Summary
Background
Acute coronary syndrome (ACS) with normal coronary angiography is a frequent clinical situation with an uncertain prognosis. Cardiac magnetic resonance imaging (CMRI) is a powerful tool for differential diagnosis between myocardial infarction (MI), acute myocarditis and Tako-tsubo cardiomyopathy (TTC). Data are sparse regarding the evolution of patients presenting an ACS with normal coronary arteries and normal CMRI.
Aims
To evaluate the evolution of patients presenting an ACS with normal coronary arteries and normal CMRI, with a 1-year follow-up.
Methods
Eighty-seven consecutive patients (mean age, 53 years; 40.2% men) presenting an ACS with troponin elevation and normal coronary arteries by angiography were prospectively included. All patients underwent CMRI at 3-Tesla. Adverse events were recorded with 1-year follow-up.
Results
A likely aetiology for the acute clinical presentation was established by CMRI in 63.2% of patients (22.7% MI, 26.4% acute myocarditis, 11.5% TTC). During follow-up, one patient in the MI group had a stroke (1.2%). In the myocarditis group, there was one initial cardiogenic shock, one episode of congestive heart failure (1.2%) and nine patients had recurrent chest pain without troponin elevation (10.3%). Two TTC group patients initially presented with cardiogenic shock (2.4%); there were no other adverse events in this group during follow-up. In the remaining 36.7% patients, no clear diagnosis could be identified by CMRI, and no adverse events occurred during follow-up.
Conclusion
CMRI is a useful tool for the management of ACS presenting with normal coronary angiography, as it helps to ascertain the diagnosis and adapt treatment in a large proportion of cases. Nonetheless, patients with no abnormalities identified by CMRI have an excellent evolution.
Résumé
Objectifs
Le syndrome coronaire aigu (SCA) avec des artères coronaires angiographiquement normales est une situation clinique fréquente mais avec un pronostic incertain. L’IRM cardiaque est un examen performant pour le diagnostic différentiel entre l’infarctus du myocarde (IDM), la myocardite aiguë et le syndrome de Tako-tsubo. Peu de données sont disponibles concernant l’évolution des SCA à coronaires normales et IRM cardiaque normale.
Méthode
Quatre-vingt sept patients (âge moyen : 53 ans, 40,2 % d’hommes) présentant un SCA avec augmentation de la troponine et des artères angiographiquement normales ont été prospectivement inclus. Tous les patients ont bénéficié d’une IRM cardiaque à 3-Tesla et les évènements indésirables étaient recensés avec un suivi d’un ans.
Résultats
Un diagnostic étiologique a été établi par l’IRM chez 63,2 % des patients (22,7 % avaient un IDM, 26,4 % une myocardite aiguë, 11,5 % un Tako-tsubo). Au cours du suivi, un patient du groupe IDM a présenté un accident vasculaire cérébral (1,2 %). Dans le groupe myocardite aiguë, il a été retrouvé un état de choc cardiogénique initial, un épisode de décompensation cardiaque et neuf patients ont présenté des récidives douloureuses thoraciques sans augmentation de la troponine (10,3 %). Deux patients du groupe Tako-tsubo présentaient initialement un état de choc cardiogénique, puis il n’a été noté aucun autre évènement indésirable au cours du suivi dans ce groupe. Finalement, pour les 36,7 % des patients restant, chez qui l’IRM ne permettait pas de poser un diagnostic final, aucun évènement indésirable n’ait survenu au cours du suivi d’un ans.
Conclusion
L’IRM cardiaque est un examen performant pour la prise en charge des SCA à coronaires angiographiquement normales qui permet d’affirmer un diagnostic et ainsi d’adapter les thérapeutiques dans une large proportion de cas. Par ailleurs, les patients qui ne présentent pas d’anomalie à l’IRM on une excellente évolution.
Introduction
Acute coronary syndromes (ACS) are one of the leading causes of death and morbidity in industrialized countries . Typical presentation includes acute chest pain and cardiac troponin I (cTnI) elevation, possibly associated with electrocardiogram (ECG) abnormalities. Subsequent coronary angiography usually reveals flow-limiting epicardial stenoses . However, coronary angiography can be normal in 1 to 12% of cases, depending on the definition of “normal” coronary arteries .
Differential diagnosis can thus be challenging and possible diagnoses include acute myocardial infarction (AMI), generally limited to the subendocardial territory , and two non-ischemic myocardial diseases, namely acute myocarditis and transient apical ballooning syndrome, also known as Tako-tsubo cardiomyopathy (TTC) .
Cardiac magnetic resonance imaging (CMRI) is a highly sensitive noninvasive technique for detecting myocardial damage in ischemic and non-ischemic cardiac disease, including acute myocarditis and TTC .
Management strategy and follow-up can be thus adapted secondarily, further to the results of CMRI, with the initiation of secondary prevention treatment, including antiplatelet therapy, when CMRI confirms a diagnosis of AMI . However, Assomull et al. demonstrated that CMRI can identify the basis for cTnI elevation in only 65% of patients presenting ACS symptoms and unobstructed coronary arteries . Moreover, the evolution of this clinical situation, usually reported in the past as uniformly benign , might in fact be less favourable, with one recent study reporting a significantly high rate of death and major adverse events .
In this context, the aim of this study was to evaluate the evolution of patients presenting with ACS and normal coronary arteries and normal CMRI, with a follow-up of 1 year. For CMRI evaluation, the latest technology available in clinical practice was used with a high field of Tesla.
Methods
Study population
Patients with clinical suspicion of ACS (ST-segment elevation myocardial infarction or non-ST-segment elevation ACS) and with strictly normal coronary arteries on coronary angiography were prospectively recruited in this observational study. Inclusion criteria were defined as: the association of new onset chest pain present at rest, lasting for longer than 30 minutes; elevated cTnI; and normal coronary angiography, performed no later than 72 hours after admission and reviewed by two experienced observers. Other initial investigations and searches for alternative causes of cTnI elevation, systematically including transthoracic echocardiography (TTE), had proved diagnostically inconclusive. Intracoronary methylergonovine testing was also systematically performed during angiography to exclude coronary vasospasm. Exclusion criteria included prior history of cardiovascular disease, previous coronary intervention, high Framingham risk score (10-year risk greater than 10%) and standard CMRI contraindications. Patients presenting with cardiac rhythm disorders, such as atrial fibrillation, which could explain cTnI elevation, were also excluded.
Data collected included baseline characteristics, medication at admission and ECG abnormalities using the Minnesota code . Daily blood samples were obtained to assess peak cTnI, lipid profile, inflammatory markers including C-reactive protein, B-type natriuretic peptide, glycaemia and creatinine concentration to evaluate glomerular filtration rate. The threshold used in the core laboratory to define a positive cTnI was 0.15 ng/mL and cTnI assessment was performed every 6 hours until peak concentrations were reached. Framingham and Global Registry of Acute Coronary Events (GRACE) risk scores were calculated for each patient to evaluate long- and short-term risk of death. Myocardial tissue biopsy was not envisaged to establish a diagnosis.
All patients initially received standard medical therapy for ACS, including antiplatelet therapy, statins, angiotensin-converting enzyme inhibitors and beta-blockers, according to international guidelines . Treatment strategy was subsequently adjusted according to the results of CMRI. In case of myocarditis, anti-inflammatory therapy was initiated if chest pain indicative of associated pericardial reaction persisted. For TTC, treatment was initiated at the physician’s discretion, combining an inhibitor of the renin-angiotensin system with beta-blocker therapy. If CMRI was normal, all therapy was stopped.
Cardiac magnetic resonance imaging
CMRI scans were performed during hospital stay or within 3 weeks of initial presentation and repeated at 3 months in case of suspected TTC to evaluate improvement in left ventricular (LV) function. CMRI studies were conducted at 3.0 field strength (General Electric Healthcare Company, Signa HD, Milwaukee, WI, USA) with a standard 40 mT/m gradient, using an eight-element phased array surface coil. Left ventricular function was assessed by ECG-gated cine steady-state free precession (SSFP) breath-hold sequences in the two-chamber and four-chamber views as well as in the short cardiac axis from base to apex (30 phases per cardiac cycle; repetition time, 3.5 ms; echo time, 1.2 ms; flip angle, 45°; typical voxel size, 1.92 × 1.25 × 8.0 mm) . T2-weighted images (triple inversion recovery; TE, 60 ms; TR, 2 × R-R interval; TI, 170 ms; slice thickness, 7 mm; flip angle, 180°; pixel size, 2.3 × 1.3 mm) were acquired in the short-axis plane. First-pass perfusion imaging was performed using a T1-weighted fast gradient echo sequence (FGRE TR/TE = 3.5 ms/1.5 ms) with a notched saturation pulse, after the injection of a bolus of gadolinium (Dota-Gd; Guerbet, Roissy, France) in a brachial vein at a single dose of 0.2 mL/kg (0.1 mmol/kg). The field of view was 400 mm and the matrix size was 256 × 224 interpolated to 256 × 256. The heart was imaged in the short-axis plane with four to six slices, 8 mm thick with a gap of 1 mm. One image per slice was acquired every two cardiac cycles, leading to a temporal resolution of 2 R-R per imaging plane. Forty frames were obtained from each imaged plane within 80 R-R. For late gadolinium enhancement (LGE) imaging at 3 and 15 minutes, a breath-hold ECG-gated T1-weighted sequence was used (TE = MinFull; field of view, 440 mm; TI optimised to obtain optimal myocardial nulling; matrix size, 256 × 224 interpolated to 256 × 256; slice thickness, 8 mm; gap, 1 mm) . The number and position of the slices were the same as used for the perfusion first-pass imaging.
Offline image analysis was performed on a dedicated workstation (General Electric Healthcare Company, Milwaukee, WI, USA). The endocardial border was drawn manually on each dynamic image. Left ventricular ejection fraction (LVEF), end-diastolic and end-systolic volume indexes (LVEDI and LVESI, respectively) and LV mass index were calculated from the short-axis view . Qualitative interpretation of CMRI scans was performed by two reviewers by consensus. Readers were blinded to the clinical situation. The portion of the heart exhibiting late enhancement and/or T2 high signal intensity was classified according to the segmentation established by the American Heart Association, which divides the left ventricle into 17 segments . LGE images and corresponding T2-weighted images were assessed for subendocardial signal abnormalities in the distribution of a coronary artery, compatible with AMI . Diagnosis of acute myocarditis was made according to criteria previously described by Friederich et al. with detection of oedema in T2-weighted images, hyperaemia and capillary leakage in myocardial early (3 minutes) gadolinium enhancement sequences and necrosis or fibrosis in LGE images detected in the midwall/subepicardial regions . TTC was suspected on initial CMRI if there were no signs of delayed enhancement, T2 signal intensity and LV dysfunction, especially with mid/apical ballooning; TTC was confirmed by complete normalization of LVEF on the follow-up CMRI at 3 months. Scans with volumes and function within the normal range, with no LGE and T2 signal intensity abnormalities were considered normal by CMRI.
Outcomes
The subsequent progress of patients was assessed by telephone contact at 1, 3, 6 and 12 months. In case of clinical events, these were confirmed by hospital records or contact with the general practitioner. Clinical endpoints evaluated were death, recurrent ACS, congestive heart failure, stroke and recurrence of chest pain without cTnI elevation.
Statistical analysis
Continuous data are expressed as mean ± standard deviation. Baseline characteristics were compared between the diagnostic CMRI group and the group in which CMRI was judged to be normal, using an independent sample t test and the Chi 2 test for categorical variables, as appropriate. All tests were two sided. A P value of less 0.05 was considered statistically significant. All statistical analyses were performed using MedCalc © (Mariakerke, Ghent, Belgium).
Results
Baseline characteristics
Among a total of 1687 patients admitted to our institution for ACS between January 2008 and September 2009, 87 (5.1%) had a normal coronary angiogram and met the inclusion criteria. Baseline characteristics are summarized in Table 1 . The mean age was 53 ± 18.5 years and 59.8% were women ( n = 52). No patient had pre-existing renal impairment, prior exposure to cardiotoxic drugs or prior chemotherapy. At the time of cTnI evaluation, no patient had symptoms suggestive of alternative cause of cTnI elevation, such as septicaemia, stroke, renal insufficiency, pulmonary embolism (PE) or aortic dissection. Eight patients (9.2%) described psychological stress before clinical presentation. The prevalence of cardiovascular risk factors and primary prevention therapy was moderate. No patient underwent thrombolytic therapy. Sixty-nine (75.8%) patients had an abnormal ECG on presentation. ST-segment depression was the most commonly detected abnormality (59.4%), followed by T wave changes (26.5%). One patient presented left bundle branch block (1.1%) and none had supraventricular or ventricular tachycardia. No patient with ST-segment elevation was eligible for inclusion. No criteria of right ventricle overload indicative of PE with cTnI elevation were found in the initial TTE and 32 patients (36.8%) underwent further investigation by CT pulmonary angiography ( n = 22) and ventilation perfusion isotope scans, which did not show evidence of embolism in any patient.
