Acute chest pain is one of the most frequent clinical problems in the ED. ACS is an important life-threatening condition causing acute chest pain. Multiple factors with interfering symptoms can cause chest pain, so the exclusion of ACS in ED is often a prolonged process requiring monitoring of cardiac serum markers especially in intermediate risk patients. This makes this process long and costly. This process is not only expensive, but also time-consuming for both physicians and patients. On the other hand, up to 11% of patients are inadvertently discharged from the ED with a missed AMI (average 2.1%), and unstable angina pain is missed in up to 4%. This misdiagnosis and inappropriate discharge leads to increased mortality for those patients [14].
Studies had demonstrated the application of TRO protocol with 16 sections in CT technology in the emergency department. The advent of 64-, 256- and 320-multislice CT scanners has improved imaging of the coronary artery. The latest generation of 320-multislice CT allows simple, rapid acquisition of the aorta, pulmonary arteries and coronary arteries in a single breath hold [15].
In patients who do not appear to be typically presenting with an ACS, other important provisional diagnoses for the cause of their chest pain that needed to be considered include non-ACS cardiac conditions such as acute pericarditis, life-threatening non-cardiac conditions such as acute pulmonary embolism, aortic dissection and perforating ulcer. The usefulness of TRO in ED has been suggested in some studies, and other studies failed to show superior clinical outcomes to TRO CT [16,17,18].
In this study, we aimed to demonstrate the diagnostic yield of TRO CTA in acute chest pain in emergency department patients and its utility to provide a definitive non-coronary diagnosis as well as its use to exclude coronary artery disease. As our main direction in this study was to access the process of clinical triage and its diagnostic outcome for our patient, our study differs from other coronary CT angiography studies that compare multislice CT with an angiographic assessment of stenosis.
Multiple studies have demonstrated that a normal coronary CT angiography ‘allows the clinician to rule out the presence of hemodynamically relevant coronary artery stenosis with a high degree of reliability’ and that no further work up is required to demonstrate the absence of coronary disease [19, 20].
Furthermore, in those patients with coronary disease, the likelihood that an individual coronary plaque will result in ACS is related to many factors independent of the degree of vascular narrowing [21].
The accepted ‘gold standard’ for the grading of coronary stenosis, conventional catheter angiography, is limited by geometric factors that may result in underestimation or overestimation of the degree of coronary stenosis in over one-third of patients [22].
On the basis of these considerations, we opted to focus mainly on clinical outcomes of patients suspected of having ACS in the ED rather than a direct comparison to catheter angiography for assessment of the degree of stenosis.
As regards the behaviour of TRO CTA in assessing coronary artery disease, we were able to say confidentially that 42% of our population had negative coronary results and non-significant coronary disease; this did not seem to be surprising as regards our included population with moderate risk to ACS.
TRO CTA behaviour in the assessment of coronary disease
As regards our positive coronary findings, we diagnosed 40% (18 of 45) of our population had significant coronary stenosis. 17.7% (8 of 45) of our patients had moderate coronary stenosis, and 22.2% (10 of 45) had severe coronary stenosis
We used 50% as a cut off value to define obstructive lesion or significant lesion, we chose this cut off value because this threshold usually requires further evaluation either with non-invasive testing to determine functional significance or with catheter angiography. This cut off has been used in many published studies, e.g., Hoffmann et al [23]. Patients with significant lesions were further subclassified as moderate disease (50–70% stenosis) and severe stenosis (> 70% stenosis).
In Thomas et al. [24] TRO CTA study, they were able to diagnose 19% of their patients who have significant coronary stenosis, they also used 320-multislice CT scanner, but they included larger population in their study.
A recent TRO CTA study was done using 320-multislice CT on 25 low to intermediate acute chest pain patients. Using TRO CTA, they were able to diagnose 50% of their patients of having coronary lesions [25]. Our result is perfectly matching with their result as we found 57.8% of our patients having coronary artery disease, but we included only intermediate risk patients in our study and larger study population.
In Kevin et al. [26] study done over 197 low to intermediate risk to access the role of TRO CTA in the emergency department, 11% of their population was found to have significant coronary artery lesions. Their study was done on 64-multislice CT, and they also included low to intermediate risk in their population. This seemed the real cause of having lower percentage with significant coronary disease.
In our study done in over 45 patients having acute chest pain with intermediate risk to ACS, 15 patients with significant coronary disease have entered invasive catheterization. Based on per segmental analysis comparison, we were able to estimate sensitivity and specificity of our study.
We were able to diagnose obstructive coronary artery disease with high specificity reaching 96%, positive predictive value reaching 69%, based on per segmental analysis and in comparison to invasive angiography. We had also high levels of sensitivity and negative predictive values reaching 95% and 99% respectively.
We compare our results with the following dedicated coronary CTA studies:
Hoffmann et al. [23] study was done in over 368 patients using 64-mutislice CT which showed high specificity and high negative predictive value in exclusion obstructive coronary artery lesions reaching to 87% and 100% respectively; this agreed with our result because their result also showed high sensitivity of 77% in the diagnosis of obstructive lesions but they got only 35% as a positive predictive value.
Our result goes well with Nieman et al. [27] and Ropers et al. [28] results with high sensitivity and specificity reaching 95% and 86% respectively in Nieman et al.’s study and 92% and 93% respectively in Ropers et al.’s study and both studies were done to assess coronary luminal stenosis greater than 50% using only 16-detector CT.
Leber et al. [29], Leschka et al. [30] and Fine et al. [31] made studies in detection of coronary artery disease over 64-mutislice CT with high sensitivity 91% and specificity 96%; they also got high negative predictive value reaching 97%.
Han et al. [32] study which was done on 345 patients using 64-multislice CT found higher specificity and positive predictive value reaching 99% and 87% respectively than we found in our results, but they got the same negative predictive value 99% and our sensitivity result was higher than they got, as they got 81%.
Raff et al. [33] and Puglise et al. [34] studies showed also high sensitivity and specificity in their studies done to detect coronary artery disease, they did their studies using 64-multislice CT and they got 91% sensitivity and 96% specificity.
Our negative predictive value agrees also with Kevin et al. [26] which was 99.4%, and they also based their study on the clinical follow-up at 30 days and further diagnostic testing in a minority of patients.
So, our TRO CTA study shows comparable results to dedicated coronary artery CTA as regards exclusion of ACS with high sensitivity and specificity.
Our TRO CTA study showed the same result of another systematic review study done by Ayaram et al. [35]; they showed that TRO CTA helped them in the diagnosis of coronary artery disease in patients with acute chest pain with 94.3% sensitivity, 97.4% specificity and 99% negative predictive value.
TRO CTA behaviour in assessment of non-coronary disease
In our study, we found 24.4% (11 out of 45) of our patients that were suspected to have ACS, having another diagnosis that explained their symptoms. 17.7% of them had clinically important non-coronary findings that explained their symptoms. This is not surprising given that ACS symptoms such as chest pain and shortness of breath can mimic a number of other diseases.
In 17.7% of our population, we were able to diagnose life-threatening conditions as aortic dissection, aortic intramural hematoma and pulmonary embolism. We were able to save the life of these patients by directing them toward the accurate cause of chest pain and from then to accurate treatment direction using single-shot acquisition imaging without any other time-consuming tests.
Among those non-coronary life-threatening conditions, we identified pulmonary embolism in 6.6% of our population and aortic dissection in 4.4%, 2.2% had intramural hematoma and 2.2% had partially thrombosed aortic aneurysm. We found that our detection rate of aortic dissection and pulmonary embolism may have been reduced by the exclusion of patients when treating physician chooses to order a dedicated CT scan rather than triple role out protocol.
Our non-coronary results agreed also with the non-coronary results in Thomas et al. [24] study. This study was done on 100 emergency patients using 320-multislice CT, in his study 20 patients demonstrated other non-coronary findings that explain chest pain including PE, pleural effusion, left ventricular hypertrophy with pleural effusion and pneumonia. They differed from us as they included pneumonia and left ventricular hypertrophy with pleural effusion as causes for chest pain.
In our study, we were able to diagnose pulmonary embolism, aortic intramural hematoma and aortic dissection as emergency causes of acute chest pain. In Soliman TRO CTA study [25] done on acute chest pain patients, he diagnosed 17 cases with non-coronary findings including pulmonary embolism (12.5%), aortic dissection (22.5%) and aortic aneurysm (10%).
White et al. [36] and Savino et al. [37] agreed also with our result that TRO CTA can diagnose non-coronary chest pain causes as well as coronary causes. White et al. TRO CTA study was done on 69 patients with acute chest pain showed significant non-coronary findings as pericarditis, pneumonia and pulmonary embolism that could account for chest pain in 4.3% of patients in their studies, while Savino et al. study done over 23 acute chest pain patients using TRO CTA found 8.7% of their populations to have pulmonary embolism.
A previous study was done by Halpern [15] and Takakuwa [16] included 201 acute chest pain patients evaluated with TRO CT, which identified a non-coronary diagnosis as the explanation for acute chest pain in 11% of patients.
Additional diagnosis in patients undergoing coronary CT angiography may be clinically important. Some diagnosis, such as hiatus hernia, may seem to be of less clinical importance than other diagnosed life-threatening conditions that should be excluded first, but may nonetheless represent the true cause for the representing complaint. Identifications of these abnormalities allow the emergency department physician to direct the patient to an appropriate physician to treat and monitor disease progression.
Among our population, two patients (4.4% of our population) had pulmonary parenchymal lesions and the first had pulmonary nodule that proved to be pulmonary malignancy while the other had right middle lobe collapse. Another two patients (4.4% of our population) were diagnosed to be complaining due to diaphragmatic hernia.
In Gruettner et al. [38] TRO CTA study, they showed even higher numbers of their patients having non-coronary disease reaching to 36%, 5 patients had pulmonary embolism, 1 patient had severe right ventricular dysfunction with pericardial effusion and one patient had incidental bronchial carcinoma. They diagnosed only 19% of their population to have significant coronary stenosis. In their study, they included intermediate risk patients as we did, but their follow-up period was for 90 days.
The ability to identify an alternative non-coronary diagnosis in patients suspected of having ACS is a major strength of the triple rule-out scan. In our study, we were able not only to diagnose other life-threatening conditions causing chest pain as pulmonary embolism and acute aortic syndrome, but the larger field of scanning helped us to diagnose other non-emergency conditions as hernia in two patients and pulmonary nodule in one patient and directing those patients toward the right physicians helping them to treat the true cause of their chest pain.
On the basis of the frequency of clinically important non-coronary findings in our study, the detection of extra coronary findings in moderate risk ACS population provided the true cause of chest pain more frequently than was expected. Although many of these non-coronary causes might have been diagnosed after further observations and diagnostic testing, the triple rule-out protocol provided these diagnoses quickly and without the need for additional testing, such as ventilation-perfusion scanning.
We found that TRO CTA is a very useful protocol in emergency department helping in rapid discharge in ED as we were able to safely and rabidly discharge 44.4% of our population based on their negative results or non-significant coronary findings and confirmed by 60 days follow-up. Our results agreed with Thomas et al. results [24], in his study 60 of 100 patients were discharged on the same day based on their negative results. None of the discharged patients showed MACE during the 90-day follow-up.
Amelia M. Wnorowski et al [39] wide study on 970 low to moderate risk patients using 256 multi-detector CT study agreed with our result as regard behaviour of TRO CTA in the emergency department as they showed that they were able to discharge 81.4% of their patients based on their negative results. They stated also in their result that they were able to diagnose 8.9% of their patients with non-coronary diagnosis as pulmonary embolism and aortic disease.
Our high negative predictive value reaching 99% enables us to recommend TRO CTA protocol to be used as a guide discharging intermediate-risk patients in acute chest pain emergency department.
Based on per segmental analysis, having a positive predictive value of 69% indicate that our TRO CTA did not completely correlate with cardiac cauterization result in severe coronary artery disease .In our opinion, this may be caused by overestimation of severity of stenosis in severely calcified lesions.
We found that despite of good result of TRO studies in dealing with rapid and accurate diagnosis in acute chest pain patients in ED, yet it is not ordered in the same frequency as dedicated coronary CTA, this was in our opinion due to lack of popularity of technique to clinicians who are accustomed to request dedicated coronary CTA in ACP patients with possible risk to ACS. Also lack of patients standardization criteria and knowledge of which patients are proposed to do this examination.
One of drawbacks of TRO examination that may lower its clinical significance is the degraded image of RCA due to patient’s respiratory motion artefacts, seen in patients suffering from dyspnoea and chest pain. In spite of very rapid scanners, RCA image may be still at motion risk.
TRO CTA and radiation exposure
We found that our estimated effective dose value (EDV = 9.7 mSv) was slightly lower than of Soliman [25] and Gruettner et al. [38] TRO studies who had EDV reaching to 10 and 12.4 mSv respectively. Our EDV was still higher than EDV of patients underwent dedicated coronary CTA in the Joachim Gruettner study as they got 8.7 mSv.
TRO study radiation exposure values pushed us to recommend this examination only in certain patients, who are of high suspicion to have an alternative diagnosis that may explain their chest pain other than acute coronary syndrome. This suspicion may be based on clinical symptoms as severe dyspnoea and high risk for pulmonary embolism (history of DVT), or based on laboratory data as positive D-Dimer test.