Best Evidence Topics
  • Send this BET as an Email
  • Make a Comment on this BET

Can epiaortic ultrasound reduce the incidence of intraoperative stroke during cardiac surgery?

Three Part Question

In [patients undergoing cardiac surgery] does [epiaortic ultrasound] reduce the incidence of [intra-operative stroke]?

Clinical Scenario

You are about to perform an on-pump quadruple coronary arterial bypass graft on a 75 year old lady with a poor ejection fraction. In addition, she has had a carotid endarterectomy 5 years ago and is known to have some peripheral vascular disease, and thus you are anxious that she may have an atherosclerotic ascending aorta, which could cause an intra-operative stroke. You would like to use your usual surgical technique of cross-clamping but you wonder whether using intra-operative ultrasound to decide on your operative approach would reduce the chance of stroke for your patient.

Search Strategy

Medline 1966-Aug 2003 using the OVID interface.
[ OR OR aortic ultraso$.mp] AND [cardiac or OR exp cardiac surgical procedures/ OR coronary arter$ bypass graf$.mp] AND [maximally sensitive RCT filter]

Search Outcome

179 papers were found of which 6 were deemed to be relevant in documenting papers that studied modifications to cardiac surgery on the basis of epiaortic ultrasound. In addition checking of the reference lists of these papers and hand searching the latest editions of cardiac surgical journals found an additional 2 papers. These are presented in the table.

Relevant Paper(s)

Author, date and country Patient group Study type (level of evidence) Outcomes Key results Study Weaknesses
Goto et al,
463 pts over 60 years old undergoing elective CABG, with management of aorta planned according to EAU results, including either moving area for cross clamping or no touch technique Aorta graded by 2 independent blinded investigators : 0 almost normal; 1 mild, <3mm intimal thickening 2 mod, >3mm intimal thickening one segment 3 severe, 3mm intimal thickening more than 1 segmentCohort study (level 2b)Incidence of atherosclerosis of aortaMod. atherosclerosis: 57 pts, (12%)
Severe atherosclerosis 76 pts (16%)
No clinical benefit shown for modification of surgical technique according to EAU findings
Poor post op neurophysiological score determined by 2 S.D. drop in Hasegawa-dementia scoreSevere atheroma 20/76 pts (26%)
Moderate atheroma 4/57 pts (7%)
No atheroma 27/330 pts (8%) p<0.001
Intraoperative CVASevere atheroma 8/76 pts (10.5%). (But none in the 10 pts operated by no touch technique)
Moderate atheroma 1/57 pts (1.8%)
No atheroma 4/330 pts (1.2%) p<0.001
Shimokawa et al,
155 consecutive pts undergoing off pump CABG Patients divided into 2 groups on the basis of epiaortic ultrasound interpreted by surgeon Group A: atherosclerotic findings in the anterior ascending aorta Group B : Non atherosclerotic findings in the anterior ascending aortaProspective cohort study (level 2b)Incidence of atherosclerosis in the anterior ascending aorta on EAU54 patients (35%) allocated to Group A, of which 19 had local disease and 35 had extensive diseaseNo inter-observer agreement or validation of allocation to Group A or B made Study is underpowered to conclude that EAU followed by a no touch technique reduces incidence of stroke
Management change on basis of EAU29 patients (19%) had grafting without partial cross clamping 13 (8%) patients had placement of partial crossclamp in a different location
Complications1 thoracic aortic dissection (had normal EAU). 1 CVA in a patient in Group A who had a no cross clamp technique operation
Hangler et al,
Epiaortic scanning performed in 352 consecutive patients undergoing primary CABG Patient's operation was modified according to EAU findings : (using an optional guideline given to surgeons) normal/mild (aortic wall 3 mm), - no modification moderate (aortic wall 3 to 5 mm) - single cross clamp or alternative cross clamp site severe (aortic wall thickness > 5 mm and/or marked calcification or mobile plaque) - no touch technique used, OPCAB or equivalentRetrospective cohort study (level 2b)Incidence of aortic atherosclerosisNormal or mild 151 patients (42.9%), moderate 167 patients (47.5%), severe 34 patients (9.6%)40% of all CABG patients operated on at this institution during this time period The group with severe atheroma were older, had higher Hx of stoke and higher Euroscore (although significant findings still found after logistic regression performed) 4 patients with severe atheroma did not have their operation modified as per protocol
Incidence of post operative stroke, confirmed by CT scanningNormal or mild 3 patients (2%), moderate 4 patients (2.4%), severe 1 patient (2.9%) p=0.935
Post operative mortalityNormal or mild 0 patients, 0%, moderate 5 patients (3.0%), severe 3 patient (8.8%) p=0.005.
Post operative MINormal or mild 3 patients (2%), moderate 5 patients (3%), severe 5 patient (15%) p=0.001
Van der Linden et al,
921 consecutive pts undergoing elective cardiac surgery in 2 centres (including 26.5% of patients requiring valve surgery) Patients categorised by EAU performed by surgeon and anaesthetist: Atheroma absent Atheroma present if intimal thickening >0.5 mm Modifications in operation included changes in position of cross-clamps, cannulae and proximal anastomosesProspective cohort study (level 2b)Identification of atheroma by manual palpation in patients with atheroma on EAUOnly 39.6% of the 241 patients with significant atheroma on EAU had palpable atheromaHigh overall stroke rate of 3.5% Dichotomous categorization of atheroma severity Valve procedures in this cohort introduces the possibility that air embolism may have been the cause of some CVAs
Post-operative stroke confirmed by CT scan and neurologistNo atheroma 12/680 (1.8%)
Atheroma present 21/241 (8.7%) p = 0.001
Royse et al,
47 patients who underwent elective multivessel CABG without concomitant procedures from 1997-1998. Two groups were studied which represented 2 surgeons normal clinical practise Echo/Y group (N=26) Had EAU and total arterial revascularisation with Y grafting. Single cross clamp technique Control group (N=21) manual palpation of aorta. Pts received the standard aorto-coronary bypass operation for multivessel coronary revascularizationCohort study (level 4)20% deterioration in late post operative neuropsychological function. 10 neuropsychometric tests were performed by a neuropsychologistEcho/Y group 3.8% pts deteriorated. Control group 38.1% pts deteriorated p=0.006The surgeon operating on the Echo/Y group was different from the surgeon in the control group Carbon dioxide insufflation of sternotomy used only in Echo/Y group Small study assessing two single surgeon's practise
Intra-operative transcranial Doppler assessment of total gas or particulate emboliEcho/Y group 10.5+/- 3.2. Control group 5.5+/- 2.2 p=0.341
Normal aorta on EAU in Echo/Y group or manual palpation in control groupEcho/Y group 35% of anterior zones free of disease. Control group 83% of aortas determined as being normal on manual palpation
Ura et al,
496 patients undergoing open heart surgery had epiaortic ultrasound before and after decannulation Grade 1 intimal thickening <3 mm Grade 2 if the intimal thickening 3-4mm Grade 3 thickening >4 mm, Management changed according to findings, either cannula movement or beating heart or fibrillating technique usedCohort study (level 2b)Incidence of atheroma of aortaGrade 1: 371 patients (74%). Grade 2: 73 patients (15%). Grade 3: 28 patients (5.6%).No uniform protocol for management of patients with grade 2 or 3 atherosclerosis Unclear as to who graded aorta and whether interobserver agreement was assessed
New lesion identified after reassessment post decannulation6 new minor lesions. 10 mod/severe new lesions (6 due to cross clamping and 4 due to aortic cannula). 3 of these patients had a stroke
Incidence of stroke determined by neurologist and CT scanGrade 1: 2 patients (0.5%). Grade 2: 2 patients (2.7%). Grade 3: 6 patients (21%).
Hammon et al,
395 patients undergoing CABG. All patients had neuropsychological testing pre and postoperatively. Group A: 203 patients, manual palpation by surgeon for assessment of aorta Group B: 192 patients , EAU assessment of aorta and modification of operation according to these resultsRetrospective cohort study (level 4)Microemboli measured by continuous intraoperative carotid ultrasoundGroup A: Median 103 microemboli
Group B: Median 90 microemboli p=NS
P value for 1 week neurobehavioural findings calculated by us to be p=0.0735 using 2-sided fishers exact test. This paper did not document which statistical tests performed. (1 month test correct) Two groups were created retrospectively , rate of EAU use in group B not stated
Decline in neurobehavioural testing after battery of 11 tests applied pre and post operatively at 1 weekGroup A: 140/203 pts (69%)
Group B: 115/192 pts (60%). p<0.05
Decline in neurobehavioural tests at 1 monthGroup A: 52/180 pts (29%)
Group B: 35/198 pts (18%). p<0.01
Duda et al,
359 consecutive patients undergoing CABG 195 patients were assessed by Epiaortic Ultrasound and the operation was modified according to results. Modifications included change to aortic cannula site, proximal graft sites, fibrillatory arrest or femoral bypass Retrospective control group of 164 patients who underwent manual palpation only for assessment of atherosclerosisCohort study (level 2b)Detection of significant atherosclerosisEAU group moderate in 20 patients (10%) severe in 7 pts (3.5%). Control group 3 patients found to have atheroma on palpationNon randomized retrospective control group Small sample size for assessment of stroke as primary outcome measure
Post-operative stroke based on permanent neurological deficit immediately after operation and mortalityEUA group: No strokes, 5 deaths (2.6%)
Control group 5 strokes (3.0%), 6 deaths (3.6%) p<0.02 for stroke
Number of patients for which the operation was modified due to atheromaEUA group 19 patients (10%)
Control group 3 pts (3%)


There are several areas of evidence that need to be considered prior to using epiaortic ultrasound (EAU) as part of routine surgery. First, it is well established that EAU detects a far higher level of atherosclerosis in the ascending aorta compared to manual palpation. Van der Linden reported that only 39.6% of significant atheroma detected on EAU was detected by manual palpation. Duda reported that when EAU was not used 1.8% of patients were thought to have significant atherosclerosis of the aorta on manual palpation, but when EAU was used 14% of patients were found to have significant atherosclerosis. The incidence of moderate atherosclerosis found by EAU ranged from 10% to 48% and severe atherosclerosis ranged from 3.5% to 16% in these studies. Second, it is well established that patients with moderate or severe atherosclerosis of the aorta have a significantly higher incidence of stroke. Van der Linden found that presence of atheroma increased the stroke rate from 1.8% to 8.7%. Ura found that severe atheroma increased the incidence of stroke from under 3 % to 21%. Goto found that neurobehavioural outcome was significantly worse in the presence of severe atherosclerosis, with 28% of these patients having a decrease in neurophysiological tests compared to only 8% in groups with normal or moderate scores. Many other papers not documented here also support these findings. Finally the more controversial issue is whether modifying the management of the ascending aorta based on epiaortic ultrasound can effectively reduce the incidence of stroke in patients attending for Cardiac Surgery. Goto was unable to reduce the incidence of stroke by a range of modifications in technique although they report that in the 10 patients with severe atherosclerosis where a no-touch technique was used, no strokes occurred. Shimokawa operated on 19% of their cohort of patients using a no touch technique for high risk patients, although 1 of these 54 patients had a stroke. Hangler performed off pump CABG with a no touch technique in their patients with severe atherosclerosis and had a low stroke rate of only 2.8% in these patients, but Van der Linden who used more minor modifications in technique including relocating the aortic cannulation site or site of proximal anastomosis had a stroke rate of 8.7%. Royse used total arterial revascularisation with Y grafting for high risk patients and found a drop from 38 % to 4% in abnormal neurobehavioural testing post-operatively. Ura used more minor modifications and still used at least a single cross clamp and could not improve the stroke rate, but Hammon reported that when epiaortic ultrasound was introduced into their clinical practise, there was a significant improvement in neurobehavioural testing at 1 month post operation. Finally Duda reported that the introduction of EAU into their clinical practise eliminated stroke in their patients when used in 195 patients. Introducing EAU into their practise Duda's results showed that a modification to their usual technique was required for every 14 ultrasound scans performed. Results are mixed as to whether the stroke rate can be reduced as a result of EAU examination. 5 of the 8 studies showed a reduction in the stroke rate as a result of modification to surgical technique after EAU. It may be that the degree of modification to the surgeon's operative technique when atherosclerosis is detected is the determining factor in these papers as to whether a reduction in stroke rate can be achieved, with a no touch technique being of most benefit in the highest risk patients.

Clinical Bottom Line

Epiaortic ultrasound is superior to manual palpation in detecting atherosclerosis in the ascending aorta, and the severity of atherosclerosis found is closely correlated to the incidence of postoperative stroke. No touch techniques in patients with severe atherosclerosis may avert this increase in the incidence of stroke.


  1. Goto T, Baba T, Matsuyama K, et a;. Aortic atherosclerosis and postoperative neurological dysfunction in elderly coronary surgical patients. Ann thorac Surg 2003;75(6):1912-1918.
  2. Shimokawa T, Minato N, Yamada N, et al. Assessment of ascending aorta using epiaortic ultrasonography during off-pump coronary artery bypass grafting. Ann Thorac Surg 2002;74(6):2097-2100.
  3. Hangler HB, Nagele G, Danzmayr M, et al. J Modification of surgical technique for ascending aortic atherosclerosis: Impact on stroke reduction in coronary artery bypass grafting. J Thoracic Cardiovasc Surg 2003;126(2):391-400.
  4. van der Linden J, Hadjinikolaou L, Bergman P, et al. Postoperative stroke in cardiac surgery is related to the location and extent of atherosclerotic disease in the ascending aorta. J Am Coll Cardiol 2001;38(1):131-135.
  5. Royse AG, Royse CF, Ajani AE, et al. Reduced neuropsychological dysfunction using epiaortic echocardiography and the exclusive Y graft. Ann Thorac Surg 2000;69(5):1431-14.
  6. Ura M, Sakata R, Nakayama Y, et al. Ultrasonographic demonstration of manipulation-related aortic injuries after cardiac surgery. J Am Coll Cardiol 2000;35(5):1303-1310.
  7. Hammon JW, Jr., Stump DA, Kon ND et al. Risk factors and solutions for the development of neurobehavioral changes after coronary artery bypass grafting. Ann Thorac Surg 1997;63(6):1613-1618.
  8. Duda AM, Letwin LB, Sutter FP, et al. Does routine use of aortic ultrasonography decrease the stroke rate in coronary artery bypass surgery? J Vasc Surg 1995;21(1):98-107.