Population

We conducted a single-centre, prospective, observational cohort study in low-risk to intermediate-risk ACS patients. From April 2008 to September 2009, 123 patients presenting to the ED of the University Hospital of Rangueil in Toulouse, France, were prospectively enrolled. Patients presenting to this ED with acute chest pain or other recent-onset angina-like symptoms, an ECG without myocardial ischemia-related abnormalities and without elevation of cardiac troponin could be prospectively enrolled. Standard chest pain assessment by the ED physicians included:

• •
  

  detailed assessment of clinical history;

  
  
• •
  

  CAD risk factors; description of symptoms;

  
  
• •
  

  physical examination;

  
  
• •
  

  18-lead ECG; chest X-ray;

  
  
• •
  

  and measurement of cardiac biomarkers (troponin I), repeated six hours later.

Patients could be enrolled at the ED if their Global Registry of Acute Coronary Events (GRACE) risk score was less than 119 ( http://www.outcomes-unmassmed.org/grace ).

Patients were excluded if aortic dissection or pulmonary embolism was particularly suspected; in these cases, a standard dedicated diagnostic algorithm was performed. Patients were also excluded if there was evidence for a cause of chest pain other than myocardial ischemia. Patients were excluded if the chest pain was related to a high-risk ACS. High-risk criteria were:

• •
  

  ST-segment elevation acute myocardial infarction;

  
  
• •
  

  diabetes mellitus known for more than 10 years;

  
  
• •
  

  a rise in serum cardiac troponin concentration;

  
  
• •
  

  ECG changes suggesting ischemia;

  
  
• •
  

  GRACE risk score greater or equal to 119;

  
  
• •
  

  history of CAD; and acute heart failure.

These patients were admitted to the cardiology intensive care unit and ICA was performed according to European recommendations .

After this evaluation, the ED physician was free to include the patient in the study if further immediate coronary evaluation was considered as being required. If the patient had no CAD risk factor, atypical chest pain and a very low GRACE score, the ED physician could allow patient to leave the ED with an appointment to see a cardiologist a few days later. The safety and efficiency of an immediate MSCT-CA has not yet been demonstrated in such cases.

At this stage, patients were not included if MSCT-CA was not feasible or was contraindicated because of: age less than 18 years; allergy to contrast agents; pregnancy; atrial fibrillation or frequent ectopy; uncontrolled heart rate; inability to perform a 20-second breath hold; body mass index (BMI) greater than 40; or renal failure (creatinine clearance less than 60 mL/min/m ² ). If the patient was suitable for inclusion, a cardiologist was called to provide a full explanation of the strategy. The patient was then included if they gave informed consent. Clinical history, CAD risk factors and description of symptoms were checked by a cardiologist at the ED. Thrombolysis In Myocardial Infarction (TIMI) and GRACE risk scores were calculated.

Technique and image reconstruction

MSCT-CA was performed using dual-source 64-slice computed tomography (SOMATOM, Dual-Source Definition; Siemens Medical Solutions, Erlangen, Germany) with retrospective electrocardiographic gating. Before acquisition, the patients received beta-blockers, targeting a heart rate less or equal to 65 beats/min. The acquisition started at the level of the carina and stopped below the heart after injection of iodine contrast. Calcium scoring was not studied (i.e. only one acquisition was performed). Moreover, no delayed acquisition was performed in order to limit radiation exposure. The acquisition delay was computed automatically by placing a region of interest in the ascending aorta, with a start threshold of 150 Hounsfield units. A biphasic injection using a dual-head injector consisted of a bolus of 1.5 cc/kg of contrast agent (iomeprol) at a concentration of 400 mg/mL (Iomeron 400, Bracco-Byk, Milan, Italy) followed by a saline flush. Collimation was 64 × 0.6 mm with a gantry rotation of 330 ms and a pitch of 0.36. The reference tube current was 320 mA. The tube voltage was adjusted to the patient’s body weight: 80 kV for patients less than 70 kg, 100 kV for patients weighing 70–90 kg and 120 kV for patients above 90 kg. Raw data were reconstructed with a slice thickness of 0.75 mm every 0.5 mm using a standard kernel, every 10% of the R-R interval. Current intensity modulation was systematically applied to reduce radiation during systolic phases (maximal current intensity between 30–40% and 70–80%, depending on the heart rate). The tube modulation programme CARE Dose 4D was also used (Siemens Medical Solutions, Erlangen, Germany).

For each patient, the images were uploaded to a dedicated workstation (Leonardo; Siemens, Erlangen, Germany) and interpreted by two experienced cardiac radiologists who were aware of the clinical data. One, two or three phases were kept for the analysis, depending on the heart rate of the patient and the presence of motion artifacts. Multiplanar reconstruction, curvilinear multiplanar reconstruction, volume rendering technique and maximum intensity projection were used for the analysis (Circulation Software; Siemens, Erlangen, Germany). A 17-segment model of the coronary arteries and a three-point grading score (normal, mild [< 50% luminal diameter narrowing], stenosis [≥ 50%]) were used to evaluate coronary stenosis; however, only segments greater than 1.5 mm in diameter were analysed. Patients were classified according to the maximal lesion. MSCT-CA was considered as ‘negative’ when there was no or mild CAD. MSCT-CA was considered as ‘positive’ when there was at least a single coronary artery stenosis greater or equal to 50%. MSCT-CA was also considered as ‘positive’ when it was not interpretable.

Diagnostic pathway

The diagnostic pathway of the patient according to the MSCT-CA result is summarized in Fig. 1 . Patients without CAD or with mild lesions on MSCT-CA were discharged early from the ED without treatment after other possible causes of chest pain had been investigated. All patients with stenosis greater or equal to 50% on MSCT-CA were admitted to the cardiology unit and had ICA. Qualitative ICA was performed by one of the nine physicians from the catheterization laboratory of the University Hospital of Rangueil in Toulouse. The interventional cardiologist was aware of the results of the MSCT-CA. The ICA was visually interpreted by two observers. Automated analysis stenosis severity assessment was also applied for lesions greater or equal to 50%, with an automated edge-detection system (Medical QCA/CMS; Medis Imaging System, Leiden, The Netherlands). A functional stress test (myocardial perfusion imaging, stress echocardiography or exercise treadmill test, according to availability) or fractional flow reserve (FFR) was performed to assess the haemodynamic significance of intermediate stenosis greater or equal to 50% but less or equal to 70% and direct myocardial revascularization indication. Only Lesions with an FFR less than 0.80 at maximum hyperaemia were considered functionally significant and were treated by mechanical revascularization . Only severe lesions and intermediate lesions with proven myocardial ischaemia were considered for myocardial revascularization. Non-obstructive CAD is defined by stenosis greater or equal to 50% but less or equal to 70% on ICA with a negative stress test or FFR test. Medical therapy, including aspirin, beta-blockers and statins, was proposed for patients with non-obstructive CAD. Obstructive CAD is defined by stenosis greater or equal to 70% on ICA or less or equal to 50% on ICA with an abnormal functional stress test or FFR less than 0.80. A revascularization strategy associated with medical therapy was proposed for obstructive CAD patients.

Study population.

Follow-up

All patients or proxies and general practitioners were contacted by the first investigator (S.H.) by a telephone recall between January 2010 and April 2011. Clinical status was queried. During follow-up, death or a history of myocardial infarction, myocardial revascularization or heart failure was searched for. The main outcome criteria were major adverse coronary events (MACE), defined as myocardial infarction, myocardial revascularization and cardiovascular death occurring after the primary hospitalization. For the patients followed in our cardiology department, the follow-up data in the medical information file were noted.

Statistical analysis

Continuous variables are expressed as mean ± standard deviation (SD) and the median is provided if the distribution of the variable departed from normality. Categorical variables are expressed as total number (percentage). The SCORE method with continuity correction was used to estimate the 95% confidence interval (CI) for proportions . Student’s t test was performed to compare means of quantitative variables between groups. Fisher’s exact test was performed to compare proportions. A p value less than 0.05 was considered to be statistically significant. All calculations were done using MedCalc statistical software, version 8.0 (MedCalc Software, Mariakerke, Belgium).