the final diagnosis was either incorrect or uncertain and the benefit of CMR was absolutely essential. In another one-third of the patients, CMR was useful for confirming the diagnoses. These results strengthen the need for using CMR in unclear ACS cases. As in many centres, CMR has not yet been included into the standard approach for these patients and our study aims to encourage making a change in these procedures, because the final diagnosis influences prescriptions for treatment. In literature, CMR diagnosis is reached in 90–93% of all cases [9, 13‐15] and CMR leads to a change in therapy in 32% of the patients [16]. Making the correct diagnosis is essential for providing adequate treatment and the appropriate risk-stratification of patients. Although the study samples in the trials were heterogeneous, some studies and meta-analysis described even worse prognoses for these patients (4.7% had a 12-months mortality) compared to patients with AMI [17].

In our study, the majority (56%) of the patients suffered from inflammatory diseases – myocarditis or perimyocarditis, while unrecognised AMI was found in 18% of the patients and Takotsubo CMP was present in 15% of the patients. The distribution of diagnoses differs among studies, mainly due to different inclusion criteria, for instance ECG changes or laboratory markers. Despite the differences in the percentage of diseases, unrecognized AMI and myocarditis are consistently the most frequent causes. In concordance with our study, very similar frequencies of diagnoses have been found by several groups. For instance, in a study by Leurent et al. [5], myocarditis was found in 59.9%, Takotsubo syndrome in 14% and myocardial infarction in 15.8% of the patients. Similarly, Stensaeth et al. [6] described myocarditis or pericarditis in 56% and stress CMP in 10% of the patients, while Assomull et al. [18] found myocarditis in 50%, AMI in 11.6% and CMP in 3.4% of the patients in their trial. Laraudogoitia Zaldumbide et al. [19] discovered myocarditis in 63%, AMI in 15% and Takotsubo CMP in 11% of the patients, and Avegliano et al. [20] diagnosed myocarditis in 60.9%, AMI in 18.7% and stress CMP in 12.5% of their patients. All these trials had very similar inclusion criteria and fewer patients than our study. Some other studies differed greatly, such as a study by Steen et al. [13] in which only 17% of the patients had myocarditis and 38% had AMI. These differences can be explained by their different inclusion criteria, especially by the absence of coronary angiography before CMR. The fact that Mahmoudi et al. [21] performed a CMR examination 2 months after coronary angiography together with missing T2 weighted sequences in the protocol, could probably explain their very low number of myocarditis patients (16%) – it could be assumed that most myocarditis patients were not diagnosed by this late CMR examination. From this comparison, we can highlight the necessity for early inclusion of CMR into the diagnostic algorithm.

CMR is an unbeatable non-invasive method for diagnosing myocarditis. The diagnostic process in myocarditis is based mainly on CMR and endomyocardial biopsy and would be very difficult without using them. Myocarditis is the third leading cause of sudden death and up to 30% of the cases can progress to dilated CMP, so that making the proper diagnosis is crucial for patient management [22, 23]. In our study, 56% of the patients had myocarditis alone or perimyocarditis. Frequently, the first clinical manifestation looks like an AMI, with oppressive chest pain, ECG changes and elevated markers of myocardial damage, so ACS is often the initial diagnosis. CMR provides functional information, but its unique contribution is in the structural information, especially in the detection of regional differences in tissue characteristics. For myocarditis, an increased T2 signal indicating increased water content in an inflamed myocardium, increased early gadolinium enhancement and epicardial, mid-wall, or patchy LGE are typical findings [4, 24]. However, recent data has described a very rapid decrease of CMR oedema markers in just a few weeks with the urgent need for CMR examination at an early stage of the disease [25]. In our study, the mean time interval between the presentation of symptoms and the CMR scan was 7.44 ± 10.03 days, so there is a high probability that all myocarditis patients were captured. The younger age and higher C-reactive protein (CRP) values in myocarditis patients are in concordance with the literature [8].

Culprit-free AMI was the second most common cause in our cohort. The 18% prevalence of these patients is consistent with the published data [8]. Several mechanisms could be proposed to explain these patients, such as coronary spasms, embolisms, atherosclerotic plaque disruptions, or spontaneous recanalization of transitory occlusions. Coronary angiography is a routine examination in patients with ACS. In cases where no visible stenoses/occlusions are found, intravascular ultrasound (IVUS) and/or coronary optical coherence tomography (OCT) can be considered. However, independently of IVUS or OCT results, the diagnosis should be confirmed by other methods. CMR with scar (and/or oedema) visualization is a perfect choice in these situations [26, 27]. A subendocardial or transmural LGE pattern in a corresponding coronary artery section leads to the correct diagnosis with subsequent therapeutic consequences and patient follow-up.

There is no specific CMR pathology for Takotsubo CMP. The diagnosis is usually made by combining transient LV dysfunction and the absence of LGE. Wall motion abnormalities could be apical in typical apical ballooning syndrome, or less frequently as mid-ventricular. Previous studies have shown a prevalence of stress CMP between 11 and 22% [8]; in our cohort, it was 15%.

In a small number of patients (8%), the final diagnosis remained unclear even after using CMR. This number is in keeping with other studies [5]. These patients had lower Troponin T levels compared to other groups and the best LV function among the groups. We could speculate about missing a very small LGE in case of microinfarctions or very mild inflammation due to limited spatial resolution or imaging artefacts. Some cases may be due to an underlying non-cardiac disease. The CMR protocol was not designed to exclude non-cardiac reasons for chest pain and biomarker elevation. Further investigation of these patients after excluding a cardiac origin of their troubles was typically not performed in our centre. There is also the possibility of false positive biochemical results.

Our study demonstrated that CMR had a powerful incremental diagnostic value in a cohort of patients with ACS and unobstructed coronary arteries. CMR is strongly recommended to be included as an inalienable part of diagnostic algorithms in these patients, because making the correct final diagnosis has direct implications on patient management.