triple-rule-out ct angiography for evaluation of acute...

Triple-Rule-Out CT Angiographyfor Evaluation of Acute ChestPain and Possible AcuteCoronary Syndrome1

Ethan J. Halpern, MDTriple-rule-out (TRO) computed tomographic (CT) an-giography can provide a cost-effective evaluation of thecoronary arteries, aorta, pulmonary arteries, and adjacentintrathoracic structures for the patient with acute chestpain. TRO CT is most appropriate for the patient who isjudged to be at low to intermediate risk for acute coronarysyndrome (ACS) and whose symptoms may also be attrib-uted to acute pathologic conditions of the aorta or pulmo-nary arteries. Although a regular cardiac rhythm remainsan important factor in coronary CT image quality, newerCT scanners with 64 or more detector rows afford rapidelectrocardiographically (ECG) gated imaging to providehigh-quality TRO CT studies in patients with a heart rate ofup to 80 beats per minute. Injection of iodinated contrastmaterial (!100 mL) is tailored to provide simultaneoushigh levels of arterial enhancement in the coronary arter-ies and aorta (!300 HU) and in the pulmonary arteries(!200 HU). To limit radiation exposure, the TRO CTexamination does not include the entire chest but is con-strained to incorporate the aortic arch down through theheart. Scanning parameters, including prospective ECGtube current modulation and prospective ECG gating withthe “step-and-shoot” technique, are tailored to reduce ra-diation exposure (optimally, 5–9 mSv). When performedwith appropriate attention to timing and technique, TROCT provides coronary image quality equal to that of dedi-cated coronary CT angiography and pulmonary arterialimages that are free of motion artifact related to cardiacpulsation. In an appropriately selected emergency depart-ment patient population, TRO CT can safely eliminate theneed for further diagnostic testing in over 75% of patients.

! RSNA, 2009

1 From the Department of Radiology, Thomas JeffersonUniversity, 132 S 10th St, Philadelphia, PA 19107-5244.Received December 31, 2008; revision requested Febru-ary 5, 2009; revision received February 10; acceptedFebruary 18; final version accepted February 25. Addresscorrespondence to the author (e-mail: [email protected] ).

" RSNA, 2009

REVI

EWS

AND

COM

MEN

TARY

#HO

WID

OIT

332 radiology.rsnajnls.org $ Radiology: Volume 252: Number 2—August 2009

Note: This copy is for your personal non-commercial use only. To order presentation-ready copies for distribution to your colleagues or clients, contact us at www.rsna.org/rsnarights.

Evaluation of chest pain in theemergency department (ED) is apublic health issue of great conse-

quence. According to the most recentavailable health statistics report fromthe Centers for Disease Control andPrevention, evaluation of acute chestpain and related symptoms was the sec-ond most common reason for a visit tothe ED by a female adult and the mostcommon reason by a male adult in theUnited States in 2006 (1). Chest painaccounted for 6 392 000 ED visits and1 976 000 hospital admissions. Overall,suspected heart disease and chest painwere the most common reasons for di-rect admission from the ED and ac-counted for 2 492 000 hospital admis-sions in 2006.

The differential diagnosis of chestpain is a complex problem for the EDphysician. The diagnosis of acute coro-nary syndrome (ACS) includes unstableangina, non–ST-elevation myocardial

infarction, and ST-elevation myocardialinfarction. Of patients presenting to theED with symptoms of ACS, only 25%ultimately have a confirmed diagnosis ofACS at the time of discharge (2). Thefailure rate for diagnosis of ACS amongpatients presenting to the ED is in therange of 2%–5% (3,4) but may be ashigh as 29% at low-volume centers (5).

Patients in whom the diagnosis ofACS is missed tend to be younger and tohave an atypical presentation and a non-diagnostic electrocardiogram (ECG) (6).The missed diagnosis of ACS is a commonreason for litigation against ED physiciansand accounts for up to 25% of the totalmalpractice liability of ED physicians (7).On the other hand, uncertainty in the di-agnosis of ACS results in the practice ofdefensive medicine and begets an in-creased number of diagnostic tests andhospital admissions (8). The cost of neg-ative inpatient cardiac evaluations is esti-mated at $6 billion in the United Stateseach year (9).

Clinical Role of Triple-Rule-Out CT

Numerous studies have demonstratedgood to excellent diagnostic accuracyof dedicated coronary computed to-mographic (CT) angiography for eval-uation of coronary disease (10), withexcellent negative predictive values(11,12). However, few reports havedescribed the application of CT as partof the triple-rule-out (TRO) examina-tion with a dedicated TRO injectionand scan protocol (13).

TRO CT is a tailored ECG-gated ex-amination designed to evaluate the aorta,coronary circulation, pulmonary arteries,and the middle to lower portion of thechest with a single scan. Application ofthe TRO examination for evaluation forsuspected ACS in the ED is possible be-cause of advances in CT technology thatprovide greater z-axis coverage with im-proved temporal resolution and de-creased radiation dose. A recent surveyof radiology practices found that 33%used CT in the ED for the work-up ofchest pain and that 18% were using aTRO protocol (14).

All patients with ACS require hospitaladmission, and many will benefit from

rapid triage to cardiac catheterization andintervention. On the other hand, when apatient’s presentation clearly suggests anoncardiac diagnosis, coronary evalua-tion is not required and is not cost effec-tive. The remaining patients suspected ofhaving ACS must be cleared of this diag-nosis prior to discharge. Given the poten-tially life-threatening consequences ofmissing a diagnosis of ACS, a high nega-tive predictive value is critical for dis-charging patients with possible ACS.

The negative predictive value of cor-onary CT angiography for ACS will de-pend on the prevalence of coronary dis-ease in the study population. A recentmulticenter trial (15) demonstrated a99% negative predictive value of coro-nary CT angiography for coronary dis-ease at both the patient and the vessellevels in a population with a diseaseprevalence of less than 25%, establish-ing coronary CT angiography as an ef-fective noninvasive examination to ruleout obstructive coronary artery steno-sis. Although another recent multi-center trial (16) demonstrated a nega-tive predictive value of only 83% forcoronary CT angiography, that studyevaluated a population with a high(56%) prevalence of obstructive coro-nary disease. On the basis of the resultsof these studies of dedicated coronaryCT angiography, it is likely that TRO CTwill be most effective in a populationwith a low prevalence ("50%) of ob-structive coronary disease.

For those with a low risk of ACS whoare evaluated with conventional nuclearstress testing, only one-third of patientswith a positive or indeterminate stresstest result are found to have clinically sig-nificant coronary disease at the time ofcatheterization (17). For the evaluation of

Published online10.1148/radiol.2522082335

Radiology 2009; 252:332–345

Abbreviations:ACS # acute coronary syndromeECG # electrocardiogramED # emergency departmentMIP # maximum intensity projectionTRO # triple rule out

Author stated no financial relationship to disclose.

Essentials

# The primary goal of triple-rule-out(TRO) CT in the emergency de-partment is to facilitate the saferapid discharge of patients judgedto be at low to intermediate riskof acute coronary syndrome.

# The detection of noncoronary le-sions that explain the presentingcomplaint is a major advantage ofthe TRO CT examination overnuclear stress testing.

# TRO studies are most appropriateand cost-effective when there is asuspicion for acute coronary syn-drome along with other diagnosessuch as pulmonary embolism,acute aortic syndrome, or nonvas-cular disease in the thorax.

# An optimized TRO protocol pro-vides excellent image quality foraortic and coronary and pulmo-nary arterial evaluation whileminimizing contrast agent doseand radiation exposure.

# Attention to the details of patientpreparation, contrast agent ad-ministration, and timing of thescan is the key to high-qualityTRO studies.

HOW I DO IT: Triple-Rule-Out CT Angiography Halpern

Radiology: Volume 252: Number 2—August 2009 $ radiology.rsnajnls.org 333

patients presenting to the ED who arejudged to be at low risk for ACS, coronaryCT angiography is at least as accurate asnuclear imaging (18) and allows the safeand rapid discharge of low- to intermedi-ate-risk ACS patients (19–21). Results ofa recent study (22) suggest that in low- tomoderate-risk patients, a CT triage modelis less costly and more effective thanstrategies based on either stress echocar-diography or stress ECG testing. The au-thors of another recent study (23) con-cluded that “compared to the other strat-egies, immediate CTA [CT angiography]is safe, identified as many patients withcoronary disease, had the lowest cost,had the shortest length of stay, and al-lowed discharge for the majority of pa-tients.” TRO CT precludes the need foradditional diagnostic testing in over 75%of patients with low to intermediate riskof ACS and provides the additional advan-tage of helping find noncoronary diag-noses that explain the presenting com-plaint in 11% of ED patients (24). TROCT eliminates the need for separate ded-icated studies for coronary disease, aorticdissection, pulmonary embolism, andother acute chest conditions. In a prop-erly selected population, coronary CT canprovide a cost-effective evaluation (25)with reduced diagnostic time, lowercosts, and fewer repeat evaluations forrecurrent chest pain, as compared withstandard diagnostic evaluation (26).

Among patients who present to theED with a low to moderate risk of ACSand who are evaluated with TRO CT, aminority ("10%) are subsequently eval-uated with conventional cardiac cathe-terization. Among those ED patientswho are studied with both TRO CT andcardiac catheterization, few normal car-diac catheterization results would be ex-pected (24). Since it would not be ethi-cal to subject most patients with a lowto moderate risk of ACS to cardiac cath-eterization, there are no studies thatconfirm the negative predictive value ofTRO CT relative to conventional arte-riography in the ED population. None-theless, if the quality of coronary imag-ing obtained with TRO CT is equivalentto that of dedicated coronary CT an-giography, one would expect the samehigh diagnostic accuracy and negative

predictive value that have been docu-mented with dedicated coronary CT an-giography.

Injection and scan techniques forTRO CT studies vary considerably fromone institution to another, resulting ininconsistent image quality. Some radiol-ogists are reluctant to perform TROstudies because of an impression thatthe TRO examination is too technicallychallenging or that the quality of thecoronary artery study is compromisedin the TRO examination. The goals ofthis article are to discuss various ap-proaches to patient preparation, bolustiming, contrast agent administration,and ECG gating and to describe astraightforward optimized techniquefor performance of TRO CT studies.An optimized TRO protocol shouldminimize contrast agent dose and ra-diation exposure to the patient whileproviding coronary arterial imagequality equivalent to that of a dedi-cated coronary CT angiogram, pulmo-nary arterial image quality equivalentto that of a dedicated CT pulmonaryarteriogram, and high-quality imagingof the thoracic aorta without pulsationartifact.

Patient Selection

Appropriate patient selection is crucialto the cost-effective application of TROCT (Fig 1). Patients who are at high riskfor ACS, with elevated cardiac biomar-kers or acute ECG changes, should beadmitted to the hospital and are likely tobenefit from direct triage to cardiaccatheterization for diagnostic purposesand timely intervention. In the remain-ing patients suspected of having ACS,the goal of TRO CT is to exclude thediagnosis of coronary disease or to de-fine an alternative diagnosis that mightexplain the presenting symptoms.

Patients who are likely to have ahigh burden of calcified coronaryplaque, because they have known cor-onary disease (including patients withprevious myocardial infarction,chronic angina, or a stent, as well aspatients who have undergone bypass)are less likely to benefit from the cor-onary imaging performed with TRO

CT, although the TRO study may stillbe useful with respect to the aorta,pulmonary arteries, and other in-trathoracic conditions. The degree ofcoronary disease is often overesti-mated in these patients owing toblooming of calcified plaque, such thatit is impossible to exclude clinicallysignificant coronary disease. Older pa-tients with multiple cardiac risk fac-tors are more likely to have extensivecoronary calcification (27). An inde-terminate coronary CT evaluation ismuch more likely in patients with anelevated calcium score (score ! 400–1000) (28). In such patients, a calciumscoring study may be useful prior toTRO CT to define whether the patientis a candidate for TRO CT.

An acceptable clinical history forTRO CT includes a symptom complexthat raises the suspicion of ACS, includ-ing symptoms such as chest pain, short-ness of breath, syncope or near syncopeor neck, shoulder, back, or arm painnot appearing to be musculoskeletal innature. Patients should be negative forinitial cardiac biomarkers (myoglobinand troponin-I) and should not havenew ECG changes suggestive of myocar-dial ischemia. Ideally, these patientsshould have signs, symptoms, and labo-ratory data that might be interpreted asconsistent with ACS or other causes of

Figure 1

Figure 1: Patient selection criteria for TRO CT.



chest pain, including pulmonary embo-lism and acute aortic syndrome.

In selected patients who test posi-tive for low levels of biomarkers, TROCT may be appropriate when the clini-cal impression favors pulmonary embo-lism or acute aortic syndrome or whenthere is a need to exclude ACS but thereis no immediate intention of sending thepatient for invasive cardiac catheteriza-tion. When clinical suspicion is trulylimited to ACS, a dedicated coronaryCT angiogram is preferred, as it willinvolve less contrast material and ex-pose the patient to a lower radiationdose. Age, sex, and clinical presentationare well-validated parameters that canbe used to define a population with possi-ble ACS that would be appropriate forTRO CT (29). While traditional cardiacrisk factors such as a family history ofcoronary disease, hypercholesterolemia,hypertension, and others clinical param-eters may be important long-term prog-nostic markers, such risk factors are oflimited clinical value in diagnosing ACS inthe ED setting and in triaging these pa-tients (30).

The presence of a cardiac arrhyth-mia presents a challenge for ECG-gated coronary imaging but is nolonger an absolute contraindication.Sinus bradycardia is the preferredheart rhythm for TRO CT. In the ab-sence of a clinical contraindication, a$-blocker should be administeredprior to TRO CT. Both heart rateand ectopy are reduced after treat-ment with an intravenous $-blocker(31,32).

New CT technology provides im-proved temporal resolution, with thecapability of scanning the entire heartin one to two heartbeats (as comparedwith four to five beats for most 64-section scanners). This new technol-ogy has reduced both the requiredphase window for diagnostic imagingof the coronary arteries and the ef-fects of variability in heart rhythm oncoronary image quality (33).

The decision as to whether a patientshould be excluded from undergoing TROCT owing to a cardiac arrhythmia mustbe based on an assessment of the magni-tude of the arrhythmia and the specific

capabilities of the scanner that will beused for the study. A regular heart rate ofup to 80 beats per minute is no longer acontraindication for many new scanners,including dual-source scanners and sin-gle-source scanners with a gantry rotationtime faster than 300 msec. An irregulartachyarrhythmia poses a more difficultproblem, but the degree of contraindica-tion depends on the frequency of ectopicbeats.

Allergies to contrast material andrenal insufficiency are relative contra-indications to administration of iodin-ated contrast material for TRO stud-ies. The presence of asthma, acuteheart failure, severe cardiomyopathy,or hypotension may limit the use$-blockers to control heart rate andthus may reduce the quality of theTRO CT images. A history of recentcocaine use or a drug screen positivefor cocaine is also a relative contrain-dication to the use of $-blockers forthe scan (34), although this contrain-dication remains controversial (35).Recent use of a phosphodiesterase in-hibitor is a relative contraindication tothe administration of nitroglycerin forcoronary vasodilatation during CT,but this does not represent a contra-indication to TRO CT.

CT Hardware and Radiation Issues

TRO CT requires a longer scanninglength than does dedicated coronary CTangiography. A mean scanning length of20 cm is required to image the chestfrom above the aortic arch through thecaudal aspect of the heart. To performthis scan during a single breath hold,the scanner should be capable of im-aging the required volume with anECG-gated technique in no more than15 seconds. This requirement limitsTRO CT studies to scanners with atleast 64 detector rows.

TRO studies are associated with ahigher radiation dose when comparedwith dedicated coronary CT angiogra-phy examinations because of the longerscanning length. Our typical scan pa-rameters include a tube voltage of 120kVp and a mean effective tube currentof 600 mAs per section (where effective

milliampere-seconds equals tube milli-amperes times gantry rotation time di-vided by pitch). Heavier patients weigh-ing over 200 lbs (!91 kg) are scannedwith higher tube current of 800–1000mAs, on the basis of a subjective esti-mate of patient body habitus by the at-tending radiologist. In our experience,mean effective TRO radiation dose forpatients evaluated in helical scan modewithout tube current modulation aver-ages 18 mSv and is decreased to 8.75mSv among patients evaluated withtube current modulation (36). Mean ef-fective tube current and/or tube voltagecan be decreased in smaller patients toreduce radiation dose.

Until recently, all coronary CT an-giography studies have been performedwith a helical acquisition (with or with-out tube current modulation). In pa-tients with a very stable heart rate,newer scanners can acquire a TRO CTstudy with prospective ECG gating byusing the “step-and-shoot” axial mode tofurther reduce radiation dose to 5–6mSv. Prospective ECG gating should bereserved only for patients with a verystable heart rate, since any change incardiac rhythm will either prolong thescan time (as the scanner waits for thenext “normal” heartbeat) or result indegraded image quality from cardiacmotion.

Images obtained with prospectiveECG gating are more sensitive to minorvariations in heart rate and cannot pro-vide information about cardiac functionand regional wall motion. Nonetheless,in appropriately selected patients evalu-ated with proper attention to technique,prospective ECG gating of coronary CTangiography can be used to reduce radi-ation dose while maintaining imagequality (37,38).

CT imaging, and coronary CT inparticular, has been criticized as an im-portant source of radiation exposureto the population (39,40). Recent ad-vances in CT technology, however, al-low a dramatic decrease in radiationdose with coronary CT. The effectiveradiation dose for a TRO CT scan with astate-of-the-art scanner compares fa-vorably with the dose of a nuclear stresstest, the dose from which has been re-



ported to range from 10 to 17 mSv (41).When one considers that the conven-tional work-up in a patient with chestpain presenting to the ED is likely toinclude a negative nuclear stress testresult, as well as another diagnostic ra-diologic examination such as a chest CTor a ventilation-perfusion study, appli-cation of a TRO CT study to an appro-priate patient population may actuallyreduce the per-patient radiation expo-sure for diagnostic studies during theED evaluation.

TRO CT Technique and Image Quality

High-quality coronary imaging is essen-tial to distinguish patients with coronarydisease from those in whom ACS maybe excluded. Careful attention to pa-tient preparation (Fig 2), scan tech-nique (Fig 3), and injection technique(Table) will result in optimal, homoge-neous aortic, coronary, and pulmonaryarterial opacification and in coronaryimage quality that is equal to that ob-tained with dedicated coronary CT an-giography (42). ECG gating of TRO

studies eliminates motion artifact re-lated to cardiac pulsation and, there-fore, provides superior definition of thepulmonary arterial tree as comparedwith dedicated pulmonary CT angiogra-phy without ECG gating. The discussionthat follows is directed primarily towardperformance of the TRO study with a 64detector–row scanner.

Patient Preparation and Monitoring

To minimize ectopy in the cardiacrhythm during coronary CT angiogra-phy, patients should refrain from stimu-lants such as caffeine on the day of theexamination. However, unlike typicaloutpatients scheduled for coronary CTangiography, TRO CT patients typicallypresent directly from the ED and cannotbe instructed to modify their dietary in-take prior to the scan. Nonetheless, theuse of intravenous $-blockers allowsrapid control of heart rate and reduc-tion of ectopy in ED patients sent forTRO CT.

Adequate intravenous access isnecessary to deliver a rapid contrast

agent bolus for coronary CT angiogra-phy. An 18 –20-gauge intravenouscatheter is placed into a large vein inthe antecubital fossa. The intravenouscatheter should be tested with a rapidsaline flush to ascertain that there isno extravasation and that the patientdoes not experience pain during injec-tion. A painful intravenous injectioncan result in a sudden change of heartrate during contrast agent administra-tion, with degradation of image qual-ity.

To reduce inadvertent compressionof the subclavian vein during contrastagent injection, the arm with the intra-venous catheter should not be extendedabove the patient’s head. With the pa-tient lying in a supine position, I preferto extend the arm with the intravenousline directly in front of the patient. Tomaintain the patient’s comfort and tokeep the arm out of the gantry, the ex-tended hand is rested on the gantry dur-ing the scan.

ECG leads are positioned above andbelow the level of the scan to preventstreak artifact. The ECG trace should be

Figures 2, 3

Figure 2: Technique for patient preparation for TRO CT. IV # intravenous. Figure 3: Technique for performing TRO CT.

Contrast Material Injection Protocols for Coronary CT Angiography and TRO CT

Examination Type First Phase Second Phase Injection Rate

Dedicated coronary CT angiography 70 mL of iodine 350 40 mL of saline 5.5 mL/secTRO CT 70 mL of iodine 350 50 mL of diluted contrast agent (25 mL iodine 350 plus 25 mL saline) 5.0 mL/secExtended (18–20 sec) TRO CT (see Fig 4) 80 mL of iodine 350 70 mL of diluted contrast agent (35 mL iodine 350 plus 35 mL saline) 5.0 mL/sec

Note.—Iodine 350 # 350 mg of iodine per milliliter.



evaluated immediately before the scan,with the patient’s arms raised into theposition that will be used for the scan. Itis important to be certain that the ECGleads will not be pulled off of the patientwhen the table moves for the scan. Aclearly defined R-wave is necessary toensure adequate ECG gating. If neces-sary, leads should be repositioned toobtain a clearly defined R-wave.

Baseline heart rate and blood pres-sure are obtained before and during ad-ministration of $-blockers and after theprocedure. A stable blood pressureshould be documented before the patientis returned to the ED.

The ideal heart rate for ECG-gatedstudies is a slow regular rhythm, usuallya sinus bradycardia at 50–60 beats perminute (43). Although precise control ofheart rate may be less critical with dual-tube CT scanners or with newer scannersthat have a faster gantry rotation and im-proved temporal resolution, image qual-ity is optimized and radiation dose is min-imized with a regular cardiac rhythm.

Oral $-blockers may be given inthe ED at least 1 hour before the scanfor control of heart rate. However,heart rate often increases with thelevel of anxiety when the patient isplaced on the CT table. I prefer toadminister metoprolol tartrate intra-venously when the patient is on the CTtable. Intravenous metoprolol has anonset of action within 1–3 minutes (ascompared with 1–2 hours for an orallyadministered dose) and allows bettertitration of heart rate prior to contrastagent injection. $-Blockers may altervascular tone, cardiac rhythm, andmyocardial contractility and can pro-mote bronchospasm. Metoprolol shouldbe used with caution, and may be contra-indicated in patients with heart block, un-compensated heart failure and asthma.

In the interest of improving patientthroughput, administration of metopro-lol is performed during acquisition ofthe scout topogram and setup of thebolus-tracking images before coronaryCT angiography. Patients who arrivewith a heart rate in the range of 60–65beats per minute are given an initialintravenous dose of 2.5 mg of metopro-lol. Patients who arrive with a heart rate

Figure 4

Figure 4: TRO CT angiogram of neck and chest in 24-year-old woman with history of Marfan syndrome whopresented with acute onset of chest pain radiating into the neck. Clinical suspicion was high for aortic dissection withpossible extension into coronary arteries or great arteries in the neck. Scan was obtained with prospective ECG gat-ing and 120 mL of iodinated contrast material. First phase of injection was increased to 80 mL of contrast agent whilesecond phase was increased to 40 mL of contrast agent plus 40 mL of saline to compensate for increased scanningtime needed to include carotid arteries. (a) Coronal slab maximum intensity projection (MIP) image demonstratesenhancement of aorta, pulmonary arteries, and great vessels extending from the aortic arch with no dissection.(b) Oblique slab MIP image demonstrates normal aortic arch and descending thoracic aorta. (c) Oblique coronalslab MIP image demonstrates normal left ventricular outflow tract extending into proximal part of aortic arch. How-ever, there is air in tissues of the neck surrounding great vessels (arrows). (d) Coronal slab MIP image through thetrachea demonstrates extensive emphysematous changes in mediastinum. On further questioning, patient com-plained of an episode of intense coughing the previous night, just before symptoms began. Mediastinal air wasthought to represent the cause of symptoms and was likely related to a ruptured pulmonary bleb (never seen). Patientrecovered without further interventional therapy.



faster than 65 beats per minute aregiven an initial intravenous dose of 5mg. After allowing several minutes toobserve the effect of the first dose, addi-tional doses of 5 mg are administeredevery 3–5 minutes until the target heartrate is achieved. A minority of patientswith acute chest pain and/or shortnessof breath will present with a tachycardiathat does not respond to $ blockade. It isunusual to administer more than 20 mg ofmetoprolol since patients who do not re-spond with a lower heart rate after thefirst 10–20 mg are unlikely to respond toa higher dose. For patients who continueto respond slowly to administration of anintravenous $-blocker, the maximum in-travenous dose that I will administer is 30mg. The patient’s heart rate and bloodpressure are monitored after every dose,and no further $-blocker is given if sys-tolic pressure declines below 100 mm Hg.

To achieve maximum coronary va-sodilatation for the study, sublingual ni-troglycerin is administered 2–3 minutesbefore the start of TRO CT (44). A re-cent study suggests that pretreatmentwith sublingual nitroglycerin may im-prove the diagnostic accuracy of coro-nary CT (45). As with metoprolol, nitro-glycerin is not administered if the sys-tolic blood pressure declines below 100mm Hg.

Although nitroglycerin and $-block-ers can combine to cause hypotension,this is not generally a problem on theCT table with the patient in the supineposition. Nitroglycerin does result in a re-flex tachycardia, but this is not generally aproblem when patients have received a$-blocker prior to the study. Relative con-traindications to nitroglycerin includeclinical scenarios such as hypovolemiaand idiopathic hypertrophic subaortic ste-nosis or recent use of a phosphodiester-ase inhibitor, where nitroglycerin may in-duce profound hypotension.

After administration of nitroglyc-erin, I generally practice the breath holdwith the patient. The patient is in-structed to take a slow small breath andto hold it for 15 seconds. A large in-spiratory effort draws more unopacifiedblood from the inferior vena cava andmay reduce the level of intravascularopacification during scanning (46). If

the patient takes a large practicebreath, I ask the patient to repeat thepractice exercise and take a smallerbreath. Although the breath hold mayseem like a trivial detail, it is impor-tant to practice the breath hold so thatthe patient is prepared for the lengthof the breath hold that will be requiredfor the scan.

Setting Up the Scan

The TRO CT examination must include theentire thoracic aorta, as well as the heart.On the basis of inspection of the scout to-pogram, TRO scans are programmed tostart 1 cm above the aortic arch, usually atthe inferior margins of the clavicular heads(Fig 5). Because radiation dose to the pa-

Figure 5

Figure 5: Scan setup and bolus-tracking images. (a) Coronal scout topogram. Scan levels are plannedwithin the green rectangle. Starting level for scan is at the inferior margin of the clavicular heads, just above theaortic arch. Inferior margin of the scan is set below the base of the heart. To limit radiation to the patient, CTangiographic acquisition is monitored in real time and manually stopped as soon as the base of the heart isimaged. Note that setup specifies two reconstruction fields of view. Smaller (25-cm) field of view is used forevaluation of aorta and coronary anatomy. Larger field of view is used to evaluate pulmonary arteries, lungs,and chest wall. (b) Precontrast bolus-tracking image. Table height and patient position are adjusted so thatheart is centered in the scanning area. CT resolution is maximized at center of the gantry. A region of interest(circle) is defined in the left atrium. (c) Low-dose bolus-tracking images are obtained every 2 seconds, begin-ning 5 seconds after start of the contrast agent injection. This image demonstrates early opacification of leftatrium. CT angiography was manually started at this time. (d) Graph shows that scanning is programmed tobegin 5 seconds after initial opacification of left atrium reaches 100 HU above baseline (horizontal line at 150HU in this plot). To maximize use of the contrast dose, however, scanning is manually started by the technolo-gist as soon as contrast material enters left atrium. Manual start time is approximately 2 seconds earlier thanthe density would have triggered an automatic start to the scan. The programmed start line serves as a backupin the event the technologist fails to start the scan earlier.



tient is directly proportional to the length ofthe scan, the lung apices above the level ofthe aortic arch are not included. Although5% of patients with pulmonary embolismhave an isolated upper lobe embolus (47),an isolated subsegmental pulmonary embo-lus above the level of the aortic arch is ex-tremely uncommon and is very unlikely tobe detected with CT angiography. In theearly days of spiral CT, before the introduc-tion of multisection CT scanners, dedicatedpulmonary CT angiography was generallylimited to a distance of 10–12 cm from theaortic arch down through the inferior pul-monary veins (48). Excluding the apicesfrom TRO CT reduces the scan length byapproximately 4–5 cm, which we estimateis associated with a 15%–20% reduction ineffective radiation dose.

Many centers traditionally acquire CTpulmonary arteriography studies in thecaudal-to-cranial direction to reduce theeffect of respiratory motion in the lowerlobes (49,50). Others have suggested ad-vantages to scanning in the cranial-to-caudal direction for CT pulmonary arte-riography (51). As discussed below, thecranial-to-caudal scanning direction is fa-vored for TRO studies because of addi-tional considerations related to timing ofthe contrast agent injection and patientheart rate. Although scanning is pro-grammed to continue through the base ofthe heart, the TRO CT angiography ac-quisition is monitored in real time and ismanually stopped as soon as the base ofthe heart is imaged. Manual stopping ofthe scan at the heart base can reducescanning length by 1.5–2.0 cm. The radi-ation dose estimates of 8.75–18 mSv citedearlier were obtained before we routinelybegan terminating the scan in real timeonce the base of the heart was imaged.We estimate that the use of real-timemonitoring to reduce acquisition lengthcan reduce the radiation dose by another7%–10%.

Contrast Agent Injection and Timing ofImage Acquisition

The goal of contrast agent administra-tion for dedicated coronary CT angiog-raphy is to maintain a high level of en-hancement in the coronary arteries(52). Coronary opacification demon-

strates a strong correlation with therate of injection, as well as with theconcentration of iodine in the contrastmaterial (53,54). For TRO CT studies, areasonable enhancement goal is an at-tenuation level higher than 300 HU inthe coronary arteries and higher than200 HU in the pulmonary arteries.

Standard injection techniques usedfor dedicated coronary CT angiographyresult in suboptimal opacification ofthe pulmonary arterial circulation(55). Opacification of the pulmonaryarteries requires an extended contrastagent injection to maintain contrast onthe right side of the heart during thescan. However, it is important not tohave full strength contrast material inthe superior vena cava at the time of thescan, because this can cause a streakartifact that may limit image quality.

Studies of aortic CT angiographyhave demonstrated an advantage to thebiphasic injection of contrast material(56). A high-injection-rate uniphasic in-jection of contrast material results in apeak of contrast enhancement during ashort interval, with attenuation minimaat the beginning and/or end of the ac-quisition. A biphasic injection protocolcan be tailored to provide a more homo-geneous enhancement profile over time.To optimize both coronary and pulmo-nary arterial enhancement for TRO CT,I prefer a biphasic contrast agent injec-tion protocol to provide an intense ho-mogeneous level of contrast enhance-ment in the left-sided circulation (aortaand coronary arteries) with a slightlylower homogeneous level of enhance-ment in the right-sided circulation. Arapid flow rate is maintained through-out the injection to minimize the effectof venous return from the inferior venacava.

Although preheating of contrast ma-terial is not required to obtain a rapidflow rate, preheating up to a tempera-ture of 37°C prior to injection reducesthe viscosity of the contrast materialand facilitates a rapid flow rate at lowerinjection pressures (57).

The biphasic injection is timed sothat the first phase of the injectionopacifies the left heart while the secondphase opacifies the right heart during

TRO CT. More specifically, for a 64-detector scanner, the biphasic injectionconsists of 70 mL of undiluted contrastmaterial (350 mg of iodine per milliliter)followed by 25 mL of contrast materialdiluted with 25 mL of saline, all injectedat 5.0 mL/sec (Table). To make efficientuse of the contrast agent bolus, imagingis triggered on the basis of opacificationof the left atrium, which begins 2–3 sec-onds before opacification of the de-scending aorta. Scanning begins 5 sec-onds after the contrast agent enters theleft atrium (Fig 5), so that the aorta andcoronary arteries are in the plateauphase of peak enhancement during theCT angiographic acquisition. The injec-tion volume and rate are optimized for ascanning time of approximately 14–15seconds. In the event of a significantlylonger scan time—such as that whichmay occur if there is an arrhythmiawhen using prospective ECG gating orwhen the CT angiogram is extended inlength—the injection would need to beprolonged to ensure adequate pulmo-nary opacification (Table). When theTRO study is properly timed, the firstphase of the injection opacifies the cor-onary arteries during image acquisition(Figs 6, 7), while the second phase ofthe injection provides simultaneous ho-mogeneous enhancement of the pulmo-nary arteries (Figs 8, 9). The thoracicaorta is also homogeneously enhancedand optimally evaluated (Figs 10, 11). Areview of our TRO studies demon-strates that this technique provides amean enhancement level of 300–350HU in the aorta, pulmonary arteries,and coronary arteries (42).

Several variations of this biphasicinjection protocol have been proposed.In one variation, a biphasic injection isused with undiluted contrast material(320 mg of iodine per milliliter), butthe flow rate starts at 5 mL/sec and isthen reduced to 3 mL/sec to avoidoverloading the right heart and supe-rior vena cava with dense contrast ma-terial (58,59). The reduced flow rate inthe second phase of the injection pro-vides a longer injection time and re-duces streak artifact from the superiorvena cava. A 50-mL saline flush is ap-plied as a third phase to flush the con-



trast agent that remains in the armveins into the right heart.

An alternative triphasic techniqueinvolves a first phase with 50 mL of un-diluted contrast material (350 mg of io-dine per milliliter) followed by 50 mL of60% contrast agent–to–40% saline and

then by 30 mL of saline, each injected at4.5 mL/sec (60). Contrast material isdiluted in the second phase of the injec-tion to reduce streak artifact from thesuperior vena cava. The use of dilutedcontrast material with a faster flow ratehas a theoretic advantage over the use

of full-strength contrast material at aslower rate. The faster injection rateresults in greater filling of the venoussystem from the injection, greater fillingof the right atrium from the superiorvena cava, and less variation related tounopacified venous flow from the infe-rior vena cava.

Although I do use a saline flush fordedicated coronary CT angiography, Ido not use a saline flush for TRO CTstudies. In my experience the salineflush can result in complete washout ofcontrast material from the right heart

Figure 6

Figure 6: TRO CT angiogram in 37-year-old woman with no relevant cardiac history who presented withsudden onset of chest pain while at work. TRO demonstrated a smooth 75% stenosis of the left anterior de-scending artery (arrow). Patient was treated with angioplasty. Slab MIP images of (a) left anterior descendingartery in long axis of the aortic root and (b) left anterior descending artery in orthogonal obliquity in the shortaxis of the aortic root.

Figure 7

Figure 7: TRO CT angiogram in 51-year-old athletic man with no relevant cardiac history who presentedwith atypical chest pain while resting at home. TRO CT images demonstrate irregular narrowing of left anteriordescending artery (LAD; arrow). Patient was treated with angioplasty. (Bright spots projecting just below LADcorrespond to contrast material in interstices of right ventricle between trabeculations adjacent to interventric-ular septum. These are imaged along with septum owing to the thickness of the slab MIP projection.) (a) SlabMIP image of LAD along the long axis of the left ventricle. (b) Slab MIP image of LAD in an orthogonal obliq-uity along short axis of the aortic root.

Figure 8

Figure 8: TRO CT angiogram in 31-year-oldwoman with chest pain that was atypical for anginabut without severe shortness of breath. TRO CTdemonstrates bilateral pulmonary embolism withnormal aorta and coronary arteries. (a) Obliquecoronal slab MIP image demonstrates largethrombus in right pulmonary and interlobar pul-monary arteries with extension into right lowerlobe segmental branches. (b) Axial slab MIP im-age demonstrates bilateral pulmonary embolismwith a smaller thrombus also present in left pul-monary artery.



and result in a nondiagnostic pulmo-nary arteriogram when the acquisitiontime is prolonged or when the start ofimage acquisition is delayed by a slowright-to-left contrast material transittime. I prefer to leave some of thediluted contrast material from the sec-ond phase of injection in the patient’sarm rather than risk using a salineflush that may wash the contrast out ofthe pulmonary circulation. Whenscanning time must be increased foradditional z-axis coverage, the injec-tion may be prolonged with additionalcontrast material. (Fig 4 demonstratesa carotid plus TRO CT study requiring20 seconds for acquisition.)

There are different approachesamong centers to the directionalityused for acquiring TRO studies. If thepatient is unable to told his or herbreath for the full acquisition, it is bestto use caudal-to-cranial scanning, sothat the heart is imaged before thepatient begins to breathe. The cranial-to-caudal scanning direction used inmy protocol introduces an additional 5seconds between the breath hold andthe cardiac portion of the scan. In myexperience, the heart rate is morevariable when the patient first takes abreath and tends to plateau at a rateslightly below the baseline level at5–15 seconds after the patient takes abreath. For the majority of patientswho have no trouble holding theirbreath, scanning in a cranial-to-caudaldirection affords better coronary im-age quality by imaging the heart dur-ing this respiratory-induced plateau ofthe heart rate. Furthermore, sincecontrast material must pass throughthe pulmonary arterial tree before itenters the coronary arteries, optimalenhancement in the pulmonary circu-lation is achieved before optimal cor-onary arterial enhancement. A scanperformed in the cranial-to-caudal di-rection can be started a few secondsearlier than a scan performed in thecaudal-to-cranial direction and cantake advantage of the earlier enhance-ment in the pulmonary circulation toreduce the overall contrast materialload.

Timing of contrast agent injection

Figure 9

Figure 9: (a) Sagittal and (b) coronal slab MIP images from TRO CT angiogram in 40-year-old man withchest pain and tachycardia show left upper lobe pulmonary embolus extending into an apical segmentalbranch of left pulmonary artery (arrow). Coronary arteries and aorta were normal.

Figure 10

Figure 10: TRO CT angiogram in 79-year-old woman with recent onset of vague chest discomfort. TRO CTdemonstrates type B aortic dissection extending from distal aortic arch to descending aorta. (a) Oblique slabMIP image shows entire aortic arch with dissection flap extending into the abdomen. (b) Coronal slab MIPimage again demonstrates the dissection with asymmetric enhancement of true and false lumina.



and image acquisition is a critical com-ponent of the study. When scanning istimed to begin 5 seconds after contrastmaterial first appears in the left atrium,image acquisition corresponds to the“peak plateau” of contrast enhancementin the coronary circulation. The entireTRO study can be performed with 100mL of contrast material with current64-section scanners. Even faster scan-ning times with shorter injections maybe achieved with newer CT systems thatuse 256 or 320 detector rows.

Image Interpretation

Interpretation of TRO CT studies in-cludes evaluation of the coronary arter-ies, as well as of other vascular andnonvascular structures. Most of thenoncoronary structures are evaluatedwith axial 3–5-mm-thick images. Thin-ner sections may be obtained whenneeded for further evaluation of abnor-malities detected on the 3–5-mm sec-tions. For patients without substantialcoronary disease, I prefer to evaluate

the coronary arteries with a combina-tion of thin-section (0.6–0.8-mm) axialimages and slab (5-mm) MIP recon-structions (Figs 6, 7).

Slab MIP images are reconstructedin real time during the interpretationsession so that no additional technolo-gist effort is required. It is important tobe certain that each segment of eachcoronary artery is evaluated in multipleprojections by rotating the slab MIP im-ages on a workstation. The aorta andpulmonary arteries are evaluated with

Figure 11

Figure 11: TRO CT angiogram in a 74-year-old man with history of coronary disease and pulmonary embolism who presented with progressive chest pain over 6months that became acutely worse on the day of presentation. Diagnosis of pulmonary embolism, a primary clinical consideration, was excluded at CT. (a) Oblique slabMIP image demonstrates ectatic aortic arch with atherosclerotic calcification and suggestion of two areas of aortic ulceration (arrowheads). Arrow # proximal right coro-nary artery. (b) A different obliquity on the slab MIP shows one of the ulcers, which measured 10 mm in width and extended 8 mm beyond the expected contour of theaorta (arrowhead). Comparison with earlier studies (not shown) demonstrated no notable change in appearance of the aortic ulcers. (c) Left anterior oblique slab MIPshows high-grade stenosis or possible proximal occlusion (arrow) in proximal right coronary artery. (d) Axial slab MIP confirms high-grade stenosis or possible occlu-sion (arrow) in proximal right coronary artery. (e, f) Curved MIP images of right coronary artery confirm the lesion (arrow) in proximal right coronary artery. Presence ofischemia in inferior wall and inferoseptum was confirmed with stress testing.



the same viewing tools as are used forthe coronary circulation (Figs 8–10).For more complicated cases where vas-cular calcifications or complex plaqueare present, vessel-tracking software isuseful to create curved MIP views thatfacilitate visualization of the coronaryarteries in multiple planes (Fig 11).

For those patients scanned with aconventional helical acquisition, mul-tiphase reconstructions are obtained at10% increments throughout the cardiaccycle. Cardiac wall motion is evaluatedon a workstation in standard echocar-diographic projections, including four-,three-, and two-chamber and short-axisviews. Any abnormality of regional wallmotion should be correlated with thecorresponding vessel on the coronaryCT angiogram to search for and/or con-firm a stenosis. Identification of a wallmotion abnormality can confirm that acoronary artery lesion is functionallysignificant.

In a patient who does not have ahigh pretest likelihood of obstructivecoronary disease, normal findings on acoronary CT angiogram serve to elimi-nate any further need for a diagnosticcoronary work-up (61). In the practiceat my institution, the diagnosis of ACS iseffectively ruled out by a TRO CT scanthat demonstrates normal coronary ar-teries or no more than minimal coro-nary disease ("25% stenosis). Whenthere is more than minimal disease, fur-ther cardiac work-up may be appropri-ate. Although both calcified and noncal-cified plaques may be associated withACS, mixed calcified-and-noncalcifiedplaques with a predominantly noncalci-fied component are better correlatedwith ACS (62). The authors of one re-cent study (63) suggest that positivevascular remodeling, low-attenuationplaque, and spotty calcification in coro-nary plaques are associated with ACS.Unfortunately, the anatomic informa-tion provided by coronary CT angiogra-phy in the presence of substantial coro-nary disease may not be definitive forthe functional diagnosis of ACS. Clinicaltrials are needed to define whetherthere is an appropriate appearance ofcoronary plaque and/or degree of coro-nary stenosis that can definitively in-

clude or exclude acute coronary syn-drome. In the future, perfusion CT willbe combined with coronary CT angiog-raphy to provide a more accurate diag-nostic tool for evaluation of ACS.

Conclusion

The TRO CT examination can be a pow-erful tool for evaluation and triage ofpatients with a low to moderate risk ofACS in whom diagnostic catheterizationis not indicated. However, unlike mostCT studies that can be performed by atechnologist using a simple protocol,TRO CT studies require more individu-alized attention. Careful considerationregarding patient selection, patientpreparation, and injection and scanningtechniques will result in high-qualityTRO CT studies to evaluate the aorta,coronary circulation, pulmonary arter-ies, and adjacent intrathoracic condi-tions. When compared with conven-tional management of acute chest painin the ED, appropriate application ofTRO CT can reduce (a) time for patienttriage, (b) number of required diagnos-tic tests, (c) ED costs, and (d) radiationexposure to the patient.

Acknowledgments: The author acknowledgesthe assistance of Anish Koka, MD, and KevinTakakuwa, MD, who reviewed the manuscriptand provided insightful suggestions.

References1. Pitts SR, Niska RW, Xu J, Burt CW. National

Hospital Ambulatory Medical Care Survey:2006 emergency department summary. NatlHealth Stat Report 2008;7:1–38.

2. Pope JH, Selker HP. Acute coronary syn-dromes in the emergency department: diag-nostic characteristics, tests, and challenges.Cardiol Clin 2005;23(4):423–451, v–vi.

3. Pope JH, Aufderheide TP, Ruthazer R, et al.Missed diagnoses of acute cardiac ischemiain the emergency department. N Engl J Med2000;342(16):1163–1170.

4. Christenson J, Innes G, McKnight D, et al.Safety and efficiency of emergency depart-ment assessment of chest discomfort. CMAJ2004;170(12):1803–1807.

5. Schull MJ, Vermeulen MJ, Stukel TA. Therisk of missed diagnosis of acute myocardialinfarction associated with emergency de-partment volume. Ann Emerg Med 2006;48(6):647–655.

6. Rusnak RA, Stair TO, Hansen K, Fastow JS.Litigation against the emergency physician:common features in cases of missed myocar-dial infarction. Ann Emerg Med 1989;18(10):1029–1034.

7. Karcz A, Korn R, Burke MC, et al. Malprac-tice claims against emergency physicians inMassachusetts: 1975–1993. Am J EmergMed 1996;14(4):341–345.

8. Katz DA, Williams GC, Brown RL, et al.Emergency physicians’ fear of malpractice inevaluating patients with possible acute car-diac ischemia. Ann Emerg Med 2005;46(6):525–533.

9. Storrow AB, Gibler WB. Chest pain centers:diagnosis of acute coronary syndromes. AnnEmerg Med 2000;35(5):449–461.

10. Meijer AB, O YL, Geleijns J, Kroft LJ. Meta-analysis of 40- and 64-MDCT angiographyfor assessing coronary artery stenosis. AJRAm J Roentgenol 2008;191(6):1667–1675.

11. Marano R, De Cobelli F, Floriani I, et al. Italianmulticenter, prospective study to evaluate thenegative predictive value of 16- and 64-sliceMDCT imaging in patients scheduled for coro-nary angiography (NIMISCAD-Non InvasiveMulticenter Italian Study for Coronary ArteryDisease). Eur Radiol 2009;19(5):1114–1123.

12. Hamon M, Morello R, Riddell JW, HamonM. Coronary arteries: diagnostic perfor-mance of 16- versus 64-section spiral CTcompared with invasive coronary angiogra-phy—meta-analysis. Radiology 2007;245(3):720–731.

13. Gallagher MJ, Raff GL. Use of multislice CTfor the evaluation of emergency room pa-tients with chest pain: the so-called “triplerule-out”. Catheter Cardiovasc Interv 2008;71(1):92–99.

14. Thomas J, Rideau AM, Paulson EK, BissetGS 3rd. Emergency department imaging:current practice. J Am Coll Radiol 2008;5(7):811–816e2.

15. Budoff MJ, Dowe D, Jollis JG, et al. Diagnos-tic performance of 64-multidetector row cor-onary computed tomographic angiographyfor evaluation of coronary artery stenosis inindividuals without known coronary arterydisease: results from the prospective multi-center ACCURACY (Assessment by Coro-nary Computed Tomographic Angiographyof Individuals Undergoing Invasive CoronaryAngiography) trial. J Am Coll Cardiol 2008;52(21):1724–1732.

16. Miller JM, Rochitte CE, Dewey M, et al.Diagnostic performance of coronary angiog-raphy by 64-row CT. N Engl J Med 2008;359(22):2324–2336.

17. Khare RK, Powell ES, Venkatesh AK,



Courtney DM. Diagnostic uncertainty andcosts associated with current emergency de-partment evaluation of low risk chest pain.Crit Pathw Cardiol 2008;7(3):191–196.

18. Gallagher MJ, Ross MA, Raff GL, GoldsteinJA, O’Neill WW, O’Neil B. The diagnosticaccuracy of 64-slice computed tomographycoronary angiography compared with stressnuclear imaging in emergency departmentlow-risk chest pain patients. Ann Emerg Med2007;49(2):125–136.

19. Rubinshtein R, Halon DA, Gaspar T, et al.Usefulness of 64-slice cardiac computed to-mographic angiography for diagnosing acutecoronary syndromes and predicting clinicaloutcome in emergency department patientswith chest pain of uncertain origin. Circula-tion 2007;115(13):1762–1768.

20. Hollander JE, Chang AM, Shofer FS,McCusker CM, Baxt WG, Litt HI. Coro-nary computed tomographic angiographyfor rapid discharge of low-risk patientswith potential acute coronary syndromes.Ann Emerg Med 2009;53(3):295–304.

21. Chang SA, Choi SI, Choi EK, et al. Useful-ness of 64-slice multidetector computed to-mography as an initial diagnostic approachin patients with acute chest pain. Am Heart J2008;156(2):375–383.

22. Khare RK, Courtney DM, Powell ES,Venkatesh AK, Lee TA. Sixty-four-slice com-puted tomography of the coronary arteries:cost-effectiveness analysis of patients pre-senting to the emergency department withlow-risk chest pain. Acad Emerg Med 2008;15(7):623–632.

23. Chang AM, Shofer FS, Weiner MG, et al. Ac-tual financial comparison of four strategies toevaluate patients with potential acute coronarysyndromes. Acad Emerg Med 2008;15(7):649–655.

24. Takakuwa KM, Halpern EJ. Evaluation of a“triple rule-out” coronary CT angiographyprotocol: use of 64-section CT in low-to-moderate risk emergency department pa-tients suspected of having acute coronarysyndrome. Radiology 2008;248(2):438–446.

25. Ladapo JA, Hoffmann U, Bamberg F, et al.Cost-effectiveness of coronary MDCT in thetriage of patients with acute chest pain. AJRAm J Roentgenol 2008;191(2):455–463.

26. Goldstein JA, Gallagher MJ, O’Neill WW,Ross MA, O’Neil BJ, Raff GL. A randomizedcontrolled trial of multi-slice coronary com-puted tomography for evaluation of acutechest pain. J Am Coll Cardiol 2007;49(8):863–871.

27. Wexler L, Brundage B, Crouse J, et al.Coronary artery calcification: pathophysi-ology, epidemiology, imaging methods, and

clinical implications—a statement forhealth professionals from the AmericanHeart Association Writing Group. Circula-tion 1996;94(5):1175–1192.

28. Hecht HS, Bhatti T. How much calcium is toomuch calcium for coronary computerized to-mographic angiography? J Cardiovasc Com-put Tomogr 2008;2(3):183–187.

29. Fraker TD Jr, Fihn SD, 2002 Chronic StableAngina Writing Committee, et al. 2007chronic angina focused update of the ACC/AHA 2002 guidelines for the management ofpatients with chronic stable angina: a reportof the American College of Cardiology/American Heart Association Task Force onPractice Guidelines Writing Group to de-velop the focused update of the 2002 guide-lines for the management of patients withchronic stable angina. J Am Coll Cardiol2007;50(23):2264–2274.

30. Han JH, Lindsell CJ, Storrow AB, et al. Therole of cardiac risk factor burden in diagnos-ing acute coronary syndromes in the emer-gency department setting. Ann Emerg Med2007;49(2):145–152.

31. Fenster PE, Quan SF, Hanson CD, CoakerLA. Suppression of ventricular ectopy withintravenous metoprolol in patients withchronic obstructive pulmonary disease. CritCare Med 1984;12(1):29–32.

32. Quan SF, Fenster PE, Hanson CD, CoakerLA, Basista MP. Suppression of atrial ectopywith intravenous metoprolol in chronic ob-structive pulmonary disease patients. J ClinPharmacol 1983;23(8:–9):341–347.

33. Steigner ML, Otero HJ, Cai T, et al. Narrow-ing the phase window width in prospectivelyECG-gated single heart beat 320-detectorrow coronary CT angiography. Int J Cardio-vasc Imaging 2009;25(1):85–90.

34. Lange RA, Hillis LD. Cardiovascular compli-cations of cocaine use. N Engl J Med 2001;345(5):351–358. [Published correction ap-pears in N Engl J Med 2001;345(19):1432.]

35. Dattilo PB, Hailpern SM, Fearon K, Sohal D,Nordin C. Beta-blockers are associated withreduced risk of myocardial infarction aftercocaine use. Ann Emerg Med 2008;51(2):117–125.

36. Takakuwa KM, Halpern EJ, Gingold EL,Levin DC, Shofer FS. Radiation dose in a“triple rule-out” coronary CT angiographyprotocol of emergency department patientsusing 64-MDCT: the impact of ECG-basedtube current modulation on age, sex, andbody mass index. AJR Am J Roentgenol2009;192(4):866–872.

37. Hirai N, Horiguchi J, Fujioka C, et al. Pro-spective versus retrospective ECG-gated 64-detector coronary CT angiography: assess-

ment of image quality, stenosis, and radia-tion dose. Radiology 2008;248(2):424–430.

38. Shuman WP, Branch KR, May JM, et al.Prospective versus retrospective ECG gatingfor 64-detector CT of the coronary arteries:comparison of image quality and patient ra-diation dose. Radiology 2008;248(2):431–437.

39. Brenner DJ, Hall EJ. Computed tomog-raphy: an increasing source of radiation ex-posure. N Engl J Med 2007;357(22):2277–2284.

40. Einstein AJ, Henzlova MJ, Rajagopalan S.Estimating risk of cancer associated with ra-diation exposure from 64-slice computed to-mography coronary angiography. JAMA2007;298(3):317–323.

41. Thompson RC, Cullom SJ. Issues regardingradiation dosage of cardiac nuclear and radi-ography procedures. J Nucl Cardiol 2006;13(1):19–23.

42. Halpern EJ, Levin DC, Zhang S, Takugawa.Comparison of image quality and arterial en-hancement with a dedicated coronary CTAprotocol versus a triple rule-out coronaryCTA protocol. Acad Radiol (in press).

43. Ferencik M, Nomura CH, Maurovich-HorvatP, et al. Quantitative parameters of imagequality in 64-slice computed tomography an-giography of the coronary arteries. Eur JRadiol 2006;57(3):373–379.

44. Dewey M, Hoffmann H, Hamm B. MultisliceCT coronary angiography: effect of sublin-gual nitroglycerine on the diameter of coro-nary arteries. Rofo 2006;178(6):600–604.

45. Chun EJ, Lee W, Choi YH, et al. Effects ofnitroglycerin on the diagnostic accuracy ofelectrocardiogram-gated coronary computedtomography angiography. J Comput Assist To-mogr 2008;32(1):86–92.

46. Wittram C, Yoo AJ. Transient interruptionof contrast on CT pulmonary angiography:proof of mechanism. J Thorac Imaging 2007;22(2):125–129.

47. Oser RF, Zuckerman DA, Gutierrez FR,Brink JA. Anatomic distribution of pulmo-nary emboli at pulmonary angiography: im-plications for cross-sectional imaging. Radi-ology 1996;199(1):31–35.

48. Remy-Jardin M, Remy J. Spiral CT angiogra-phy of the pulmonary circulation. Radiology1999;212(3):615–636.

49. Washington L, Goodman LR, Gonyo MB. CTfor thromboembolic disease. Radiol ClinNorth Am 2002;40(4):751–771.

50. Wittram C. How I do it: CT pulmonary an-giography. AJR Am J Roentgenol 2007;188(5):1255–1261.



51. Hargaden GC, Kavanagh EC, Fitzpatrick P,Murray JG. Diagnosis of pulmonary emboliand image quality at CT pulmonary angiog-raphy: influence of imaging direction withmultidetector CT. Clin Radiol 2006;61(7):600–603.

52. Cademartiri F, Mollet NR, Lemos PA, et al.Higher intracoronary attenuation improvesdiagnostic accuracy in MDCT coronary an-giography. AJR Am J Roentgenol 2006;187(4):W430–W433.

53. Rist C, Nikolaou K, Kirchin MA, et al. Con-trast bolus optimization for cardiac 16-slicecomputed tomography: comparison of con-trast medium formulations containing 300and 400 milligrams of iodine per milliliter.Invest Radiol 2006;41(5):460–467.

54. Rist C, Becker CR, Kirchin MA, et al. Opti-mization of cardiac MSCT contrast injectionprotocols: dependency of the main boluscontrast density on test bolus parametersand patients’ body weight. Acad Radiol 2008;15(1):49–57.

55. Dodd JD, Kalva S, Pena A, et al. Emergency

cardiac CT for suspected acute coronarysyndrome: qualitative and quantitative as-sessment of coronary, pulmonary, and aorticimage quality. AJR Am J Roentgenol 2008;191(3):870–877.

56. Fleischmann D, Rubin GD, Bankier AA,Hittmair K. Improved uniformity of aortic en-hancement with customized contrast mediuminjection protocols at CT angiography. Radiol-ogy 2000;214(2):363–371.

57. Cademartiri F, Mollet NR, van der Lugt A,et al. Intravenous contrast material admin-istration at helical 16-detector row CT cor-onary angiography: effect of iodine concen-tration on vascular attenuation. Radiology2005;236(2):661–665.

58. Vrachliotis TG, Bis KG, Haidary A, et al.Atypical chest pain: coronary, aortic, andpulmonary vasculature enhancement at bi-phasic single-injection 64-section CT angiog-raphy. Radiology 2007;243(2):368–376.

59. Haidary A, Bis K, Vrachliotis T, Kosuri R,Balasubramaniam M. Enhancement perfor-mance of a 64-slice triple rule-out protocol vs

16-slice and 10-slice multidetector CT-an-giography protocols for evaluation of aorticand pulmonary vasculature. J Comput AssistTomogr 2007;31(6):917–923.

60. Litmanovich D, Zamboni GA, Hauser TH,Lin PJ, Clouse ME, Raptopoulos V. ECG-gated chest CT angiography with 64-MDCTand tri-phasic IV contrast administrationregimen in patients with acute non-specificchest pain. Eur Radiol 2008;18(2):308–317.

61. Achenbach S. Computed tomography coro-nary angiography. J Am Coll Cardiol 2006;48(10):1919–1928.

62. Feuchtner G, Postel T, Weidinger F, et al.Is there a relation between non-calcifyingcoronary plaques and acute coronary syn-dromes? a retrospective study using multi-slice computed tomography. Cardiology2008;110(4):241–248.

63. Motoyama S, Kondo T, Sarai M, et al. Mul-tislice computed tomographic characteristicsof coronary lesions in acute coronary syn-dromes. J Am Coll Cardiol 2007;50(4):319–326.



triple-rule-out ct angiography for evaluation of acute...

Documents