Since its original description in 1992, computed tomography angiography (CTA) has advanced from the method of using a simple single-detector to a powerful imaging multidetector system capable of acquiring 16 channels of data with rapid imaging acquisition and higher spatial resolution, while simultaneously allowing patient coverage of more than 120cm with a single scan. Computed tomography (CT) has undergone a remarkable evolution over the last two decades and now allows subsecond scan times, submillimeter scan thickness and the acquisition of more than 1,000 slices per examination with a single venous injection. Optimized contrast enhancement, improved three-dimensional (3-D) volumetric data analysis, and sophisticated computer software and workstations are but several advantages of the current technology and is why several sources are predicting that multidetector CTA (MDCTA) will ÔÇ£replace 80% to 90% of all conventional diagnostic angiographyÔÇØ.1 MDCTA has already demonstrated comparability and even diagnostic superiority over conventional diagnostic angiography (CDA) in several vascular applications including the aorta, carotid, renal, iliac, and pulmonary vessels.
CDA with digital subtraction angiography (DSA) remains the clinical 'gold standardÔÇÖ for vascular imaging but has multiple limitations. Magnetic resonance angiography (MRA) has been advocated to address the limitations of DSA but MRA also possesses significant limitations. In comparing MDCTA with DSA and MRA, noteworthy MDCTA advantages include lower costs, less invasiveness (venous stick),much faster operating times, potentially less contrast use and radiation exposure, allowance for 3-D reconstruction, it does not require an angiosuite or cathode laboratory team, and has potentially fewer complications (see Box 1).
MDCTA can also be utilized in evaluating coronary artery disease (CAD), chronic vascular thrombus or calcification, and non-vascular soft tissues and osseous structures. Since MDCTA does not require an arterial access, patients on anticoagulation or with hypercoagulable states need no pre- or peri-procedural preparations, and MDCTA can be utilized in patients with limited vascular access (grafts, severe peripheral vascular disease (PVD), or absent pulses). This article describes the authorsÔÇÖ early 16-channel and current 64-channel out-patient office experience with MDCTA, and how it has changed and influenced patient care.
The three major principles of MDCTA are:
- to achieve an adequate level of arterial contrast enhancement during acquisition;
- to provide cephalocaudad coverage of the targeted anatomy during an early sustainable breath-hold interval (less than 20 seconds); and
- to time the onset of CT acquisition after contrast injection accurately, so the first circulation enhancement is obtained from the start to the end of acquisition.
An important aspect of efficient clinical MDCTA is the accurate determination of the suspected vascular pathology of the patient, and therefore the clinical information requested. This requires developing close communication between the physician and MDCTA team, particularly during the early experience.
Typically, 80-100mL of an intravenous (IV) non-ionic contrast agent is administered through a 20-gauge plastic venous cannula in an antecubital vein at an injection rate of 4-6mL per second. In clinical MDCTA, the contrast enhancement, acquisition parameters, coverage speed, and circulating time are accurately determined by protocols based on the clinical information requested, imaging modality used, and patient-specific variables (age, body size, cardiac output, and renal function, etc.). With current technology, it is reasonable to obtain images from the clavicles or abdomen down to the knees or the abdomen and pelvis to the patientÔÇÖs feet, all within 30 seconds. After the primary author observed his first MDCTA, he commented, ÔÇ£this is like shooting 16 angiograms all from different angles, simultaneously, in color, and in just a few secondsÔÇØ.
Abdominal Aortic Aneurysm
Conventional CT scanning and CDA have been considered the 'gold standardÔÇÖ for abdominal aortic aneurysm (AAA) diagnosis, treatment planning recommendations, and post-operative surveillance, particularly since the introduction of aortic stent grafts and endovascular aneurysm reconstruction (EVAR). Besides being non-invasive, MDCTA offers several advantages over traditional imaging including:
- superiority in identifying mural thrombus and evaluating peri-aortic tissues for rupture, endoleak, and inflammatory AAA identification;
- more precise determinations of size, length, angulation, and transverse dimensions of the AAA superior neck;
- more accurate characterization of post-procedural juxtarenal AAA endograft deformation, kinking, or migration with identification of branch vessels;
- 3-D reconstruction allows improved assessment of iliac tortuosity and detection of endoleaks; and
- post-EVAR AAA volume determinations may allow earlier detection of device failure, rupture or endoleak.2
Traditional imaging can overestimate the true AAA neck diameter and length because of obliquity, therefore influencing the treatment recommendations and outcomes. MDCTA has replaced traditional CT scanning and DSA in almost all aspects of the authorsÔÇÖ AAA pre-therapy recommendations and post-treatment follow-up (see Figure 1a-d).
Carotid Artery Disease
The degree of stenosis of the internal carotid artery (ICA) has been scrutinized more than any other peripheral vessel because the degree of stenosis is associated with the risk of stroke. Landmark trials over the last two decades have demonstrated reduced stroke rates in both symptomatic and asymptomatic patients with ICA stenosis of ÔëÑ50% to 60% treated by carotid endarterectomy (CEA).3,4 Several equally land-marked interventional carotid artery stent (CAS) trials have shown at least equivalency and even improved early outcomes with ICA stenting compared with surgery challenging the 'gold standardÔÇÖ for ICA disease treatment.5,6 This will also challenge the authorsÔÇÖ 'gold standardÔÇÖ for ICA imaging since optimal CAS will require more detailed and comprehensive information on the aortic arch and entire extra- and intracranial carotid artery system, which is currently required for surgery.
CDA with DSA remains the 'gold standardÔÇÖ for ICA imaging but there is a definite risk of periprocedural stroke. DSA is performed in limited projections while MDCTA provides multiple views. DSA has recently been shown to underestimate the degree of ICA stenosis when surgical specimens of cross-sectional lumens were compared with DSA.7 Using helical MDCTA, Elgersma et al.8 identified additional (16%) ICA suitable for CEA compared with DSA, therefore further underscoring the complexity of the ICA bifurcation and the need for multiple views to accurately determine the degree of stenosis.
Carotid duplex ultrasound (DU) has been known to have a sensitivity and specificity of more than 90% in diagnosing ICA disease but is still operator-dependent, moderately time-consuming, and gives limited information on the distal extracranial ICA and intrathoracic common carotid artery (CCA), and no information on the intracranial or arch vessels. Several reports have shown the diagnostic accuracy of carotid MDCTA for 70% to 99% ICA stenosis to have a sensitivity of 100% and specificity of 94% to 100%.9,10 Multiplanar reconstruction (MPR) methods will allow cross-sectional luminal and plaque morphology evaluation, and potentially provide additional pertinent clinical information.11
A routine carotid MDCTA examination includes 30cms of cephalocaudad coverage from the ostium of the aortic arch vessels to the circle of Willis at the base of the skull. This allows an accurate decision-making recommendation regarding a patientÔÇÖs CAS candidacy with a single non-invasive examination to be given. Currently, intracranial vessels are not being routinely imaged, although this capability is possible. MDCTA neuroimaging has been found to compare favorably with DSA in evaluating acute stroke, arterial dissections, and intracranial aneurysms with reduced patient risks.12 Excellent proximal vertebral artery imaging is available on a routine carotid MDCTA allowing identification of additional symptomatic patients who could benefit from endovascular intervention.
Aortic arch vessel anatomy, tortuosity, and ostial arch vessel disease (AVD) along with distal ICA tortuosity are often significant limitations to carotid stenting, since placement of a distal protection device has been required in most CAS trials. Isolated ostial AVD is thought to be rare with a reported incidence of 1.8%, and is reported to be even less (0.6%) in patients undergoing CEA.13,14 This most likely underestimates the true incidence of AVD as many selective carotid angiograms are performed without arch angiography, and CEA is often recommended on DU alone. The authorsÔÇÖ group reported a 4.3% (34/784) incidence of simultaneous significant ostial AVD in patients requiring CEA, further raising questions regarding the underestimation of AVD.15 MDCTA now allows accurate assessment of ICA stenosis, access vessel and distal ICA tortuosity, and the detection of ostial AVD, therefore potentially allowing a determination of CAS versus CEA candidacy non-invasively without the inherent risks of DSA. In the authorsÔÇÖ experience, carotid MDCTA has almost replaced CDA and DSA as a diagnostic tool, and increasingly important carotid treatment and surveillance decisions are made relying on MDCTA information alone (see Figure 2a-d).
Renal and Mesenteric Arteries
The benefits of renal artery stenosis (RAS) diagnosis and revascularization have been proven in patients with severe hypertension, refractory CHF, angina and patients with declining renal functionÔÇötherefore underscoring the importance of a safe, accurate, simple non-invasive diagnostic tool. The limitations of abdominal vessel DU are well known. Three-dimensional abdominal MDCTA has addressed many of these limitations, and Berengi et al.16 recently reported 100% accuracy and sensitivity comparing MDCTA with DSA in RAS. Galanski et al.17 reported the sensitivity and specificity of MDCTA compared with DSA in RAS as 92% and 95%, respectively.
Recent reports by AbuRahma et al.18 and the authorsÔÇÖ group19 have identified benefits with PTA/stenting in appropriately diagnosed patients with mesenteric artery disease (MAD). There are no such reports of MDCTA evaluating MAD although the authorsÔÇÖ initial experience indicates that MDCTA will be as accurate in diagnosing MAD as in RAS. Stents are very well imaged, allowing the potential for improved accuracy in detecting restenosis, which has a 10% to 20% incidence post-renal and mesenteric PTA/stenting.
Abdominal MDCTA is rapidly replacing DU as a primary diagnostic and surveillance tool in evaluating RAS and MAD. In patients with RAS, the MDCTA scanner can 'pinpointÔÇÖ the juxtarenal aorta and shorten acquisition times to less than 10 seconds and reduce contrast load to less than 40mL.
Asymptomatic and symptomatic critical MAD disease (celiac and superior-inferior MAD) has been difficult to diagnose both clinically and by traditional imaging techniquesÔÇöit is therefore underappreciated, underdiagnosed, and undertreated. This is likely to be very analogous to the understanding and treatment of RAS two decades ago. An abdominal aorta with distal run-off CTA will begin acquiring 3-D images at the level of the diaphragm, and will therefore start to identify a significant amount of MAD. It is likely that MAD is as common as RASÔÇöthis will therefore be a new vascular territory available for percutaneous revascularization, because the endovascular treatment techniques are very similar to treating RAS. Traditionally, the treatment of MAD has required difficult major vascular surgical reconstructions with high mortalities and morbidities. It is possible that with MDCTA and endovascular treatments, the incidence of treating MAD will greatly increase in the near future.
Aortoiliac Occlusive Disease
An 'endovascular firstÔÇÖ approach toward aortoiliac occlusive disease (AIOD) revascularization was adopted in the authorsÔÇÖ practice well over a decade ago, and acceptable long-term results in iliac PTA/stenting have substantiated this policy. DSA can fail to detect aneurysmal disease and the diagnostic accuracy is adversely affected by vascular calcification. Single planar DSA frequently 'missesÔÇÖ eccentric lesions, which are so posteriorly prevalent at the distal aortic bifurcation and entire iliac artery segment. DSA often poorly visualizes distal vessels to a more proximal stenosis or occlusion.
Native common femoral artery (CFA) disease is not uncommon and patients frequently present with 'stick site injuriesÔÇÖ from multiple prior procedures. Avoiding a 'diagnostic stickÔÇÖ with DSA has significant clinical implications. The authors have found MDCTA to be particularly useful in avoiding 'sticking into diseaseÔÇÖ and planning endovascular treatments.
Several recent studies have shown the sensitivity and specificity of MDCTA in detecting significant AIOD to be more than 96%20 and Rubin et al.21 recently reported 100% concordance between MDCTA and DSA in AIOD using 2.5mm slices covering 120mm in 60 seconds acquisition time. MDCTA has been shown to be superior to DSA in evaluating vascular trauma, dissections, and popliteal aneurysms. Several asymptomatic AAA and significant iliac and popliteal aneurysms bigger than 5cm are diagnosed each month on patients imaged for suspected AIOD. Recent reports have shown a reduction in contrast use and a four-fold reduction in radiation exposure comparing MDCTA and DSA in diagnosing AIOD.22
The authorsÔÇÖ routine MDCTA for AIOD includes cephalocaudad coverage from the supraceliac aorta to the proximal thigh (approximately 30-40cm). Scanning parameters can be adjusted to achieve a reduced contrast load (less than 75cc) with scan times Ôëñ30 seconds. It is always recommended to scan both CFAs, since most endovascular procedures are performed via a CFA approach. MDCTA is now the authorsÔÇÖ diagnostic mode of choice for diagnosing and following AIOD, and more than 80% of their surgical and endovascular treatment recommendations are influenced by MDCTA.
The SFA and crural vessels are among the most calcified vessels in the body, therefore are a challenge to accurate assessment of the degree of stenosis. The high incidence of vessel occlusions, asymmetric disease, and vascular calcifications are highly characteristic of infrainguinal disease and known limitations of DSA and DU. Known advantages of MDCTA include improved accuracy in vascular occlusions, calcifications, and patients with asymmetrical disease.
There is no published data comparing DSA or DU with MDCTA in infrainguinal vessel, but several promising image-enhancement processing techniques are available, especially in infrapopliteal vessels. Curved planar reformation (CPR) and semi-transparent volume rendering (STVR) with automated measurements are new 3-D imaging modalities that improve the accuracy of MDCTA in highly calcified vessels.23
The editing of bony structures (osseous segmentation) is now available at the workstation using automatic region growing imaging techniques. Segmentation of the tibia, fibula, and tarsal osseous structures can have significant clinical implications in achieving limb salvage. Maximum intensity projection (MIP) is a 3-D workstation that allows maximal contrast opacification and vessel interrogation.
Rubin et al.21 reported the identification of 26 additional infrainguinal arterial segments that were not identified with DSA utilizing MDCTA due to the improved arterial opacification distal to an occluded segment. Identification of 'distal targetsÔÇÖ could identify patients for tibial bypass or endovascular revascularization who otherwise may only be offered amputation. This becomes particularly important with the excellent six- to 12-month limb salvage rates reported by Laird et al.24 (more than 90%) utilizing the excimer laser (Spectranetics Corporation, Colorado Springs, CO) in the multicentered Laser Angioplasty for Critical Limb Ischemia (LACI) trial. Similar results were reproduced by the authorsÔÇÖ group25 in a 'LACI-equivalentÔÇÖ report. MRA is considered the 'gold standardÔÇÖ in imaging pedal vessels but MRA is not reliable in calcified vessels, is time-consuming, not widely available, and affected by retrograde flow artifacts, existing stents, and pacemakers. The authors have found infrainguinal MDCTA invaluable in the diagnosis, treatment, and planning and follow-up of their infrainguinal interventions, particularly in infrapopliteal vessels where DU is rarely helpful (see Figure 3a-f).
Currently, the authorsÔÇÖ most frequently utilized imaging is an abdominal MDCTA with bilateral run-off to the feet. This can usually be accomplished with 80-100mL of contrast injected over approximately 120cm with a scanning duration of less than 30 seconds.
Bypass Graft Surveillance
Both infrainguinal bypass grafts and endovascular treated vascular segments have been shown to require a 20% to 30% secondary intervention rate to achieve an acceptable one- to two-year patency, underscoring the need for non-invasive post-procedural surveillance.26 DU has been the 'gold standardÔÇÖ in the post-surgical patient, but DU limitations not found with MDCTA include difficulty in following non-anatomically placed grafts, inaccuracy in cases with inflow lesions, multiple graft lesions, A-V fistulas, inability to 'pinpointÔÇÖ stenosis, and failure to display tibial or pedal vessels distal to the anastomosis. Willmann et al.27 reported 98% sensitivity and specificity in comparing DSA with four-channel MDCTA in 85 bypass grafts. In the authorsÔÇÖ experience, MDCTA has allowed accurate diagnosis and planning for graft reintervention, while DU has not.
Cardiac MDCTA differs from routine MDCTA for PVD primarily due to the dynamic motion of the heart and static nature of PVD. Detector, table, and imaging parameters vary greatly from these used in PVD.The scan times are much longer and oral and IV ╬▓-blockers are required to keep the heart rate at less than 60bpm. Current MDCTA is unable to accurately grade coronary artery stenosis, particularly in small vessels. Cardiac MDCTA has shown a high negative-predictive value for significant disease and the sensitivity, specificity, and reproducibility of MDCTA for detection of coronary calcification has been found to be higher than electron beam CT (EBCT).28
MDCTA has been found to identify intracoronary thrombus much like in PE, and may have a promising role in diagnosing acute coronary syndromes. Other potential future applications include post-PCI stent evaluation and vulnerable plaque identification. The authors are not currently performing cardiac MDCTA for CAD screening or patients with stable angina but they are increasingly using MDCTA to evaluate coronary artery bypass grafts (CABGs) and pulmonary venous anatomy in planning electrophysiological procedures.
Additional Vascular Applications
MDCTA is rapidly replacing traditional imaging modalities in patients with acute and chronic aortic syndromes (dissections and aneurysms), acute and chronic pulmonary embolus (PE), congenital cardiovascular anomalies, cardiothoracic and vascular neoplasms and trauma, acute stroke, hepatic vascular evaluations, organ donor transplantations, complex vascular anomalies and malformations, and superior vena cava syndrome.11,19,20,28 MDCTA has shown a sensitivity and specificity of 78% to 100% in diagnosing PE, as contrast enhancement techniques using sub-1.25mm slice thickness allow the visualization of small PE, even in fifth- and sixth-generation pulmonary vessels.29 With more than 650,000 people newly diagnosed with PE each year in the US, MDCTA has provided a quick, safe, and accurate non-invasive tool to diagnose PE where early treatment can reduce mortality from 30% to approximately 2% to 10%.29
Additional Non-vascular Applications
With eight cardiothoracic surgeons on staff, contrast enhanced thoracic CT is becoming increasingly important in evaluating pulmonary nodules, differentiating between benign and malignant disease, staging, treatment, and in thoracic surgical follow-up. Swensen et al.30 reported a 98% sensitivity and 58% specificity for malignancy using contrast enhanced CT in evaluating 356 lung nodules from 5-40mm in diameter. Widespread single and dual detector CT screening for lung cancer has not been thought to be cost-effective and has a high false-positive rate. It is very likely that in the future, MDCTA screening in high-risk patient populations will be clinically and cost-effective.
The lifetime risk of colorectal cancer in the US is 5.7%, and screening for fecal occult blood has shown a 33% reduction in mortality.31,32 CT colonography is gaining popularity as a non-invasive screen for cancer and has a sensitivity of 90% to 94% and specificity of 72%. Currently, this requires special patient preparation, colon inflation, and is time-consuming. Total body CT scanning for 'occultÔÇÖ neoplasms has been promoted, but currently lacks randomized cost and clinical analysis. We have not adopted 'cancer screeningÔÇÖ but do realize the potential clinical benefits to the large elderly populations at risk and the potential to build referral relationships with primary care physicians (PCPs), oncologists, and radiologists.
Currently, the authors have a radiologist interpret all non-vascular CT imaging and 'over readÔÇÖ all vascular MDCTA studies. Their cardiologist and surgeons also currently interpret each MDCTA according to their CDA and DSA policy. They are increasingly utilizing traditional CT imaging for more non-cardiovascular applications (i.e. CT of the head, chest, and lumber spine, etc.). Having this imaging modality rapidly available in their out-patient office setting has greatly simplified and facilitated the overall clinical work-up and follow-up of the patient population who often have multiple medical problems.
Contrast induced nephropathy (CIN) has been the single biggest concern and adjustment to the authorsÔÇÖ out-patient practice. Most patients with a creatinine (CR) Ôëñ 1.8mg/dl can tolerate an IV dose of 100-120cc of non-ionic low osmolar contrast without clinical sequelae. Multiple protocols are available for MDCTA in patients with chronic renal insufficiency. The physicians, nurses, and MDCTA team must evaluate the renal function of each patient considered for MDCTA. In low-risk CIN patients (CR <1.4mg/dl) no preparation is necessary.
In higher risk CIN patients (diabetic, CR 1.4-2mg/dl) a 'renal preparationÔÇÖ protocol is instituted that includes Mucomyst (Acetylcysteine) 600mg po bid for four doses beginning the day before MDCTA. An IV drip of D5 1/2 NS at 100cc/hr for six hours prior to MDCTA is administered in the authorsÔÇÖ holding area for patients with the greatest CIN risk (CR 1.7-2mg/dl). Contrast conserving measures are utilized. For patients with CR >2mg/dl, hospitalization for nephrology consultation and further renal preparation is recommended. A one-week follow up CR is obtained on all out-patient MDCTA patients.
The 16- to 64-channel multidetector CTA imaging system has revolutionized the authorsÔÇÖ PVD practice in only a few short months. MDCTA has become their non-invasive imaging modality of choice and is rapidly replacing CDA and DSA as their 'gold standardÔÇÖ for diagnostic PVD evaluation, and also often for post-procedural follow-up. MDCTA allows them to confidently make clinical planning and treatment decisions on their endovascular and surgical patients much more efficiently, accurately, and with less risk and costs to the patient than with conventional imaging techniques before. Each week they are discovering new clinical uses for this new technology, which will translate to improved overall outcomes for patients. Ôûá
The authors would like to thank Mrs Kelly Tilbe for her technical help with manuscript preparation.