Microemboli During Carotid Angiography
Association With Stroke Risk Factors or Subsequent Magnetic Resonance Imaging Changes?
Background and Purpose Carotid angiography is associated with a 1% risk of major stroke. Recently, transcranial Doppler ultrasonography (TCD) has shown cerebral microemboli during carotid angiography. To determine their significance, we correlated the number of microemboli during angiography with clinical characteristics, angiography findings, and preangiography and postangiography cerebral magnetic resonance imaging (MRI).
Methods One middle cerebral artery was monitored with TCD in 24 patients during angiography for carotid territory ischemia. The number of microemboli was correlated with angiographic and clinical characteristics. T2-weighted cerebral MRI was performed before and ≤48 hours after angiography, and the number of new ischemic lesions was determined in a blinded review.
Results Microemboli were seen in all patients, with an average of 51 per procedure (range, 12 to 154). The majority of microemboli had signal characteristics typical of air. Sixteen of the 24 patients had both preangiography and postangiography MRI. One of 24 patients had an angiographic stroke, with a single new thalamic lesion on MRI. No other patient had a new lesion. The microembolus count correlated with the angiographic contrast volume (P<.001) but not with any other radiological or clinical characteristic.
Conclusions This study confirmed the presence of numerous cerebral microemboli during carotid angiography. The microembolic signal characteristics and the correlation with contrast volume indicate that introduced air is the cause. These microemboli are usually clinically silent and do not lead to new changes on cerebral MRI.
Cerebral infarction is an uncommon but potentially devastating complication of carotid angiography, with ≈1% of patients sustaining a permanent neurological deficit. Minor stroke or TIA occurs in an additional 3% of cases.1 The precise mechanism is often unclear, but a number of risk factors have been identified, including age, carotid artery stenosis, the presence of a preexisting neurological deficit, the length of the procedure, the volume of contrast injected, and the number of catheters used.2 Recent studies3 4 5 with TCD have shown frequent high-intensity transient signals recorded from the MCA during routine carotid angiography, but the clinical significance of these is unknown. These signals are seen frequently during cardiac bypass surgery6 7 and are encountered in other clinical settings such as atrial fibrillation,8 9 10 symptomatic and asymptomatic carotid stenosis,11 12 and patients with prosthetic heart valves.13 14 There is now considerable in vitro and in vivo evidence that these high-intensity transient signals are caused by microemboli, either particulate or gaseous.15 16 An average of 73 microemboli were observed during each procedure in the largest study,5 and they occurred in all phases of the angiogram as well as during the injection of contrast.
There is evidence that microemboli seen during angiography are gaseous,4 but some of the signals are of lower intensity, confined to the Doppler blood flow spectrum, and might be accounted for by particulate matter. Furthermore, microscopic air embolism might not be benign.17
To determine the significance of microembolisms during carotid angiography, we conducted a combined TCD and MRI study. We aimed to correlate the number of microemboli with the reported risk factors for angiographic stroke and to determine whether microemboli occurring during carotid angiography were associated with new cerebral lesions detectable by high-resolution MRI.
During a 12-month period, we studied 24 individuals chosen from elective intra-arterial digital subtraction angiography patients being investigated for carotid territory TIAs (13 cases ) and ischemic strokes (11 cases). During that period, >300 angiograms were performed; this series represents a necessarily small sample because of the severe logistical constraints of the study protocol. Patients with atrial fibrillation were excluded. Patients were also excluded if it was likely that they would not have completed the protocol because of factors such as MRI unavailability on weekends (affecting all Friday cases), angiography scheduled for the evening, or a high likelihood of immediate carotid surgery. Patients with suspected severe carotid stenosis were not excluded necessarily, but only if they already had an endarterectomy scheduled for soon after admission, leaving no time for completion of the study protocol. The study was approved by the hospital's medical research and ethics committees, and informed consent was obtained from all patients.
A neurological examination was performed by a neurologist just before the investigations. Cerebral MRI was performed in the 24 hours before angiography. T2-weighted axial images were obtained by use of a modified version of the MRI protocol for a trial of β-interferon in multiple sclerosis, so that an exact slice-for-slice correlation of the preangiography and postangiography MRI scans could be made.18 Intra-arterial digital subtraction angiography was performed by one of three staff radiologists or one of four radiology fellows. Six patients had unilateral selective carotid angiography, and the remainder had bilateral carotid angiography. Views of the carotid bifurcation were obtained in at least two planes, as well as anteroposterior and lateral cerebral images. Carotid bifurcation stenosis was measured according to the North American Symptomatic Carotid Endarterectomy Trial criteria.19 Details of the angiogram procedure were recorded on standard forms. The volume of contrast was the total of angiographic contrast injections in the arch and on the monitored side.
TCD monitoring (Medasonics CDS, 2 MHz pulsed-wave probe) of one MCA was performed for a 15-minute epoch immediately before angiography, with the probe fastened over the temporal acoustic window by a hook-and-loop fastener headband. The signal was recorded on audiotape with a Tascam 134 Syncaset for later review. Monitoring was continuous during the angiogram, and the elapsed time of the tape recording was referenced to the stages of the procedure, including contrast injections and catheter manipulations. Bilateral recording was not performed because it was not possible to keep two radiopaque transducers from obscuring the lateral cerebral angiogram. The MCA of the symptomatic hemisphere was monitored when possible. The asymptomatic side was monitored when the symptomatic MCA signal was unobtainable. The six patients who had unilateral carotid angiography were monitored on the symptomatic side. Tape recordings of the TCD data were reviewed later by an observer blinded to the clinical and angiographic findings. High-intensity transient signals (microemboli) were counted by use of criteria recommended by the Consensus Committee of the Ninth International Cerebral Hemodynamics Symposium.20 The total number of microemboli for the entire angiogram was used for the analysis. Microemboli were not subclassified into gaseous and possibly particulate categories.
Further neurological assessments were made by a neurologist immediately after angiography and again 4 and 24 hours later. Formal neuropsychological assessments were not performed. Repeat cerebral MRI was performed on the next day, between 21 and 33 hours after the angiogram. The number of new lesions on postangiography MRI was determined by a neuroradiologist. To ensure blinding, preangiography and postangiography MRI scans of each patient were displayed side by side, with the earlier or later scan randomly assigned to the left position on the viewing screen. Opaque film masks were made to obscure the date and patient identification information on each slice. Corresponding lesions on equivalent slices were recorded, and a new lesion was scored when there was no lesion at the corresponding site on the equivalent slice of the other MRI.
Linear regression was used to assess the relation between embolus counts and continuous patient- and procedure-related variables, and a logarithmic transformation was used in the analysis of carotid stenosis versus embolus count because of the large number of cases with 0% stenosis. Student's t test was used to compare the mean embolus counts in symptomatic and asymptomatic hemispheres, the mean embolus counts for angiograms examined by fellows and radiologists, and the mean embolus counts for a dichotomized percentage carotid stenosis result.
There were 16 men and 8 women aged 41 to 84 years (mean±SD, 63.7±10.7 years). Preangiography MRI was performed in all 24 patients, and 16 also had postangiography MRI. Eight patients withdrew consent for the second MRI because of a dislike of the procedure. The historical risk factors are recorded in Table 1⇓. None of the patients were taking anticoagulants.
The symptomatic MCA was monitored with TCD during carotid angiography in 14 patients and the asymptomatic in 10. There were no microemboli recorded during the 15-minute epochs before angiography. Cerebral microemboli were seen during angiography in all subjects. There were a total of 1214 microemboli recorded in the 24 patients, with a range of 12 to 154 (mean±SD, 50.6±33.8) per procedure (Table 2⇓). There was no difference in mean microembolus counts in symptomatic (43.6±29.2) versus asymptomatic MCAs (60.3±38.8). There was no difference in mean embolus counts for angiograms performed by fellows versus radiologists. The high-intensity signals were predominantly bidirectional, indicating amplitude overload,21 and occurred most frequently at the time of contrast injections (Fig 1⇓). As a rule, microemboli were detected only when the carotid artery being investigated angiographically was on the same side as the MCA that was being monitored, but in two bilateral carotid angiography cases, a microembolus was seen during selective angiography of the carotid artery contralateral to the monitored side. The signals occurred soon after contrast injection. In both cases, the symptomatic side was the monitored side, and in neither case was cross flow demonstrated on the cerebral views. There was no stenosis of the contralateral carotid artery. The precise number of microembolic signals during catheter manipulations alone was not recorded separately from the incidental occurrence of signals unrelated to any particular angiographic event.
The microembolus count correlated significantly with the volume of angiographic contrast injected on the monitored side (P<.001; Fig 2⇓). There was no association between microembolus count and the presence or degree of carotid stenosis, the duration of the procedure, age, sex, or history of diabetes, hypertension, or smoking.
One of the 24 patients had an angiographic stroke. At the end of carotid and vertebral angiography, a 70-year-old man developed severe vertigo. He was ataxic with horizontal nystagmus for the next 48 hours but made a complete recovery. The postangiography MRI showed a new thalamic infarct, consistent with posterior circulation ischemia. He had 54 microemboli recorded in the right MCA during the entire angiographic procedure. No other patient had a stroke or a new lesion on postangiography MRI.
The bidirectional high-intensity TCD signals seen during carotid angiography were due to gaseous microemboli.4 15 These signals were of higher intensity than the signals caused by particulate emboli in patients in atrial fibrillation or with carotid stenosis, and they appeared bidirectional because of amplitude overload.21 These intense signals resulted from the great difference in the acoustic impedance of air and blood and the specular properties of some air bubbles. Whereas small air microemboli may account for microembolic signals of lower intensity within the spectrum of the Doppler trace, even very large particulate emboli will not give rise to the embolic signals characteristic of air embolism. A reliable distinction between a gaseous or particulate origin for microembolic signals confined to the Doppler velocity spectrum is not possible in the context of angiography, or even in a closed vasculature.14
We found a significant association between microembolus count and the volume of angiographic contrast. The prevalence of microembolic signals at times of contrast injection also indicated that introduced microbubbles were the likely cause. Dagirmanjian et al5 recorded a similar number of microemboli per procedure (73) and also found that they were most numerous during injection of contrast or flushing of the catheter. Markus et al4 showed how microemboli during angiography can be reduced by altered contrast preparation and injection techniques. The occurrence of air emboli at other times, when there is no injection of saline or contrast, might be related to the inadvertent introduction of microbubbles after reinsertion of a guidewire or cavitation effects (air microbubbles being generated in an agitated fluid) due to the presence of intravascular instruments in fast-flowing blood.14
Gaseous microemboli have been documented during cardiopulmonary bypass,7 and the subtle neuropsychological deficits found in coronary artery graft surgery patients have been attributed to air embolism.17 Padayachee et al7 found microemboli were no longer detectable in the MCA of coronary bypass patients in whom membrane oxygenators were used, and with similar cardiopulmonary bypass circuits, Walsh et al22 showed that there is no neuropsychological deterioration after coronary artery bypass surgery. Frequent gaseous microembolism might account for some of the transient neurological disturbances seen after angiography,4 17 but whether microemboli, seen in all our cases and all of those in previous studies, contribute to the uncommon occurrence of angiographic stroke remains unclear.
No previous study has evaluated MRI changes after cerebral angiography. In the first hour after elective coronary artery bypass graft surgery, brain swelling, but not focal lesions, has been seen on MRI.23 After carotid endarterectomy, new lesions seen on MRI in 4 of 40 patients correlated with a higher number of microemboli during intraoperative TCD.24 All 4 subjects remained asymptomatic. Despite numerous microemboli during carotid angiography in the present study, only one new brain lesion was seen in the 16 patients who had preangiography and postangiography MRI, and this was in a patient who suffered an angiographic cerebral infarct. The patient had angiography to investigate transient left-sided weakness with some mild giddiness, but nothing else to indicate vertebrobasilar ischemia. Although the total of 54 microemboli recorded in that patient's right MCA was close to the mean for the whole group, it is possible that there were more microemboli in the vertebrobasilar circulation. Unfortunately, the difficulty of insonating the posterior circulation precluded remote monitoring of the basilar artery during angiography.
There was no association between the microembolus count and any of the other reported risk factors for angiographic stroke. In the case of carotid stenosis, a link may have been missed because of the low number of cases with higher-grade lesions.
We conclude that cerebral microemboli, detected routinely during catheter angiography, are chiefly gaseous, clinically silent, and generally not associated with cerebral tissue changes on T2-weighted MRI.
Selected Abbreviations and Acronyms
|MCA||=||middle cerebral artery|
|MRI||=||magnetic resonance imaging|
|TCD||=||transcranial Doppler ultrasound|
|TIA||=||transient ischemic attack|
We gratefully acknowledge financial support from the Australian Stroke and Neuroscience Institute and from the Slezak Trusts. We acknowledge the great assistance of the angiography staff and the medical imaging technologists in the MRI unit, and also Dominique Cadilhac, RN, who performed some of the TCD recordings.
- Received December 8, 1995.
- Revision received June 4, 1996.
- Accepted June 4, 1996.
- Copyright © 1996 by American Heart Association
Hankey GJ, Warlow CP, Sellar RJ. Cerebral angiographic risk in mild cerebrovascular disease. Stroke.. 1990;21:209-222.
Dion JE, Gates PC, Fox AJ, Barnett HJM, Blom RJ. Clinical events following neuroangiography: a prospective study. Stroke.. 1987;18:997-1004.
Pugsley W. The use of Doppler ultrasound in the assessment of microemboli during cardiac surgery. Perfusion.. 1986;4:115-122.
Padayachee TS, Parsons S, Theobold R, Linley J, Gosling RG, Deverall PB. The detection of microemboli in the middle cerebral artery during cardiopulmonary bypass: a transcranial Doppler ultrasound investigation using membrane and bubble oxygenators. Ann Thorac Surg.. 1987;44:298-302.
Tegeler CH, Hitchings LP, Eicke M, Leighton J, Fredericks RK, Downes TR, Stump DA, Burke GL. Microemboli detection in stroke associated with atrial fibrillation. J Cardiovasc Tech.. 1990;9:283-284. Abstract.
Tong DC, Bolger A, Albers GW. Incidence of transcranial Doppler-detected cerebral microemboli in patients referred for echocardiography. Stroke.. 1994;25:2138-2141.
Sliwka U, Job FP, Wissuwa D, Diehl RR, Flaschkampf F-A, Hanrath P, Noth J. Occurrence of transcranial Doppler high-intensity transient signals in patients with potential cardiac sources of embolism: a prospective study. Stroke.. 1995;26:2067-2070.
Siebler M, Kleinschmidt A, Sitzer M, Steinmetz H, Freund H-J. Cerebral microembolism in symptomatic and asymptomatic high-grade internal carotid artery stenosis. Neurology.. 1994;44:615-618.
van Zuilen EV, Moll FL, Vermeulen FE, Mauser HW, van Gijn J, Ackerstaff RG. Detection of cerebral microemboli by means of transcranial Doppler monitoring before and after carotid endarterectomy. Stroke.. 1995;26:210-213.
Georgiadis D, Mackay TG, Kelman AW, Grosset DG, Wheatley DJ, Lees KR. Differentiation between gaseous and formed embolic materials in vivo: application in prosthetic heart valve patients. Stroke.. 1994;25:1559-1563.
Russell D, Madden KP, Clark WM, Sandset PM, Zivin JA. Detection of arterial emboli using Doppler ultrasound in rabbits. Stroke.. 1991;22:253-258.
Markus HS, Brown MM. Differentiation between different pathological cerebral embolic materials using transcranial Doppler in an in vitro model. Stroke.. 1993;24:1-5.
Moody DM, Bell MA, Challa VR, Johnston WE, Prough DS. Brain microemboli during cardiac surgery or aortography. Ann Neurol.. 1990;28:447-486.
Paty DW, Li DKB, UBC MS/MRI Study Group, IFNB Multiple Sclerosis Study Group. Interferon beta 1b is effective in relapsing-remitting multiple sclerosis, II: MRI analysis results of a multicenter, randomized, double-blind, placebo-controlled trial. Neurology.. 1993;43:662-667.
North American Symptomatic Carotid Endarterectomy Trial (NASCET) Steering Committee. North American Symptomatic Carotid Endarterectomy Trial: methods, patient characteristics, and progress. Stroke.. 1991;22:711-720.
Consensus Committee of the Ninth International Cerebral Hemodynamic Symposium. Basic identification criteria of Doppler microembolic signals. Stroke.. 1995;26:1123. Special Report.
Walsh K, Drury J, Worcester M. Cognitive function following coronary artery bypass surgery. In: Donnan GA, Berkovic SF, Vajda FJE, eds. Stroke. Proceedings of workshop held at the Austin Hospital, John Lindell Lecture Theatre, Melbourne, Australia. 1988:133-138.
Jansen C, Ramos LM, van Heesewijk JP, Moll FL, van Gijn J, Ackerstaff RG. Impact of microembolism and hemodynamic changes in the brain during carotid endarterectomy. Stroke.. 1994;25:992-997.