- Email: [email protected]
Cerebral microemboli during left heart catheterization
Valvular and Congenital Heart Disease
Cerebral microemboli during left heart catheterization Andreas Fischer, MD,a Cem Özbek, MD,b Wolfgang Bay, MD,b and Gerhard F. Hamann, MD, PDa,c Homburg/Saar and Munich, Germany
Background The aim of this study was to describe the rate of microemboli signals (MES) during left heart catheterization (LHC).
Methods A monitoring of both middle cerebral arteries using transcranial Doppler ultrasonography was performed to investigate cerebral microemboli during LHC. Seventy-two patients undergoing LHC and 29 patients with LHC followed by coronary intervention were studied.
Results During a standardized LHC (n = 52), 95 ± 45 MES were detected of which 67.5% occurred during injection of contrast media or saline solution, 30% during movements of wire and catheter, and 2% during catheter manipulation. During coronary interventions only, rotablation (n = 2) was followed by a massive increase in MES. The use of injection fluids prepared with minor gas content reduced the MES rate by 67% (P < .05). All MES were clinically silent.
Conclusions Cerebral microembolism is a current finding during LHC. The dependence of the MES rate during diagnostic LHC on the gas content of the injection fluids provides evidence that most of the MES are caused by microbubbles and not by solid emboli. The high rate of MES during coronary rotablation may be explained by the formation of cavitation bubbles. The clinical results of the MES during LHC appear to be benign. (Am Heart J 1999;137:162-8.)
Neurologic complications during left heart catheterization (LHC) have an incidence of 0.1% to 1%.1-3 Clotting phenomena at the surfaces of guide wires and catheters with subsequent embolism into the brain circulation may be possible causes of these complications as well as the mobilization of vascular atheroma during catheter maneuvers.3,4 Air embolism during injection of contrast media or catheter flushing has been described in other angiographic techniques and must be taken into account.5,6 Finally, there are known adverse reactions to contrast agents causing damage of the blood brain barrier,7 injuries of the endothelium,8 or direct neurotoxic effects.9 Doppler ultrasonography was first used in 1965 for From the aDepartment of Neurology and bCardiology, University of the Saarland, and the cDepartment of Neurology, Ludwig-Maximilians-University. This paper funded by an institutional grant from the Department of Neurology and Cardiology, University of the Saarland, Germany. Submitted February 3, 1998; accepted May 1, 1998. Reprint requests: Gerhard F. Hamann, MD, PD, Department of Neurology, LudwigMaximilians-University, Marchioninistr. 15, D-81377 Munich, Germany. E-mail: [email protected] Copyright © 1999 by Mosby, Inc. 0002-8703/99/$8.00 + 0 4/1/91404
the detection of air bubbles in extracorporeal circuits.10 More recently, in vitro and animal studies were performed showing that Doppler emboli detection of atheroma, fat, platelet, and thrombus is also possible.11,12 Detection of emboli is based on the different acoustic impedances of air, platelet aggregates, or atheroma in respect to the blood. In the case of the passage of an emboli, this causes a sudden higher rate of reflection of the ultrasound beam and, therefore, an intensity increase of the Doppler signal, called microemboli signals (MES). Using transcranial Doppler ultrasonography (TCD), Spencer et al13 first reported nongaseous MES in 1990 in patients undergoing carotid endarterectomy. Since then, there have been a great number of studies demonstrating MES in carotid artery stenosis,14 atrial fibrillation,15 or prosthetic heart valves16,17 and also during cardiac bypass surgery.18,19 In angiographic techniques, MES have been reported in carotid and cerebral angiograms.5,6 The aim of this study was to perform continuous TCD monitoring during LHC to detect possible MES, and further to show the time of occurrence in respect to different catheter manipulations or possible accidental injections of air
American Heart Journal Volume 137, Number 1
bubbles and to prove the relation between MES and neurologic deficits.
Patients and methods One hundred one consecutive patients (70 men, 31 women aged from 24 to 83; median 63 years) undergoing LHC were studied in the cardiac catheterization laboratories of the University of Saarland in Homburg, Germany. Informed consent was obtained from all patients. The study was performed in accordance with the Helsinki declaration and after approval of an institutional committee on human clinical research. A 2-MHz TCD device (Multi Dop X, Fa. DWL Systems) with 2 separate channels was used for bilateral transcranial monitoring of the middle cerebral artery (MCA). The Doppler spectra were analyzed by 64-point fast-Fourier transform with color coded display. The emboli signals were recorded by an integrated emboli detection program (MDX TCD-7 emboli detection software). The detection threshold was put at 14 dB. At this level the random Doppler speckle did not trigger the emboli detection program and all audible and visible MES were detected. No additional tape recording was used. The ultrasound transducers were fixed at the temporal bone with a special metallic head band 10 to 15 minutes before the initiation of LHC. The depth of insonation of MCA was 50 to 55 mm, sample volume was 15 mm, and power and gain were adjusted to the lowest level to obtain blood flow spectra of low intensity. This allowed optimization of the dynamic range of the spectral analyzer for the detection of the high-intensity emboli signals. During the monitoring the examiner was able to identify emboli signals (1) visually on the monitor display with the Doppler spectra of both MCA and (2) by listening to the acoustic changes of the Doppler signal. At the same time, each MES was automatically stored by the integrated emboli detection software. The Doppler spectra around the detected event were displayed in the lower part of the monitor indicating exact time, side of occurrence, and signal intensity in steps of 4 dB. When the buffer of the monitor display was full, the spectra were written automatically to disk. Furthermore the examiner wrote down each registered emboli signal in a table, again indicating exact time, visual and acoustic quality, and current action of the catheter operator and patient. The monitoring ended immediately after LHC had finished. Directly before and after the catheterization a short neurologic exploration was performed including questions in respect to neuropsychologic functions (time, person, and place). During off-line analysis, the MES recorded by the automatic emboli detection program and the manual notations of the experienced examiner were compared. MES were selected by following the standard criteria of Spencer et al.13
Fischer et al 163
Artefacts from patient speech, coughing, or movements were excluded. Afterward, the MES were summed up in groups in respect to different catheterization maneuvers: (1) injection of contrast media or saline solution, (2) movements of the catheter with the wire inside, either both same direction or against each other, (3) maneuvers with the catheter in the aorta or left ventricle, or (4) signal without relation to any actual maneuver or injection. To analyze emboli signal amplitude, 250 emboli signals from maneuvers 1, 2, and 3 above were studied using the relative intensity increase in decibels calculated by the automatic emboli detection program. The determination of the MES rate during diagnostic LHC was performed in a subgroup of 52 patients who underwent a standardized diagnostic LHC (1 ventriculography, 1 angiogram of the right, and 3 to 4 angiograms of the left coronary artery) and who presented no other possible cerebral emboli risk factors. In 3 patients all injection fluids underwent short-term hypobaric pressure to reduce their gas content. Following the rule that a hypobaric pressure decreases the concentration of dissolved gases, we tapped the top of the injection syringe with the finger and applied a negative pressure suction of 5 seconds by pulling the syringe. Afterwards the gas fraction was eliminated. The emboli signal rate of this subgroup was compared with the patients with unprepared injection fluids. For statistical elaboration, the Mann-Whitney U test was used. The LHC was performed with the transfemoral technique (Judkins); only 1 patient was examined by the transbrachial approach (Sones). All patients received a sheath for introducing the catheter. Preshaped 5F or 6F polyurethane catheters (Cordis Corporation) and coated steel wires (Corotec Medizintechnik GmbH) were used. LHC consisted of ventriculography (40 mL contrast injected at 14 mL/sec by a power injector using a pigtail catheter). Afterwards the catheter was changed and 1 angiogram of the right and 3 to 4 angiograms of the left coronary artery were performed, all biplane. In coronary angiography, 8 to 9 mL contrast was quickly injected manually. Additionally, 12 patients underwent aortic root angiography (40 mL contrast, 18 mL/sec) and 15 patients had bypass angiography. Twentynine patients had coronary interventions. Guide catheters (7F, 9F, or 10F) were therefore used that rested at the coronary artery orifice while the different interventional catheters were introduced. Twenty-five patients had balloon dilation, 2 underwent directional atherectomy, and 2 had high-speed rotablation. During rotablation no vasodilator was injected, but in between the cutting sequences this injection was performed. During the cutting there was only the injection of saline solution for cooling of the rotor. The following contrast media were used: iopamidol (Solutrast 370, Byk Gulden) in 34 patients, iopromid (Ultravist 370, Schering) in 64 patients, and ioxagline acid (Hexabrix 320,
American Heart Journal January 1999
164 Fischer et al
Figure 1
Figure 2
Comparison of MES amplitude of 250 MES during injections of fluids, movements of catheter and wire, and coiling maneuvers with catheter. Doppler flow spectra of the MCA with typical microemboli signal (arrow) detected during coiling maneuver with cardiac catheter in aortic arch. Left scale: flow velocity (cm/s); right scale: Doppler signal intensity (dB), color coded.
Laboratoire Querbet) in 3 patients. The use of heparin (2500 IU intra-arterial, n = 76) or aspirin (100 to 500 mg orally, n = 69) was documented.
Results In all patients, cerebral MES were detected. During puncture of the femoral artery, passage of the abdominal aorta, or interruptions of the exploration up to 20 minutes, no MES were observed. Once the catheter was advanced beyond the aortic arch, MES were recorded. MES were related to coiling maneuvers with the cardiac catheter in the aortic root, aortic valve, or left ventricle to movements between catheter and guide wire and to injections of contrast media and saline solution. The MES were short (<0.1 second) unidirectional signals that occurred randomly in the cardiac cycle. They presented a relative intensity increase calculated by the emboli detection software of between 16 to 32 dB. They were accompanied by a typical acoustic sound (cracking or chirping). Fig 1 shows a typical cerebral MES occurring during a coiling maneuver with the catheter in the aortic arch. Receiver overload with bidirectional widespread Doppler signals was seen, especially after injection of
great volumes of contrast media or saline solution. Artefacts from head movement or speech also produced bidirectional Doppler signals, but were of longer duration and lower audio frequencies. The analysis of the relative intensity increase of the MES in respect to their association to injections of contrast, movements of wire, and catheter or coiling maneuvers with the catheter is shown in Fig 2. There were no significant differences of the relative intensity increase of the MES between these groups. A standardized diagnostic LHC showed a mean rate of 95 ± 45 MES. Nearly half of them occurred during ventriculography (45.5 ± 28 MES) and half of them during coronary angiography (49.5 ± 36 MES). Most of the MES (67.5%) were due to injections of contrast media and saline solution; 30% were related to parallel movements of the wire and the catheter. Only 2% of MES happened during coiling maneuvers with the catheter in the aorta or left ventricle (Fig 3). In the subgroup of patients who received prepared contrast media and saline solution with reduced gas content, we saw a significant reduction of 67% of the MES rate (P < .05). Only the MES related to injections of contrast or saline solution (65 versus 25 MES, P < .05) and parallel movements of catheter and wire (29 versus 3 MES, P < .01) were significantly decreased (see Fig 3). The different catheters or contrast media did not influence the MES rate significantly. Aortography (n = 12) showed 37 ± 18 MES and angiography of bypass (n = 15) up to 691 MES (mean
American Heart Journal Volume 137, Number 1
Fischer et al 165
Figure 3
Distribution of MES during standardized LHC (n = 55) and influence of injection of contrast media and saline solution with low gas content on mean MES rate.
90 MES), nearly all of them related to injections of contrast media. In angiography of the left internal mammarian artery, bypass MES were detected only in the left MCA. During coronary interventions, MES caused by coiling maneuvers with the thicker and less flexible guide catheter in the aortic root were not more frequent than during diagnostic LHC (0 to 7 MES, mean 1 MES per patient). Ninety percent of MES were due to contrast injections and flushing of the catheter with saline solution. During the interventional procedures themselves we observed only 2 MES during 143 balloon insufflations in 25 patients, during directional arterectomy (n = 2) only 1 MES, but during rotablation (n = 2) the first patient presented 684 MES (total time 150 seconds) and the second 238 MES (total time 60 seconds). There were more MES when proximal parts of coronary artery underwent rotablation. The MES rates in coronary interventions are shown in Fig 4. MES were only detected during the cutting sequences of the rotablation. In patients with sclerosis or stenosis of the aortic valve (n = 11), we found significantly higher rates of MES (8 MES versus 1.5 MES, P < .01) during coiling maneuvers with the catheter at the valve. The presence of aneurysm or thrombus of the left ventricle (n
= 7), atrial fibrillation (n = 6), prosthetic heart valve (n = 1), and carotid artery stenosis (n = 9) did not show any influence on the MES rate during maneuvers with the catheter inside the left ventricle. The use of aspirin or heparin did not result in significant changes in the detection rate of MES. None of the patients included in this study developed any clinical signs of cerebral ischemia or neuropsychologic deterioration. No major changes in the neurologic examination before and after the left heart catheterization were noticed.
Discussion This study demonstrates that a large number of Doppler MES in both MCAs is detectible during left heart catheterization. The MES presented the same characteristics that have been demonstrated in other clinical and in vitro Doppler emboli studies.11-16 Turbulences20 and artefacts from speech and movement21 are completely different. The MES had a minimum relative intensity increase of 16 dB. This is a relatively high threshold in comparison to other studies. But we used the relative intensity increase calculated by the automatic emboli detection software, which compares the signal intensity of the whole screen including areas free of Doppler spectra with the emboli signal. This
166 Fischer et al
American Heart Journal January 1999
Figure 4
Cerebral microemboli signals during coronary interventions.
leads to higher decibel values of MES intensity. Kaps et al17 reported similar experiences using the same emboli detection software. The determination of the underlying embolus material in our study is difficult because air bubbles might be detected as well as solid emboli. The accidental introduction of air when the contrast is drawn up and immediately injected is possible and has been described in cerebral and carotid angiography.5,6 In our study 70% of the MES were associated directly to the injection of contrast media and saline solution, so that there is great evidence that this part of the MES represents air emboli. The analysis of the relative intensity increase of each MES did not help in our study to identify gas emboli. With the use of in vitro and animal models it has been shown that air bubbles produce more intensive MES than solid emboli because of the greater difference of acoustic impedance between air and blood.11,12 But it was also demonstrated that the size of the emboli influence the intensity of its MES. In our study with unknown size and quality of the emboli, we did not know whether a signal of low intensity was due to a very small air bubble or a bigger solid embolus. The results of Markus et al5 in respect to cerebral angiography (consistent find-
ings of very high intensity increases of MES with a usual receiver overload) could not be confirmed in this study. Receiver overload was seen especially at the beginning of injections of great volumes of contrast media. These differences may be explained by a dispersion of the injected gas bubbles during their way from the aortic arch up to the MCA in comparison with the air emboli injected directly in the carotid artery in cerebral angiography. The introduction of gas bubbles into the cerebral arterial circulation during LHC was further supported by our finding that patients who received injections of fluids containing reduced gas presented a significant decrease in the MES rate of nearly 70%. Markus et al5 saw lower MES rates when the fluids were allowed to stand a few minutes before an injection. An important finding of our study was that not only the MES that occurred during the injections but also those occurring during the movements of wire and catheter were reduced. This suggests that multiple air bubbles are resting at the surface of the catheter and wire and they are mobilized afterwards by movements of the wire. The deposition of platelet aggregates at the catheter surfaces4 and its embolism is also possible, but may be reflected only in a small part of the MES.
American Heart Journal Volume 137, Number 1
Nevertheless, solid microemboli also seem to occur during LHC, especially in respect to coiling maneuvers with the catheter at the aortic arch or at the aortic valve. During these maneuvers the MES rate in patients with and without minor gas content of injection fluids was constant. Patients with sclerosis or stenosis of the aortic valve presented a significantly higher embolism rate during these maneuvers. The presence of hundreds of cerebral MES during coronary rotablation was an interesting finding of this study. One possible explanation is the continuous injection of saline solution for cooling during the rotablation, which might result in bubble transfer to the aorta. Another mechanism may be the backstream of emboli due to high local pressure formed by the high-speed (180,000 revolutions/min) rotablator. MES may reflect cavitation bubbles as well as small atheromatous particles. The clinical importance of the detected MES remains unclear. None of the patients included in this study developed cerebral ischemic symptoms, therefore the clinical results of the detected MES seem to be benign. There is only 1 case report of multiple MES caused by a compression of common carotid artery with immediate development of contralateral paresis,22 and the majority of MES studies did not show obvious clinical sequelae.13-17 One explanation of this lack of major clinical symptoms may be the size of microemboli. In animal models it was not possible to cut solid microemboli smaller than 300 to 500 µm, but emboli of this size were detected easily.11,12 In the case of use of glass microspheres, even the detection of particles of 5µm was shown.12 Consequently microemboli potentially may produce only very small cortical infarctions or, in the case of particles smaller than 10 µm, even the passage of the capillary bed is possible. Furthermore, gas emboli are dissolvable and may disappear on their way to the microcirculation. But what is the clinical impact of cerebral microinfarction? Changes of neuropsychologic function (over 50%) have been reported in patients with coronary bypass surgery and high rates of cerebral microemboli signals during the cardiopulmonary bypass.18,19 These MES have been related to gas bubbles introduced by the extracorporeal oxygenators. Neuropathologic studies in patients after bypass surgery showed great numbers of dilatations in cerebral arterioles and capillaries as possible indicators of persisting gaseous microemboli.23 The use of arterial filters in bypass surgery led to a reduction of
Fischer et al 167
the MES rate and neuropsychologic deficits.18 Therefore, in our opinion cerebral microembolism, caused by microbubbles or solid emboli, cannot be generally thought to be clinically harmless. During LHC a neuropsychologic impairment caused by hundreds of microemboli may be possible, especially when numerous consecutive LHC or coronary interventions like rotablation are carried out. Further studies have to consider whether there exists a correlation between high rates of MES and neuropsychologic dysfunction during left heart catheterization. We thank the team of cardiac catheterization laboratories, Department of Cardiology, University of Saarland. We also thank Mrs Seiche for her language editing of this manuscript.
References 1. Adams DF, Abrams HL. Complications of coronary arteriography. Cardiovasc Radiol 1979;2:89-96. 2. De Bono D. Complications of diagnostic cardiac catherization: results from 34,041 patients in the United Kingdom. Br Heart J 1993;70:297-300. 3. Bourassa MG, Noble J. Complication rate of coronary arteriography. Circulation 1976;53:106-14. 4. Takaro T, Hultgen HN, Littmann D, Whright EC, Oteen NC. An analysis of deaths occurring in association with coronary arteriography. Am Heart J 1973;86:587-97. 5. Markus HS, Loh A, Israel D, Buckenham T, Clifton A, Brown MM. Microscopic air embolism during cerebral angiography and strategies for its avoidance. Lancet 1993;341:784-7. 6. Gerraty RP, Bowser DN, Infeld B, Mitchell PJ, Davis SM. Microemboli during carotid angiography, association with stroke risk factors or subsequent MRI imaging changes? Stroke 1996;27:1543-7. 7. Sage MR, Drayer B, Dubois P, Heinz E, Osborne D. Increased permeability of the blood brain barrier after carotid Renografin-76. AJNR Am J Neuroradiol 1981;2:272-4. 8. Laerum F. Injurious effects of contrast media on vascular endothelium. Invest Radiol 1985;20;598-9. 9. Drayer BP, Velay R, Bird R, Albright R, Roberts L, Allen S, et al. Comparative safety of intracarotid iopamidol, iothalamate meglumine, diatroate meglume for cerebral angiography. Invest Radiol 1984; 19:212-8. 10. Austen WG, Howry DH. Ultrasound as a method to detect bubbles or particulate matter in the arterial line during cardiopulmonary bypass. J Surg Res 1965;5:283-4. 11. Russell D, Madden KP, Clark WM, Sandset PM, Zivin JA. Detection of arterial emboli using Doppler ultrasound in rabbits. Stroke 1991; 22:253-8. 12. Markus HS, Loh A, Brown MM. Detection of circulating cerebral emboli using Doppler ultrasound in a sheep model. J Neurol Sci 1994;122:117-24. 13. Spencer MP, Thomas GI, Nicholls SC, Sauvage LR. Detection of middle cerebral artery emboli during carotid endarteretomy using transcranial Doppler ultrasonography. Stroke 1990;21:415-23. 14. Siebler M, Sitzer M, Steinmetz H. Detection of intracranial emboli in
168 Fischer et al
15.
16.
17.
18.
patients with symptomatic extracranial carotic artery disease. Stroke 1992;23:1652-4. Tegeler CH, Hitchings LP, Eicke M, Leighton J, Fredericks RK, Downes TR, et al. Microemboli detection in stroke associated with atrial fibrillation. J Cardiovasc Technol 1990;9:283-4. Georgiadis D, Mallison A, Grosset DG, Lees KR. Coagulation activity and emboli counts in patients with prosthetic cardiac valves. Stroke 1994;25:1211-4. Kaps M, Hansen J, Weiher M, Tiffert K, Kayser I, Droste DW. Clinically silent microemboli in patients with artificial prosthetic aortic valves are predominantly gaseous and not solid. Stroke 1997;28:322-5. Pugsley W, Klinger L, Paschalis C, Treasure T, Harrison M, Newman S. The impact of microemboli during cardiopulmonary bypass on neuropychological functioning. Stroke 1994;25:1393-8.
Fischer et al
19. Clark RE, Brillman J, Davis DA, Lowell MR, Price TRP, Magovern GJ. Microemboli during coronary artery bypass grafting, genesis and effect on outcome. J Thorac Cardiovasc Surg 1995;2:249-58. 20. Spencer MP, Whisler D. Transorbital Doppler diagnosis of intracranial arterial stenosis. Stroke 1986;17:916-21. 21. Spencer MP. Detection of cerebral arterial emboli. In: Newell DW, Aaslid R, editors. Transcranial Doppler. New York: Raven Press; 1992. p. 215-30. 22. Khaffaf N, Karnik R, Winkler WB, Valetin A, Slany J. Embolic stroke by compression maneuver during transcranial Doppler sonography. Stroke 1994;25:1056-7. 23. Moody DM, Bell MA, Challa VR, Johnston WE, Prough DS. Brain microemboli during cardiac surgery or aortography. Ann Neurol 1990;28:477-86.
Cerebral microemboli during left heart catheterization Andreas Fischer, MD,a Cem Özbek, MD,b Wolfgang Bay, MD,b and Gerhard F. Hamann, MD, PDa,c Homburg/Saar and Munich, Germany
Background The aim of this study was to describe the rate of microemboli signals (MES) during left heart catheterization (LHC).
Methods A monitoring of both middle cerebral arteries using transcranial Doppler ultrasonography was performed to investigate cerebral microemboli during LHC. Seventy-two patients undergoing LHC and 29 patients with LHC followed by coronary intervention were studied.
Results During a standardized LHC (n = 52), 95 ± 45 MES were detected of which 67.5% occurred during injection of contrast media or saline solution, 30% during movements of wire and catheter, and 2% during catheter manipulation. During coronary interventions only, rotablation (n = 2) was followed by a massive increase in MES. The use of injection fluids prepared with minor gas content reduced the MES rate by 67% (P < .05). All MES were clinically silent.
Conclusions Cerebral microembolism is a current finding during LHC. The dependence of the MES rate during diagnostic LHC on the gas content of the injection fluids provides evidence that most of the MES are caused by microbubbles and not by solid emboli. The high rate of MES during coronary rotablation may be explained by the formation of cavitation bubbles. The clinical results of the MES during LHC appear to be benign. (Am Heart J 1999;137:162-8.)
Neurologic complications during left heart catheterization (LHC) have an incidence of 0.1% to 1%.1-3 Clotting phenomena at the surfaces of guide wires and catheters with subsequent embolism into the brain circulation may be possible causes of these complications as well as the mobilization of vascular atheroma during catheter maneuvers.3,4 Air embolism during injection of contrast media or catheter flushing has been described in other angiographic techniques and must be taken into account.5,6 Finally, there are known adverse reactions to contrast agents causing damage of the blood brain barrier,7 injuries of the endothelium,8 or direct neurotoxic effects.9 Doppler ultrasonography was first used in 1965 for From the aDepartment of Neurology and bCardiology, University of the Saarland, and the cDepartment of Neurology, Ludwig-Maximilians-University. This paper funded by an institutional grant from the Department of Neurology and Cardiology, University of the Saarland, Germany. Submitted February 3, 1998; accepted May 1, 1998. Reprint requests: Gerhard F. Hamann, MD, PD, Department of Neurology, LudwigMaximilians-University, Marchioninistr. 15, D-81377 Munich, Germany. E-mail: [email protected] Copyright © 1999 by Mosby, Inc. 0002-8703/99/$8.00 + 0 4/1/91404
the detection of air bubbles in extracorporeal circuits.10 More recently, in vitro and animal studies were performed showing that Doppler emboli detection of atheroma, fat, platelet, and thrombus is also possible.11,12 Detection of emboli is based on the different acoustic impedances of air, platelet aggregates, or atheroma in respect to the blood. In the case of the passage of an emboli, this causes a sudden higher rate of reflection of the ultrasound beam and, therefore, an intensity increase of the Doppler signal, called microemboli signals (MES). Using transcranial Doppler ultrasonography (TCD), Spencer et al13 first reported nongaseous MES in 1990 in patients undergoing carotid endarterectomy. Since then, there have been a great number of studies demonstrating MES in carotid artery stenosis,14 atrial fibrillation,15 or prosthetic heart valves16,17 and also during cardiac bypass surgery.18,19 In angiographic techniques, MES have been reported in carotid and cerebral angiograms.5,6 The aim of this study was to perform continuous TCD monitoring during LHC to detect possible MES, and further to show the time of occurrence in respect to different catheter manipulations or possible accidental injections of air
American Heart Journal Volume 137, Number 1
bubbles and to prove the relation between MES and neurologic deficits.
Patients and methods One hundred one consecutive patients (70 men, 31 women aged from 24 to 83; median 63 years) undergoing LHC were studied in the cardiac catheterization laboratories of the University of Saarland in Homburg, Germany. Informed consent was obtained from all patients. The study was performed in accordance with the Helsinki declaration and after approval of an institutional committee on human clinical research. A 2-MHz TCD device (Multi Dop X, Fa. DWL Systems) with 2 separate channels was used for bilateral transcranial monitoring of the middle cerebral artery (MCA). The Doppler spectra were analyzed by 64-point fast-Fourier transform with color coded display. The emboli signals were recorded by an integrated emboli detection program (MDX TCD-7 emboli detection software). The detection threshold was put at 14 dB. At this level the random Doppler speckle did not trigger the emboli detection program and all audible and visible MES were detected. No additional tape recording was used. The ultrasound transducers were fixed at the temporal bone with a special metallic head band 10 to 15 minutes before the initiation of LHC. The depth of insonation of MCA was 50 to 55 mm, sample volume was 15 mm, and power and gain were adjusted to the lowest level to obtain blood flow spectra of low intensity. This allowed optimization of the dynamic range of the spectral analyzer for the detection of the high-intensity emboli signals. During the monitoring the examiner was able to identify emboli signals (1) visually on the monitor display with the Doppler spectra of both MCA and (2) by listening to the acoustic changes of the Doppler signal. At the same time, each MES was automatically stored by the integrated emboli detection software. The Doppler spectra around the detected event were displayed in the lower part of the monitor indicating exact time, side of occurrence, and signal intensity in steps of 4 dB. When the buffer of the monitor display was full, the spectra were written automatically to disk. Furthermore the examiner wrote down each registered emboli signal in a table, again indicating exact time, visual and acoustic quality, and current action of the catheter operator and patient. The monitoring ended immediately after LHC had finished. Directly before and after the catheterization a short neurologic exploration was performed including questions in respect to neuropsychologic functions (time, person, and place). During off-line analysis, the MES recorded by the automatic emboli detection program and the manual notations of the experienced examiner were compared. MES were selected by following the standard criteria of Spencer et al.13
Fischer et al 163
Artefacts from patient speech, coughing, or movements were excluded. Afterward, the MES were summed up in groups in respect to different catheterization maneuvers: (1) injection of contrast media or saline solution, (2) movements of the catheter with the wire inside, either both same direction or against each other, (3) maneuvers with the catheter in the aorta or left ventricle, or (4) signal without relation to any actual maneuver or injection. To analyze emboli signal amplitude, 250 emboli signals from maneuvers 1, 2, and 3 above were studied using the relative intensity increase in decibels calculated by the automatic emboli detection program. The determination of the MES rate during diagnostic LHC was performed in a subgroup of 52 patients who underwent a standardized diagnostic LHC (1 ventriculography, 1 angiogram of the right, and 3 to 4 angiograms of the left coronary artery) and who presented no other possible cerebral emboli risk factors. In 3 patients all injection fluids underwent short-term hypobaric pressure to reduce their gas content. Following the rule that a hypobaric pressure decreases the concentration of dissolved gases, we tapped the top of the injection syringe with the finger and applied a negative pressure suction of 5 seconds by pulling the syringe. Afterwards the gas fraction was eliminated. The emboli signal rate of this subgroup was compared with the patients with unprepared injection fluids. For statistical elaboration, the Mann-Whitney U test was used. The LHC was performed with the transfemoral technique (Judkins); only 1 patient was examined by the transbrachial approach (Sones). All patients received a sheath for introducing the catheter. Preshaped 5F or 6F polyurethane catheters (Cordis Corporation) and coated steel wires (Corotec Medizintechnik GmbH) were used. LHC consisted of ventriculography (40 mL contrast injected at 14 mL/sec by a power injector using a pigtail catheter). Afterwards the catheter was changed and 1 angiogram of the right and 3 to 4 angiograms of the left coronary artery were performed, all biplane. In coronary angiography, 8 to 9 mL contrast was quickly injected manually. Additionally, 12 patients underwent aortic root angiography (40 mL contrast, 18 mL/sec) and 15 patients had bypass angiography. Twentynine patients had coronary interventions. Guide catheters (7F, 9F, or 10F) were therefore used that rested at the coronary artery orifice while the different interventional catheters were introduced. Twenty-five patients had balloon dilation, 2 underwent directional atherectomy, and 2 had high-speed rotablation. During rotablation no vasodilator was injected, but in between the cutting sequences this injection was performed. During the cutting there was only the injection of saline solution for cooling of the rotor. The following contrast media were used: iopamidol (Solutrast 370, Byk Gulden) in 34 patients, iopromid (Ultravist 370, Schering) in 64 patients, and ioxagline acid (Hexabrix 320,
American Heart Journal January 1999
164 Fischer et al
Figure 1
Figure 2
Comparison of MES amplitude of 250 MES during injections of fluids, movements of catheter and wire, and coiling maneuvers with catheter. Doppler flow spectra of the MCA with typical microemboli signal (arrow) detected during coiling maneuver with cardiac catheter in aortic arch. Left scale: flow velocity (cm/s); right scale: Doppler signal intensity (dB), color coded.
Laboratoire Querbet) in 3 patients. The use of heparin (2500 IU intra-arterial, n = 76) or aspirin (100 to 500 mg orally, n = 69) was documented.
Results In all patients, cerebral MES were detected. During puncture of the femoral artery, passage of the abdominal aorta, or interruptions of the exploration up to 20 minutes, no MES were observed. Once the catheter was advanced beyond the aortic arch, MES were recorded. MES were related to coiling maneuvers with the cardiac catheter in the aortic root, aortic valve, or left ventricle to movements between catheter and guide wire and to injections of contrast media and saline solution. The MES were short (<0.1 second) unidirectional signals that occurred randomly in the cardiac cycle. They presented a relative intensity increase calculated by the emboli detection software of between 16 to 32 dB. They were accompanied by a typical acoustic sound (cracking or chirping). Fig 1 shows a typical cerebral MES occurring during a coiling maneuver with the catheter in the aortic arch. Receiver overload with bidirectional widespread Doppler signals was seen, especially after injection of
great volumes of contrast media or saline solution. Artefacts from head movement or speech also produced bidirectional Doppler signals, but were of longer duration and lower audio frequencies. The analysis of the relative intensity increase of the MES in respect to their association to injections of contrast, movements of wire, and catheter or coiling maneuvers with the catheter is shown in Fig 2. There were no significant differences of the relative intensity increase of the MES between these groups. A standardized diagnostic LHC showed a mean rate of 95 ± 45 MES. Nearly half of them occurred during ventriculography (45.5 ± 28 MES) and half of them during coronary angiography (49.5 ± 36 MES). Most of the MES (67.5%) were due to injections of contrast media and saline solution; 30% were related to parallel movements of the wire and the catheter. Only 2% of MES happened during coiling maneuvers with the catheter in the aorta or left ventricle (Fig 3). In the subgroup of patients who received prepared contrast media and saline solution with reduced gas content, we saw a significant reduction of 67% of the MES rate (P < .05). Only the MES related to injections of contrast or saline solution (65 versus 25 MES, P < .05) and parallel movements of catheter and wire (29 versus 3 MES, P < .01) were significantly decreased (see Fig 3). The different catheters or contrast media did not influence the MES rate significantly. Aortography (n = 12) showed 37 ± 18 MES and angiography of bypass (n = 15) up to 691 MES (mean
American Heart Journal Volume 137, Number 1
Fischer et al 165
Figure 3
Distribution of MES during standardized LHC (n = 55) and influence of injection of contrast media and saline solution with low gas content on mean MES rate.
90 MES), nearly all of them related to injections of contrast media. In angiography of the left internal mammarian artery, bypass MES were detected only in the left MCA. During coronary interventions, MES caused by coiling maneuvers with the thicker and less flexible guide catheter in the aortic root were not more frequent than during diagnostic LHC (0 to 7 MES, mean 1 MES per patient). Ninety percent of MES were due to contrast injections and flushing of the catheter with saline solution. During the interventional procedures themselves we observed only 2 MES during 143 balloon insufflations in 25 patients, during directional arterectomy (n = 2) only 1 MES, but during rotablation (n = 2) the first patient presented 684 MES (total time 150 seconds) and the second 238 MES (total time 60 seconds). There were more MES when proximal parts of coronary artery underwent rotablation. The MES rates in coronary interventions are shown in Fig 4. MES were only detected during the cutting sequences of the rotablation. In patients with sclerosis or stenosis of the aortic valve (n = 11), we found significantly higher rates of MES (8 MES versus 1.5 MES, P < .01) during coiling maneuvers with the catheter at the valve. The presence of aneurysm or thrombus of the left ventricle (n
= 7), atrial fibrillation (n = 6), prosthetic heart valve (n = 1), and carotid artery stenosis (n = 9) did not show any influence on the MES rate during maneuvers with the catheter inside the left ventricle. The use of aspirin or heparin did not result in significant changes in the detection rate of MES. None of the patients included in this study developed any clinical signs of cerebral ischemia or neuropsychologic deterioration. No major changes in the neurologic examination before and after the left heart catheterization were noticed.
Discussion This study demonstrates that a large number of Doppler MES in both MCAs is detectible during left heart catheterization. The MES presented the same characteristics that have been demonstrated in other clinical and in vitro Doppler emboli studies.11-16 Turbulences20 and artefacts from speech and movement21 are completely different. The MES had a minimum relative intensity increase of 16 dB. This is a relatively high threshold in comparison to other studies. But we used the relative intensity increase calculated by the automatic emboli detection software, which compares the signal intensity of the whole screen including areas free of Doppler spectra with the emboli signal. This
166 Fischer et al
American Heart Journal January 1999
Figure 4
Cerebral microemboli signals during coronary interventions.
leads to higher decibel values of MES intensity. Kaps et al17 reported similar experiences using the same emboli detection software. The determination of the underlying embolus material in our study is difficult because air bubbles might be detected as well as solid emboli. The accidental introduction of air when the contrast is drawn up and immediately injected is possible and has been described in cerebral and carotid angiography.5,6 In our study 70% of the MES were associated directly to the injection of contrast media and saline solution, so that there is great evidence that this part of the MES represents air emboli. The analysis of the relative intensity increase of each MES did not help in our study to identify gas emboli. With the use of in vitro and animal models it has been shown that air bubbles produce more intensive MES than solid emboli because of the greater difference of acoustic impedance between air and blood.11,12 But it was also demonstrated that the size of the emboli influence the intensity of its MES. In our study with unknown size and quality of the emboli, we did not know whether a signal of low intensity was due to a very small air bubble or a bigger solid embolus. The results of Markus et al5 in respect to cerebral angiography (consistent find-
ings of very high intensity increases of MES with a usual receiver overload) could not be confirmed in this study. Receiver overload was seen especially at the beginning of injections of great volumes of contrast media. These differences may be explained by a dispersion of the injected gas bubbles during their way from the aortic arch up to the MCA in comparison with the air emboli injected directly in the carotid artery in cerebral angiography. The introduction of gas bubbles into the cerebral arterial circulation during LHC was further supported by our finding that patients who received injections of fluids containing reduced gas presented a significant decrease in the MES rate of nearly 70%. Markus et al5 saw lower MES rates when the fluids were allowed to stand a few minutes before an injection. An important finding of our study was that not only the MES that occurred during the injections but also those occurring during the movements of wire and catheter were reduced. This suggests that multiple air bubbles are resting at the surface of the catheter and wire and they are mobilized afterwards by movements of the wire. The deposition of platelet aggregates at the catheter surfaces4 and its embolism is also possible, but may be reflected only in a small part of the MES.
American Heart Journal Volume 137, Number 1
Nevertheless, solid microemboli also seem to occur during LHC, especially in respect to coiling maneuvers with the catheter at the aortic arch or at the aortic valve. During these maneuvers the MES rate in patients with and without minor gas content of injection fluids was constant. Patients with sclerosis or stenosis of the aortic valve presented a significantly higher embolism rate during these maneuvers. The presence of hundreds of cerebral MES during coronary rotablation was an interesting finding of this study. One possible explanation is the continuous injection of saline solution for cooling during the rotablation, which might result in bubble transfer to the aorta. Another mechanism may be the backstream of emboli due to high local pressure formed by the high-speed (180,000 revolutions/min) rotablator. MES may reflect cavitation bubbles as well as small atheromatous particles. The clinical importance of the detected MES remains unclear. None of the patients included in this study developed cerebral ischemic symptoms, therefore the clinical results of the detected MES seem to be benign. There is only 1 case report of multiple MES caused by a compression of common carotid artery with immediate development of contralateral paresis,22 and the majority of MES studies did not show obvious clinical sequelae.13-17 One explanation of this lack of major clinical symptoms may be the size of microemboli. In animal models it was not possible to cut solid microemboli smaller than 300 to 500 µm, but emboli of this size were detected easily.11,12 In the case of use of glass microspheres, even the detection of particles of 5µm was shown.12 Consequently microemboli potentially may produce only very small cortical infarctions or, in the case of particles smaller than 10 µm, even the passage of the capillary bed is possible. Furthermore, gas emboli are dissolvable and may disappear on their way to the microcirculation. But what is the clinical impact of cerebral microinfarction? Changes of neuropsychologic function (over 50%) have been reported in patients with coronary bypass surgery and high rates of cerebral microemboli signals during the cardiopulmonary bypass.18,19 These MES have been related to gas bubbles introduced by the extracorporeal oxygenators. Neuropathologic studies in patients after bypass surgery showed great numbers of dilatations in cerebral arterioles and capillaries as possible indicators of persisting gaseous microemboli.23 The use of arterial filters in bypass surgery led to a reduction of
Fischer et al 167
the MES rate and neuropsychologic deficits.18 Therefore, in our opinion cerebral microembolism, caused by microbubbles or solid emboli, cannot be generally thought to be clinically harmless. During LHC a neuropsychologic impairment caused by hundreds of microemboli may be possible, especially when numerous consecutive LHC or coronary interventions like rotablation are carried out. Further studies have to consider whether there exists a correlation between high rates of MES and neuropsychologic dysfunction during left heart catheterization. We thank the team of cardiac catheterization laboratories, Department of Cardiology, University of Saarland. We also thank Mrs Seiche for her language editing of this manuscript.
References 1. Adams DF, Abrams HL. Complications of coronary arteriography. Cardiovasc Radiol 1979;2:89-96. 2. De Bono D. Complications of diagnostic cardiac catherization: results from 34,041 patients in the United Kingdom. Br Heart J 1993;70:297-300. 3. Bourassa MG, Noble J. Complication rate of coronary arteriography. Circulation 1976;53:106-14. 4. Takaro T, Hultgen HN, Littmann D, Whright EC, Oteen NC. An analysis of deaths occurring in association with coronary arteriography. Am Heart J 1973;86:587-97. 5. Markus HS, Loh A, Israel D, Buckenham T, Clifton A, Brown MM. Microscopic air embolism during cerebral angiography and strategies for its avoidance. Lancet 1993;341:784-7. 6. Gerraty RP, Bowser DN, Infeld B, Mitchell PJ, Davis SM. Microemboli during carotid angiography, association with stroke risk factors or subsequent MRI imaging changes? Stroke 1996;27:1543-7. 7. Sage MR, Drayer B, Dubois P, Heinz E, Osborne D. Increased permeability of the blood brain barrier after carotid Renografin-76. AJNR Am J Neuroradiol 1981;2:272-4. 8. Laerum F. Injurious effects of contrast media on vascular endothelium. Invest Radiol 1985;20;598-9. 9. Drayer BP, Velay R, Bird R, Albright R, Roberts L, Allen S, et al. Comparative safety of intracarotid iopamidol, iothalamate meglumine, diatroate meglume for cerebral angiography. Invest Radiol 1984; 19:212-8. 10. Austen WG, Howry DH. Ultrasound as a method to detect bubbles or particulate matter in the arterial line during cardiopulmonary bypass. J Surg Res 1965;5:283-4. 11. Russell D, Madden KP, Clark WM, Sandset PM, Zivin JA. Detection of arterial emboli using Doppler ultrasound in rabbits. Stroke 1991; 22:253-8. 12. Markus HS, Loh A, Brown MM. Detection of circulating cerebral emboli using Doppler ultrasound in a sheep model. J Neurol Sci 1994;122:117-24. 13. Spencer MP, Thomas GI, Nicholls SC, Sauvage LR. Detection of middle cerebral artery emboli during carotid endarteretomy using transcranial Doppler ultrasonography. Stroke 1990;21:415-23. 14. Siebler M, Sitzer M, Steinmetz H. Detection of intracranial emboli in
168 Fischer et al
15.
16.
17.
18.
patients with symptomatic extracranial carotic artery disease. Stroke 1992;23:1652-4. Tegeler CH, Hitchings LP, Eicke M, Leighton J, Fredericks RK, Downes TR, et al. Microemboli detection in stroke associated with atrial fibrillation. J Cardiovasc Technol 1990;9:283-4. Georgiadis D, Mallison A, Grosset DG, Lees KR. Coagulation activity and emboli counts in patients with prosthetic cardiac valves. Stroke 1994;25:1211-4. Kaps M, Hansen J, Weiher M, Tiffert K, Kayser I, Droste DW. Clinically silent microemboli in patients with artificial prosthetic aortic valves are predominantly gaseous and not solid. Stroke 1997;28:322-5. Pugsley W, Klinger L, Paschalis C, Treasure T, Harrison M, Newman S. The impact of microemboli during cardiopulmonary bypass on neuropychological functioning. Stroke 1994;25:1393-8.
Fischer et al
19. Clark RE, Brillman J, Davis DA, Lowell MR, Price TRP, Magovern GJ. Microemboli during coronary artery bypass grafting, genesis and effect on outcome. J Thorac Cardiovasc Surg 1995;2:249-58. 20. Spencer MP, Whisler D. Transorbital Doppler diagnosis of intracranial arterial stenosis. Stroke 1986;17:916-21. 21. Spencer MP. Detection of cerebral arterial emboli. In: Newell DW, Aaslid R, editors. Transcranial Doppler. New York: Raven Press; 1992. p. 215-30. 22. Khaffaf N, Karnik R, Winkler WB, Valetin A, Slany J. Embolic stroke by compression maneuver during transcranial Doppler sonography. Stroke 1994;25:1056-7. 23. Moody DM, Bell MA, Challa VR, Johnston WE, Prough DS. Brain microemboli during cardiac surgery or aortography. Ann Neurol 1990;28:477-86.