Advertisement
Advertisement
September 2007
CTA Case Studies
Clinical utility of cardiac CTA in patients with previous CABG.
By Jason T. Bradley, MD; Mitchell N. Rashid, MD; and Matthew W. Watkins, MD, FACC
Multidetector computerized tomography (MDCT) has been studied most in patients at low to moderate risk.1,2 CTA has been best validated for the evaluation of coronary arteries in patients without known coronary disease. In this patient group, MDCT has a high sensitivity (>95% in recent reports) and negative predictive value.2 However, there are limited data on the accuracy and use of cardiac CTA for the evaluation of patients with known CAD and in particular, patients with previous coronary artery bypass graft (CABG).
For patients with previous CABG, there is no "normal" result, with most coronary segments and all patients having varying degrees of CAD. The coronary anatomy is more complex, and MDCT results must address bypass graft patency, bypass graft stenosis, and the severity of native coronary disease in nongrafted segments both proximal and distal to the anastomosis site. The case studies we present demonstrate the utility of MDCT in CABG patients, focusing on the information that MDCT adds to diagnosis and management. The subsequent literature review and discussion highlights recent results comparing accuracy of CTA to invasive angiography in bypass patients and our current approach to MDCT in the management of post-CABG patients.
EXAMPLES OF MDCT IN PATIENTS WITH PREVIOUS CABG
Case 1
A 74-year-old man presented with known CAD and a recent history of progressively worsening symptoms of chest pain and shortness of breath with exertion. These episodes are similar to previous anginal episodes. There were no symptoms at rest. The patient had CABG in 1991 with three distal anastomoses. A previous coronary angiogram in 2005 revealed no significant disease in the bypass grafts. The patient was seen in the clinic, and a 64-slice gated cardiac CTA was performed to evaluate for interim changes in bypass graft disease in the setting of increasing stable angina.
Retrospectively gated CTA images were acquired at .625-mm thickness, reconstructed at .8 mm, and reviewed at 10% increments of the cardiac cycle. The CTA showed a patent left internal mammary artery (LIMA) to the left anterior descending artery (LAD) with no significant disease below the LIMA touchdown. The saphenous vein graft (SVG) to an obtuse marginal branch had a focal plaque with 50% to 75% luminal narrowing in the midportion of the graft (Figure 1A, B). The SVG to the distal RCA contained a 50% to 75% stenosis in the proximal portion (Figure 1A, C). The native coronaries had proximal diffuse calcification and >50% to 75% stenosis in the LAD, circumflex, and RCA, but native runoff below the patent bypass grafts was patent with <50% stenosis.
A few days after the cardiac CTA, the patient was admitted to an outside hospital due to chest pain and shortness of breath. The patient was ruled out for a myocardial infarction and was transferred for cardiac catheterization based on the abnormal findings of his cardiac CTA. Invasive coronary angiography was performed to evaluate the SVGs. The SVG to the obtuse marginal artery had a new 80% stenosis in the midportion that corresponds to the lesion on CTA (Figure 2A). The SVG to the distal RCA contained a new 70% proximal stenosis corresponding to the CTA (Figure 2B). Corresponding to CTA results, there was also <50% stenosis in the native vessel runoff below these grafts. The patient underwent percutaneous coronary intervention of the proximal portion of the SVG to the distal RCA with a single 4-mm bare-metal stent with good result. The lesions in the proximal and midportions of the SVG to the obtuse marginal were successfully treated with a total of three 4-mm bare-metal stents.
Case 2
A 62-year-old man presented with known CAD and remote CABG more than 15 years ago at an outside hospital. He had placement of SVGs to the RCA and diagonal arteries. Invasive angiography in 2000 demonstrated an occlusion of the SVG to the diagonal and a focal, complex aneurysm of the proximal SVG to the RCA (Figure 3A). Subsequent invasive angiography in 2003 revealed enlargement of the aneurysm in the proximal SVG to the RCA graft with a continued patent lumen (Figure 3B). The patient has had a recent increase in complaints of atypical chest pain with a pleuritic component leading to an outside chest CT that showed a mass abutting the right atrium. Given the history of SVG aneurysm, a 64-slice gated cardiac CTA was performed for further evaluation of SVG disease and characterization of the pericardiac mass.
Retrospectively gated CTA images were acquired at .625-mm thickness, reconstructed at .8 mm, and reviewed at 10% increments of the cardiac cycle. The SVG to the diagonal artery is clearly occluded proximally, and a cast of calcification is seen in the midsection of the remnant graft (Figure 4A). Analogous to previous invasive angiography, the lumen of the SVG to the RCA shows a large, irregular aneurysm (Figure 4A). However, CTA imaging also reveals a mass surrounding the lumen of the SVG, up to 6 cm X 4.5 cm in diameter, with a density of 35 to 40 Houndsfield units; this is consistent with chronic mural thrombus inside the SVG (Figure 4B, C). The complex relationship between the patent SVG lumen and the massive surrounding mural thrombus, as well as its location filling a space between the right atrium and the sternum, is shown clearly by CTA imaging and not by previous invasive angiography. CTA also demonstrated >75% proximal stenosis in the nongrafted LAD and circumflex system. Based on the cardiac CTA findings, the patient is currently being evaluated for possible surgical ligation of the massive SVG aneurysm and repeat CABG.
DIAGNOSTIC CHALLENGE OF POST-CABG PATIENTS
The evaluation of patients with recurrent chest pain after bypass surgery remains a challenge for the cardiologist. Patients develop recurrent angina in a time-dependent manner following CABG, with at least 30% of patients developing anginal symptoms by 10 years after surgery.3 Patients are likely to acquire ischemic symptoms both due to progression of native coronary disease and secondary to progressive disease in bypass grafts. Multiple studies evaluating long-term patency of bypass grafts demonstrate vein graft patency at 10 years as low as 40% to 50%, and a more durable arterial graft patency of 51% to 98% at 15 years.4-6 Currently, the standard of care in managing coronary bypass patients includes a detailed history and physical, aggressive medical management, and continued cardiac risk factor reduction. For those individuals with new symptoms, either typical or atypical for angina, noninvasive testing is generally performed. Often, these patients have multiple comorbidities, and exercise treadmill testing is limited. Pharmacologic tests with SPECT imaging or stress echocardiography are among the most commonly used methods of evaluation. However, in this patient population, with underlying diffuse CAD, there is a high incidence of abnormal noninvasive studies, and many of these patients go on to invasive catheterization with coronary and graft angiography. For MDCT to meet this diagnostic challenge, it must be able to accurately assess graft patency, graft stenosis, and native vessel disease, while identifying severe stenosis in patients with diffuse background CAD.
MDCT APPLIED TO CABG PATIENTS
Cardiac CTA has progressed rapidly in the diagnosis and management of cardiac disease since the advent of multislice spiral or "helical" CT in 2000. Improved gantry rotation of <.5 seconds has led to a temporal resolution of <250 ms. This combined with the patientÕs ECG allows retrospective image reconstruction at quiescent cardiac phases and evaluation of the coronary arterial anatomy largely free of cardiac motion artifact. Spatial resolution has improved with current-generation 40- and 64-slice scan resolution of <.8 mm.7,8 Patient selection and preparation remain critical to obtaining high-quality cardiac CTA results. These factors include a stable heart rhythm and rate <65 bpm (in our experience this requires prescan beta blockers in almost all cases) and patient coaching to improve breath holding and minimize motion. Several factors related to MDCT image quality are of particular importance in CABG patients. A larger region of interest is examined in graft cases, including the aortic arch, which translates to longer scan times and increased contrast administration. In our experience, this requires a patient breath hold of 20 seconds and a contrast volume of 120 mL administered at 5 mL/s with a saline bolus chase. Bypass graft motion is less than native vessels, but rigorous attention to heart rate control remains important to improve image quality in the native vessels, which typically have extensive CAD. Finally, CTA image quality in CABG patients may be significantly limited by the presence of sternal wires, surgical clips along grafts, and diffuse native vessel calcium (Figure 5). The beam hardening and blooming artifacts associated with these factors remains a significant limitation in some cases.9-14
ACCURACY OF MDCT COMPARED TO ANGIOGRAPHY
In general, studies evaluating the accuracy of MDCT in CABG patients have compared CTA results in patients subsequently undergoing elective cardiac catheterization. Of note, bypass grafts and coronary segments that are deemed noninterpretable, often >5% of the total segments/grafts, are typically omitted from the calculation of MDCT accuracy compared to invasive angiography.
Table 1 compares the sensitivity, specificity, negative predictive value, and positive predictive value for the evaluation of graft occlusion, graft stenosis, and native vessels in five recent studies utilizing either a 16- or 64-slice MDCT. In contrast to CTA in nonbypass patients in whom >95% sensitivity and >90% specificity has routinely been reported,2,7,8 CTA accuracy in bypass patients is both lower and, in our view, more dependent on the improved temporal and spatial resolution of 64-slice and greater scanners. As an example, the distal graft anastomosis site could be visualized in only 74% of cases in a representative 16-slice study.10 Overall, 16-slice MDCT has been reported to accurately detect bypass occlusion and significant graft stenosis, but is limited downstream in the evaluation of distal graft anastomotic sites and native disease in calcified, diseased runoff vessels and in minimizing artifacts due to motion or metal clips.9-12
Recent reports with 64-slice scanners have been more encouraging. Ropers et al, using 64-slice MDCT, were able to evaluate and correctly classify 138 SVGs as patent or occluded with 100% sensitivity and 94% specificity.13 Importantly, downstream imaging accuracy was improved, with 40% of all graft stenoses seen being detected at the distal anastomosis site and 91% of distal runoff vessels able to be evaluated. In the distal native vessels, sensitivity was 86% and specificity was 76% for detection of stenosis >50%. Meyer et al, using 64-slice MDCT, studied 138 consecutive patients (418 bypass grafts) with recurrent chest pain for the detection of >50% stenosis in grafts only.14 In contrast to Roper et al, native vessels were not evaluated. However, in this study, 30% of patients had arrhythmias during CTA, a common finding in this patient group, and were not excluded from the analysis. Bypass grafts were interpretable in 98% of cases (motion artifacts and metallic clips led to no interpretation in nine of 418 grafts). In all 84 cases, MDCT correctly identified total graft occlusions, and 29 of 32 graft stenoses were correctly identified by MDCT. Despite the imaging improvements reported previously, modern MDCT technology still faces important hurdles in postbypass patients, which include metal artifacts and accurate evaluation of calcified and diffusely diseased native vessels.
An additional recent application of MDCT, which is pertinent to patients after CABG, has been to facilitate planning for PCI for chronic total coronary occlusion. These procedures have previously been guided by angiography alone. The course of the totally occluded segment cannot be ascertained by angiography. Pre-PCI MDCT has proven useful to define the course of the occluded segment.
SUMMARY
MDCT imaging in patients with previous CABG is inherently more challenging compared to patients without revascularization. These patients have more diffuse native CAD, frequent postsurgical metal artifacts, and fewer normal vascular segments. High-quality CTA scans are both more difficult to perform in this group and more challenging to interpret. However, recent improvements in MDCT technology have led to an impressively low incidence of noninterpretable bypass grafts at 0% to 2%. This is a lower rate of noninterpretable segments compared to the 5% to 9% rate reported for native vessels in patients without CABG.15 The 64-slice MDCT improves on previous-generation scanners to allow visualization of coronary bypass grafts including the anastomotic site and native vessels. MDCT is highly accurate for the diagnosis of overall bypass graft patency with a sensitivity and specificity approaching 98% and 94% in recent reports. MDCT diagnosis of significant graft stenosis is less accurate than graft occlusion. In contrast, MDCT is significantly less accurate in detecting native vessel lesions in CABG patients due to a high incidence of calcium and diffuse CAD. Definitive recommendations regarding appropriateness and indications for MDCT in CABG patients are evolving and await prospective trials. We agree with recent consensus statements that MDCT is not generally indicated either as a screening tool in asymptomatic CABG patients or in patients presenting with unstable coronary syndromes.16,17 An exception to the former is a clinical scenario in which documentation of chronic bypass graft patency will change patient care, such as establishing LIMA status in a CABG patient with an unexpected anterior perfusion defect. In conclusion, cardiac CTA is most likely to alter clinical management in selected symptomatic patients, in whom the presence of severe graft stenosis or native disease in nongrafted vessels would lead to invasive testing and possible interventional therapy.
The authors would like to acknowledge Heidi Streeter, BS, RT(R)(M)(CT), and Frank Flynn, AS, RT(R)(CT), 3D/CT imaging technologists, for their assistance in the preparation of the images for this manuscript.
Jason T. Bradley, MD, is from the Department of Cardiology at Fletcher Allen Health Care/University of Vermont, Burlington, Vermont. He has disclosed that he holds no financial interest in any product or manufacturer mentioned herein. Dr. Bradley may be reached at (802) 847-3734; jason.bradley111@gmail.com.
Mitchell N. Rashid, MD, is from the Department of Cardiology at Fletcher Allen Health Care/University of Vermont, Burlington, Vermont. He has disclosed that he holds no financial interest in any product or manufacturer mentioned herein. Dr. Rashid may be reached at (802) 847-3734; rashid2@earthlink.net.
Matthew W. Watkins, MD, FACC, is from the Department of Cardiology at Fletcher Allen Health Care/University of Vermont, Burlington, Vermont. He has disclosed that he holds no financial interest in any product or manufacturer mentioned herein. Dr. Watkins may be reached at (802) 847-2700; matthew.watkins@vtmednet.org.
Advertisement
Advertisement
