Clinical impact of computerised tomographic angiography performed for preoperative evaluation before coronary artery bypass grafting^☆ Free

Abstract

Objective: In our institution, computerised tomographic (CT) angiography has been performed as a near-routine test before coronary artery bypass grafting (CABG) to evaluate the aorta and its branches. We aimed to determine its impact on operative procedure and perioperative management. Methods: From January 2006 through December 2008, neck-to-leg systemic arteries were evaluated by 64-slice multidetector CT scan in 284 patients before CABG. For them, medical records and cardiac surgery database were retrospectively reviewed to investigate the incidence of pathologic lesions of the aorta and its major branches, other incidental lesions and procedure- or management-related decision affected by CT angiography. Results: In 36 patients (12.7%), cannulation and clamping of the ascending aorta were thought too dangerous because of severe atherosclerotic change. The use of cardiopulmonary bypass was inevitable in 10 of them, and postoperative mortality rate was significantly higher (3/10, 30%) than in the remaining patients (2/274, 0.7%). Conduit selection or grafting strategy was changed due to severe left subclavian artery stenosis or leg vein varicosity in 18 patients (6.3%). For the following problems, surgical interventions were performed in 34 patients (12.0%) concomitantly or during the same admission with CABG; ≥90% stenosis in carotid (5.6%), iliofemoral (11.6%) or renal (4.6%) arteries, aortic aneurysm (14.4%, six of them met surgical indication) and incidental malignant neoplasm (2.8%). Counting all of the above, preoperative CT angiography affected risk assessment, CABG procedure itself, perioperative management or follow-up plan in 142 patients (50%). Conclusions: Because preoperative CT angiography considerably affected management in one half of patients, we recommend its use as a routine test before CABG, unless contraindicated.

1 Introduction

With the remarkable advance of technology, computed tomography (CT) scan has become the most useful tool for evaluation of vascular diseases. Diverse clinician-friendly high-resolution images provided by current equipment and software made fast and non-invasive diagnosis of arterial lesions possible with high sensitivity and accuracy. For patients with coronary artery disease, CT angiography is now used for several purposes such as screening for coronary artery disease in both outpatient and emergency settings [1–3], follow-up after percutaneous intervention [4], and evaluation of bypass graft after surgical revascularisation [5]. Besides the coronary artery, whole-body systemic arteries can be evaluated conveniently with up-to-date 64-slice CT scanner and several contrast- and radiation-saving protocols [6]. Many radiologists consider multi-slice whole-body CT the clinical standard for assessment of atherosclerosis burden in patients at high risk for cardiovascular events. However, despite of wide application of CT angiography for a variety of purposes, few studies have focussed on the role or potential of CT angiography as a preoperative evaluation tool in patients undergoing coronary artery bypass grafting (CABG).

It has been known that atherosclerotic diseases of the aorta and its major branches are prevalent in patients with coronary artery disease and affect short- and long-term prognosis considerably [7–11]. Many studies reported that multisystem arterial diseases are more frequently found in patients who have multivessel or left main coronary disease, the major indications of CABG [12,13]. In addition, atherosclerotic lesion of the thoracic aorta is relatively frequent in CABG patients and is one of the major risk factors of postoperative mortality and morbidity [14,15]. Consequently, preoperative identification of atherosclerotic disease of the aorta and non-coronary systemic arteries would affect perioperative management and outcome [16,17], long-term risk factor control and prediction of prognosis.

Unlike ultrasonography (US) and magnetic resonance (MR) imaging, which have been frequently used for preoperative examination of CABG patients with suspicious cerebrovascular or other systemic arterial diseases, CT angiography provides surgeon-friendly three-dimensional images of multiple systemic arteries and veins with fast one-stop scanning (Fig. 1 ). Since opening of our institution 6 years ago, our team made it a rule to perform neck-to-leg CT angiography in as many CABG patients as possible, 1 or 2 days prior to surgery. The aim of this study is to determine the clinical impact and usefulness of CT angiography in procedure-related decision making and perioperative management.

2 Materials and methods

2.1 Patients

From January 2006 through December 2008, a total of 418 patients underwent CABG in our institution. Among them, there were 25 patients who needed CABG in combination with valvular or aortic surgery and 33 patients who required emergency CABG for cardiogenic shock or complication of percutaneous intervention. Among the remaining 360 patients, who underwent elective or urgent CABG for coronary artery disease as the primary diagnosis, excluding 76 patients, who had renal dysfunction (serum creatinine ≥2.0 mg dl⁻¹), allergy to contrast dye, severe angina or congestive symptom or refused CT, neck-to-leg CT angiography was done before surgery in 284 patients. The male:female ratio was 200:84 and the mean age was 66.3 ± 8.7 years (range 33–89 years). Patients, who underwent CABG before 2006, were excluded because there was a change in the operating surgeon that resulted in change of strategy in operative procedure and perioperative management.

2.2 Equipment and protocol of CT scan

The CT scan was done with a 64-slice (Brilliance 64; Philips Medical Systems) multidetector row scanner. Images were acquired from the mandibular level to the feet at 5-mm intervals. Following a bolus intravenous injection of 100–120 ml non-ionic contrast material (Ultravist 370; Schering, Berlin, Germany) and flushing with 20 ml saline at a flow rate of 4 ml s⁻¹, the enhanced image was acquired at 5 s after the signal density level in the ascending aorta reached the predefined threshold of 150 Hounsfield Units. Additional delayed scanning was performed after 3 min.

2.3 Review of clinical data and images

By retrospectively reviewing the radiologists’ report, the incidence of steno-occlusive lesions of major branches of the aorta and incidental non-cardiovascular lesions was recorded. To determine the severity of atherosclerotic change of the aorta, one surgeon (the first author) reviewed the CT images in person and divided patients into five groups according to the location and morphology of atherosclerotic lesions in the thoracic and abdominal aorta (Fig. 2 ).

The in-department database and hospital records were retrospectively reviewed to investigate the incidence of perioperative management affected by CT angiographic findings including operative techniques, grafting strategy, out-of-department consultation and referral and follow-up plan at the time of discharge. Early postoperative (30-day and in-hospital) mortality and major morbidity were also investigated. Major postoperative morbidities counted were stroke and severe delirium, new haemodialysis for renal dysfunction, re-operation for bleeding, mediastinal or deep sternal infection, mechanical ventilation for longer than 48 h and gastrointestinal complications requiring surgical exploration. Mortality and morbidity were investigated also for the 76 patients, who did not undergo preoperative CT angiography. For statistical analysis, the chi-square (χ²) test or the Fisher’s exact test was used to compare categorical variables with SPSS 15.0 for Windows (SPSS Inc., Chicago, IL, USA). The Institutional Review Board approved this study and waived the individual patient’s consent.

3 Results

3.1 Aortic atherosclerosis

In 107 patients (37.7%), significant atherosclerotic change of the aorta (group III–V) was observed (Table 1 ). For the seven patients in group IV who needed cardiopulmonary bypass, cannulation was done at the more proximal part of the ascending aorta after inspection with epi-aortic US. Although cannulation or clamping of the ascending aorta was considered too dangerous in group V, use of cardiopulmonary bypass was deemed necessary during preoperative planning in six patients. For them, the right axillary artery was cannulated and CABG was done after replacement of the ascending aorta with a prosthetic graft during hypothermic circulatory arrest (Table 2 ). In another four patients, who needed intra-operative conversion from off-pump to on-pump surgery, graft anastomosis was done on beating heart or during hypothermic ventricular fibrillation to avoid aortic cross-clamping. The other 26 patients of group V underwent no-aorta-touch CABG, that is, off-pump revascularisation only with pedicle and composite arterial grafts.

As shown in Table 1 and Fig. 3 , postoperative outcome was different according to the location and severity of aortic atherosclerosis as assessed by CT angiography. While there was no mortality after off-pump surgery in any group, mortality rate after on-pump surgery was higher in the patients who had more extensive, more proximal and more severe lesions in the aorta. The causes of death were low cardiac output (two), severe bowel ischaemia (two) and respiratory distress syndrome (one). Death caused by low cardiac output occurred in the patients of group V who required conversion to on-pump surgery during off-pump CABG. The patients who died of bowel ischaemia had severe atherosclerotic change in the entire descending aorta (groups III and IV). Incidence of major morbidity showed the same trend with the mortality rate in both on-pump and off-pump groups. Stroke occurred in one patient who underwent on-pump CABG. She belonged to group II, but her left internal carotid artery had >90% stenosis and brain infarction developed in the corresponding region.

In the patients who did not undergo preoperative CT angiography, the mortality rate was higher than in those who underwent CT (5.3% vs 1.8%) and morbidity rate was not different (11.8% vs 16.8%). Further, in these patients, mortality occurred only after on-pump CABG and the causes were combined stroke and acute renal failure leading to multi-organ failure (two), bowel ischaemia (one) and low cardiac output (one). Stroke occurred in three patients; two after off-pump and one after on-pump CABG. All of them had previous history of stroke and significant vertebral or intracranial artery lesions that were documented by magnetic resonance (MR) angiography performed after stroke.

3.2 Findings related with bypass conduit

Asymptomatic varicosity of the greater saphenous vein was found in eight patients (2.8%). Vein graft was used in four of them, and decision of the harvesting site was guided by the CT findings. In 10 patients (3.5%), the left internal thoracic artery (LITA) was found to be inadequate for use as a pedicle graft because of severely stenotic or occlusive lesions of proximal left subclavian artery. In four of five patients in whom total occlusion of the left subclavian artery had already been diagnosed during coronary angiography, LITA was harvested because it was patent in CT. Including them, LITA was used as a free graft in eight patients (Fig. 4 ). In one patient, LITA was used as a pedicle graft after percutaneous angioplasty of the left subclavian artery. The other patient was revascularised with only saphenous vein grafts because both subclavian arteries had severe disease and internal thoracic arteries (ITAs) were very thin. In this patient, the radial and right gastroepiploic arteries were not used because he was on long-term haemodialysis via forearm arteriovenous fistula and had severe atherosclerotic change in the entire descending aorta.

3.3 Vascular diseases and aortic aneurysm

Severe stenosis of at least one major artery was found in 17.6% of the patients. Among the 16 patients who had severe carotid stenosis, only eight had previous history of stroke or transient ischaemic attack. Leg claudication was present in 13 of 33 patients, who had severe iliofemoropopliteal disease.

Insertion of an IABP catheter, in case it was needed, was considered to need caution or carry the risk of leg ischaemia or embolic complications because of diffuse or severe atherosclerotic lesions of the iliofemoral arteries, severe tortuosity of such vessels or vulnerable atheroma of the descending aorta in 54 patients (19.0%). Among the five patients who actually needed postoperative IABP support, three patients belonged to this high-risk group. Decision with regard to the side of IABP catheter insertion and duration of support was aided by the CT findings in these patients. There was no complication related with IABP.

Aortic aneurysm was found in 41 patients (14.4%). The size criteria defining an aneurysm were 4.5 cm for the ascending aorta, 4.0 cm for the arch, 4.5 cm for the thoraco-abdominal aorta and 3.5 cm for the infrarenal aorta. Surgical intervention was indicated in six of them because of large size, that is, maximal diameter ≥6.0 cm.

3.4 Non-cardiovascular lesions

As the result of CT findings, surgical resection of tumour was done concomitantly with CABG in six patients; lung lobectomy in two, nephrectomy in one, laparoscopic gastrectomy in one, extended thymectomy in one and resection of intrathoracic goitre in one patient. There were 59 patients, who required preoperative consultation or postoperative transfer to other departments after CABG, for further management or work-up of incidentally found non-cardiovascular lesions. Benign lesions such as hepatic or renal cysts, small thyroid nodule and gallbladder stone were not included in the count.

3.5 Overall incidence of important findings

Considering all of the above, CT angiography affected assessment of surgical risk, perioperative consultation, CABG procedure itself, perioperative management or follow-up plan in 142 patients (50%), including 34 patients (12.0%), who needed additional surgical or interventional procedures concomitantly or during the same admission with CABG (Table 2). The incidence of such important findings according to the patients’ age group was as follows; 14.3% in the 40s or younger, 34.2% in the 50s, 50.4% in the 60s, and 59.6% in the 70s or older.

4 Discussion

Current 64-slice CT can provide high-resolution images of coronary and systemic arteries with one-stop examination that takes less than 30 min. For that reason, some radiologists recommend whole-body CT angiography as the standard tool for assessment of the atherosclerosis burden in patients at high risk for cardiovascular diseases [6]. Unlike ultrasonography (US) that has been widely used for screening of the carotid and peripheral arteries before CABG, CT provides surgeon-friendly three-dimensionally reconstructed images not only of the aorta and its major branches but also of bypass conduit candidates such as the ITA, radial artery, gastroepiploic artery and greater saphenous vein. Due to these aspects, CT angiography has a potential to be used for preoperative evaluation before cardiac surgery. However, only few reports have dealt with this issue. This study showed a promising and positive result in that preoperative CT angiography affected operative strategy and perioperative management including follow-up planning in a substantial proportion of patients.

Atherosclerotic change of the thoracic aorta is a well-known risk factor of stroke that is one of the most grave complications of CABG [14–17]. Significant atherosclerotic change of the ascending aorta is found in as many as 25% of adult patients undergoing cardiac surgery and necessitates change in surgical decision making in many patients [16]. However, there is no clinical factor that is strongly associated with aortic atherosclerosis and thus enables us to predict its presence before surgery [18]. Although intra-operative trans-oesophageal echocardiographic examination of the aorta is helpful, the sensitivity to find out ascending aortic lesion is not high [19]. Epi-aortic US, although it is the most sensitive modality, cannot help risk assessment or planning because it can be performed only after sternotomy. On the contrary, our result shows that severity of aortic atherosclerosis determined by CT angiography may help not only surgical decision but also preoperative determination of risk. Based on current results, we are planning future clinical research, which will focus on the relationship between the severity of aortic atherosclerosis and postoperative outcome after CABG in a larger number of patients. We admit that grouping of the aortic atherosclerosis devised by ourselves may be rather arbitrary and, because we considered only such CT findings as mobile plaque/protruding atheroma/ulcer/calcification, non-calcified intramural atheroma that can be found by epi-aortic US with very high sensitivity may be overlooked. This issue would also be addressed in future studies.

Our strategy for CABG has been to perform off-pump surgery using bilateral ITAs. The ITAs are used as a composite graft (free right ITA anastomosed to the left ITA pedicle in Y- or I-configuration) to realise no-aorta-touch total arterial revascularisation in as many patients as possible. When additional graft is needed, the greater saphenous vein is used in most patients. Although we prefer off-pump CABG, conventional on-pump technique is liberally used when the target coronary arteries are diffusely diseased and small or when haemodynamic condition deteriorates during surgery. The surgeon’s threshold of using cardiopulmonary bypass is strongly affected by the status of the thoracic aorta assessed by preoperative CT or epi-aortic US that is routinely done for the patients, who did not undergo preoperative CT angiography. When cardiopulmonary bypass has to be used in patients with severely diseased ascending aorta, manipulation of the ascending aorta was avoided or minimised by appropriate alteration of the operative technique; arterial cannulation at alternative sites, anastomosis on the beating heart or hypothermic fibrillation or replacement of the ascending aorta. We think that these strategies worked well because the incidence of stroke was low despite the high prevalence of severe aortic atherosclerosis in our patients. Strokes that developed in our series were not embolic and all were attributed to intracranial or vertebral artery stenosis. Permanent neurological deficit remained in only one patient. Based on our results, we agree with those who contend that patient stratification based on aortic atheroma burden should be addressed to tailor treatment strategies in the era of off-pump CABG. Although there has been controversy about the benefit of off-pump CABG versus on-pump CABG, there are patients in whom off-pump surgery, especially with no-aorta-touch technique, is definitely advantageous or mandatory [15,16]. We believe that CT angiography is the most useful tool in detecting such patients before they enter the operating room.

In concordance with previous studies [7–13], the prevalence of systemic arterial lesions or aortic aneurysm was high (one-third) in our series and about 25% of them underwent surgical or interventional treatment simultaneously with or early after CABG. We think that CT angiography, by giving abundant information with just one-stop examination, helped to establish fast and efficient treatment plans. In addition, it also offers useful baseline information about patient’s atherosclerotic burden that helps to tailor intensity of risk factor control and follow-up after surgery.

Another advantage of CT angiography is that it can find incidental but potentially important pathology of the organs other than cardiovascular system. Our incidence is similar to the results of previous studies that reported 13.1–20.4% incidence of significant extracardiac findings at cardiac CT angiography performed in non-surgical patients or after CABG [20,21]. In our series, 23% of patients needed preoperative consultation or regular postoperative follow-up by other specialties for incidental non-cardiovascular findings including eight cases (2.8%) of hidden malignant neoplasm. Because we did not include in this study several patients in whom CABG was cancelled due to incidental finding of inoperable malignancy at preoperative CT, the actual incidence of hidden malignancy in patients, who were consulted for CABG, would be higher. A pulmonary nodule was the most frequently found benign incidental lesion. This can be explained by tuberculosis that was one of the most prevalent diseases in our country until 3 decades ago. Although the lesion was benign, most patients were informed of the presence of pulmonary nodule for the first time and the CT findings would be the control data for later follow-up [22].

While our results showed that preoperative CT angiography considerably affects surgical decision making and perioperative management in CABG patients, we could not address the following questions: (1) Does it helps to improve early outcome after CABG? (2) Is it cost-effective in every patient? If it is not, which patients can get benefits from undergoing preoperative CT? (3) Is it safe in terms of radiation exposure [23,24] and contrast dye load? (4) Is atherosclerotic burden assessed by CT an independent risk factor for long-term prognosis after CABG? With regard to the first question, our study cannot provide an appropriate answer. Although the early mortality rate was lower in the patients who underwent preoperative CT angiography than in those who did not (1.8% vs 5.3%), the comparison is not fair. It is because many patients who were not evaluated by CT had well-known risk factors of mortality such as renal dysfunction and severe congestive or angina symptoms. There was another bias such as excluding many patients who were younger than 50 years. To address this issue, we are planning future studies that will investigate the difference in early outcome between well-matched groups according to whether they undergo preoperative CT or not.

Considering the theoretical risk of radiation-associated cancer, an argument can be made against safety of routine CT angiography. Especially in younger patients, the risk of cancer is higher and cost-effectiveness can be an issue because the incidence of significant incidental findings was relatively low. Our rationale and background of performing CT angiography in as many as patients are as follows: there is no strong clinical predictor of atherosclerotic burden of the aorta and its branches that can help to stratify the need of CT; impact of incidental findings, if present, outweighs the risk of radiation-associated cancer; the risk of cancer associated with current CT protocol is rather theoretical than proven with epidemiological evidence; the cost and waiting period taken for CT are much less than those of multiple radiological examinations such as carotid and vascular US and MR imaging; and CT is being widely performed in our country as a part of health-promotion surveillance check. In our country, CT angiography of the aorta and its branches costs less than 300 euros (about 3% of total cost for CABG) and is reimbursed by the government-monopolised health insurance. Considering that CT angiography was not added to carotid and vascular US and sometimes cerebral MR angiography (all of them are reimbursed) but replaced them as the pre-CABG evaluation tool, the additional cost burden can be minimal or even negative. Nevertheless, studies investigating the incidence of cancer and cost-effectiveness of CT angiography should be performed in a large number of patients and it will take up a considerable amount of time.

In conclusion, neck-to-leg CT angiography performed for preoperative evaluation of CABG patients had a positive clinical impact on surgical decision making and perioperative management in half of patients. Considering the convenience of examination and cost-effectiveness of CT, we recommend its use as a routine evaluation tool before CABG at least in patients older than 60 years unless contraindicated.

Appendix A

Conference discussion

Dr S. Attaran (Liverpool, UK): I would like to know who pays for the CT's in your hospital?

Dr Park: The payer?

Dr Attaran: The payment for each CT.

Dr Park: The cost of a CT scan in our country is not expensive. It is about 300€. In the case of coronary artery bypass grafting patients, 90% are paid by the government-controlled insurance.

Dr Attaran: Well, as you mentioned, CT angiogram is becoming more and more popular and advanced in detecting cardiovascular disease. However, considering what your findings were, I have some concerns which I would like to address. First of all, you mentioned that the strategy has been changed in 50% of the cases, but there is no control group who didn’t have the CT scan to see how much operative strategy would have changed simply with history taking, examination chest X-ray, ultrasound or epi-aortic scanning. You mentioned that 13% of your cases did have calcification of the aorta. Did you compare it with the rate of detection of calcification by palpation or epi-aortic scanning?

Also regarding the conduit choice, yes, with a CT scan you can see everything, but with a simple examination you can detect the varicose veins, as we have been doing. For mammary arteries, as you know, there is no correlation between the stenosis of the subclavian artery and the flow of the mammary artery. So I don’t believe that a CT scan can guide us for a better choice of conduit.

And also, an incidental finding of the cancer patient — it is a very novel approach, but it is impractical to put every single patient coming to a cardiac department at the risk of radiation and the cost to identify cancer.

I would also like to comment on the mortality rate. Your mortality rate was 1.8%, same mortality rate reported nationally and internationally. You have compared the mortality rate with the group that were declined a CT scan because of renal impairment or other comorbidities, which clearly increases the EuroSCORE. Therefore, I do not believe that CT scan in everybody preoperatively is able to decrease the mortality.

Finally, I believe that a postop CT angio of the heart is more helpful — what is your view on that?

Dr Park: Actually our strategy is not changed by the CT scan finding. I am trying to use bilateral internal thoracic artery grafts, in the form of the composite arterial grafts. About 80% of our patients undergo pure arterial revascularisation based on the bilateral internal thoracic arteries. However, if the heart is so large and the patient is small and the right internal thoracic artery shows a rather thin diameter, in that case we need a third graft. So in that case you have to worry about the state of the ascending aorta. Although epi-aortic scan is a very useful tool in assessing the aorta, it can be done only after opening the chest. CT scan can help us to decide before surgery whether to cut the vein or to use the gastroepiploic artery as a pedicle graft. So I think in that respect the CT scan can help us to determine the operative strategy. And because we use the left internal thoracic artery as the single source of all coronary territories, it is like the new left main coronary artery, so its importance is very, very large. It is not like a single internal thoracic artery plus multiple saphenous vein grafts.

So I think that even 50% stenosis, or even if the stenosis is not severe but it has unstable plaque, the prognostic impact should be large.

Your second question?

Dr Attaran: Don’t you think that postop CT angio of the heart might be even more useful?

Dr Park: We are doing a postop CT scan. However, all the information provided by CT angio would be more useful if known before surgery.

References