Original Paper Received: September 17, 1999 Accepted: February 17, 2000 Cerebrovasc Dis 2000;10:403–408 Cerebral Microembolus Detection in an Unselected Acute Ischemic Stroke Population Christian Lund a Jørgen Rygh b Brynhild Stensrød b Per Morten Sandset b Rainer Brucher a David Russell a a Department of Neurology, The National Hospital, and b Department of Medicine, Ullevaal University Hospital, University of Oslo, Norway Abstract Objective: The aims of this study were firstly to determine prevalence, frequency, and clinical significance of cerebral microemboli in an unselected acute ischemic stroke population and secondly to examine how this information may improve ischemic stroke subtype classification. Methods: We intended to perform transcranial Doppler (TCD) microembolus monitorings of the middle cerebral artery (MCA) in the symptomatic hemisphere for 45 min in 120 consecutive patients with internal carotid artery territory ischemia. The first examination was performed within 72 h from start of symptoms and the second 5 B 1 days later. Platelet and coagulation system activation were measured following TCD monitoring in 38 patients. The strokes were subtyped using the TOAST classification criteria, and the patients’ clinical status was assessed at discharge using the Scandinavian Stroke Scale and the Barthel Index. Results: Microembolus monitoring was technically possible in 83 (69.2%) of the ABC © 2000 S. Karger AG, Basel 1015–9770/00/0105–0403$17.50/0 Fax + 41 61 306 12 34 E-Mail firstname.lastname@example.org www.karger.com Accessible online at: www.karger.com/journals/ced 120 patients. Thirty-two (26.6%) patients had an inadequate temporal bone acoustic window or were too restless to allow long-time monitoring. In 5 (4.2%) patients the relevant MCA was occluded. Twenty-two (26.5%) of the 83 patients had microemboli despite the fact that over 90% were receiving an antiplatelet or an anticoagulant treatment. The mean frequency of microemboli was 6.7 B 13.6 per 45 min. Microemboli were more prevalent in assumed cardioembolic stroke than in other subtypes of ischemic stroke (p = 0.047). We found no association between the presence of cerebral microemboli and the clinical outcome or the parameters for platelet or coagulation system activation. The presence of microemboli was not associated with in-hospital deaths (p = 0.17), whereas MCA occlusion was (p = 0.01). Conclusions: Cerebral microemboli are frequent in unselected acute ischemic stroke patients despite antiplatelet or anticoagulant treatment. TCD detection of microemboli provides valuable pathophysiological information and may, therefore, improve current ischemic stroke subtype classification. Copyright © 2000 S. Karger AG, Basel Christian Lund, MD Department of Neurology, The National Hospital University of Oslo N–0027 Oslo (Norway) Tel. +47 22868910, Fax +47 22868890, E-Mail email@example.com Downloaded by: Kings's College London 188.8.131.52 - 10/25/2017 6:53:36 PM Key Words Cerebral ischemia W Embolism W Skull W Ultrasonography W Transcranial Doppler W Stroke classification Many clinical trials have been carried out to assess the efficacy of different treatment regimens in acute ischemic stroke. The results of these trials have generally been disappointing  which may at least partly be due to the fact that ischemic stroke is not a single disease entity. Although the clinical findings are similar, the pathogenesis may be very different . Strokes frequently recur, and the second or third stroke may have an etiology different from the first . It is hoped, therefore, that ischemic stroke classification criteria may be improved by information gained from neurosonological and brain imaging studies [2, 4–6]. Experimental and clinical studies have shown that Doppler ultrasound may be used to detect arterial emboli [7, 8]. The demonstration of cerebral microemboli using transcranial Doppler (TCD) in acute cerebral ischemia provides evidence that embolic material is entering the brain which indirectly suggests that the patient’s symptoms may also be due to brain embolism. The aims of this study were firstly to determine prevalence, frequency, and clinical significance of cerebral microemboli detected in unselected consecutive patients admitted to a stroke unit following acute cerebral ischemia in the carotid territory and secondly to examine how this information may improve ischemic stroke subtype classification. Patients and Methods Study Population A total of 120 consecutive patients, 66 women and 54 men with a mean age of 74 B 10 years were included in the study. All had experienced symptoms of acute cerebral ischemia in the vascular territory of a carotid artery. The only exclusion criteria was hospital admission 1 72 h after start of symptoms. In Norway, most stroke patients are served by the stroke unit of their regional hospital. The stroke unit at Ullevaal University Hospital, Oslo, admits stroke patients from a catchment area with 193,200 inhabitants (January 1, 1997). This population is comparable with the general Norwegian population with regard to age and sex distribution . The diagnostic and therapeutic procedures followed were in accordance with the stroke unit’s normal guidelines, and the patients or their relatives gave informed consent. Investigations A complete clinical neurological examination, a cerebral CT examination, and a Doppler ultrasound examination of the extracranial carotid arteries were carried out in all patients on admission. The treating physicians decided whether additional neuroradiological examinations or echocardiography should be performed according to the usual routine of the stroke unit. The number of in-hospital days and the number of in-hospital deaths were recorded. 404 Cerebrovasc Dis 2000;10:403–408 Neurological impairment and functional outcome at hospital discharge were assessed using the Scandinavian Stroke Scale (SSS) and the Barthel Index (BI), respectively. The TOAST classification system was used to subtype the patients for stroke etiology [10, 11]. This system has five subtypes of ischemic stroke: (1) cardioembolism; (2) large-artery atherosclerosis; (3) small-vessel occlusion; (4) stroke of other determined etiology, and (5) stroke of undetermined etiology. All SSS and BI ratings, as well as the TOAST subtyping, were performed by physicians who were blinded to knowledge regarding the TCD findings. When performing the TOAST classification, historical patient data as well as findings from clinical and supplementary investigations were used. Hematological and Biochemical Examinations Blood tests (hemoglobin, platelet count, C-reactive protein, glucose, creatinine, total cholesterol, and triglycerides) were carried out on admission in all patients. The level of activated coagulation (prothrombin fragment 1+2) and platelet activation (ß-thromboglobulin) were measured within 12 h after the TCD examination in 38 cases. Blood for the prothrombin fragment 1+2 analysis was collected in 0.1 vol of 129 mM citrate. The ß-thromboglobulin analysis was collected in 0.1 vol of CTAD anticoagulant (109 mM citrate, 15 mM theophylline, 3.7 mM adenosine, 0.2 mM dipyridamole). Vacutainer tubes (Becton Dickinson, Combourg, France) were used. The blood samples were centrifuged (2,000 g, 20 min), and plasma was pipetted off and stored in aliquots at –80 ° C until analysis. Prothrombin fragment 1+2 was analyzed using a commercial kit (Behringwerke, Marburg, Germany). ß-Thromboglobulin was analyzed using a kit from Stago (Asnières, France). TCD Recording The initial TCD examination was performed within 72 h after start of symptoms. A second TCD examination was carried out 5 B 1 days later if the patient was still in hospital. Every examination was made at the bedside in the Stroke Unit by the same observer (C.L.) using a Pioneer TC 4040 (Nicolet-EME, Überlingen, Germany) equipped with a 2-MHz monitoring probe. The middle cerebral artery (MCA) in the symptomatic hemisphere was insonated through the transtemporal acoustic window with two gates. A stable probe position was maintained using a fixation device. Both axial sample volumes were 10 mm in length, the spatial distance between the gates was 6 mm, and the insonation depths were 55.9 B 2.1 and 49.9 B 2.3 mm, respectively. The fast Fourier transform resolution used was 128 points, the fast Fourier transform-time window overlap was held constant at 67%, the high-pass filter setting was 120 Hz, and the sweep speed on screen was 3.1 s. The monitoring time was 45 min. The Doppler audio signals were recorded on-line onto a digital audiotape (Tascam DA 38, Tokyo, Japan). When performing the offline analysis, the Doppler audio signals were reintroduced into the Doppler system using the same instrumentation settings as when the recordings were performed. The identification of microemboli was made off-line by an experienced observer (D.R.) who was blinded to clinical information. International consensus criteria [12, 13], which were updated by our own group, were applied for microembolus identification; a relative peak intensity increase of 64 dB from the background signal, a random appearance in the cardiac cycle, a short duration (! 0.2 s), unilateral movement in the Doppler spectra, and typical sound. In addition, the microembolic signals had to appear with a time lag 1 3 ms in the two sample volumes. Lund/Rygh/Stensrød/Sandset/Brucher/ Russell Downloaded by: Kings's College London 184.108.40.206 - 10/25/2017 6:53:36 PM Introduction Statistics We considered a patient to be microembolus-positive, if any cerebral microemboli were detected on one or both examinations. Comparison of group proportions for categorical data within fourfold tables were calculated with Fisher’s exact test. Continuous nonparametric distributed variables were compared using the Mann-Whitney test and continuous parametric data using the Student t test. The level of statistical significance was p ^ 0.05. The statistical analyses were performed with the StatView (version 4.5) statistical program. The data are presented as mean values B SD. Results TCD Monitoring Cerebral Microembolus Detection Fig. 1. Eighty-three patients examined for cerebral microemboli assessed according to the TOAST classification system. The dark columns represent the patients without microemboli; the dotted columns represent the patients with microemboli. C = Cardioembolic stroke; L = stroke due to large-artery atherosclerosis; S = stroke due to small-vessel occlusion; U = stroke of undetermined etiology; TIA = transient ischemic attack. Incidence and Frequency of Microemboli Microemboli were detected in a total of 22 (26.5%) of the 83 patients: in 14 (16.9%) of the 83 patients at the first examination and in 14 (22.2%) of the 63 patients at the second examination. Eighteen microembolus-positive patients had two examinations: 6 patients showed microemboli at both examinations and 12 patients at only one examination. When microemboli were detected, their frequency was 6.7 B 13.6 (range 1–67) per 45 min. Stroke Subtypes The 83 patients, in whom TCD microembolus recordings were sucessfully performed, were evaluated for stroke etiology. Seventy-three (88.0%) of the 83 patients suffered an ischemic stroke, and 10 (12.0%) patients had a transient ischemic attack (TIA). When the 73 stroke patients were assessed according to the TOAST classification, we found that 19 (26.0%) had a cardioembolic stroke, 7 (9.6%) a large-artery atherosclerotic stroke, 17 (23.3%) stroke due to small-vessel occlusion, and 30 (41.1%) stroke of undetermined etiology. Ten (52.6%) patients with cardioembolic stroke, 1 (14.3%) with large-artery atherosclerotic stroke, 3 (17.6%) with small-vessel occlusion, and 5 (16.7%) with stroke of undetermined etiology had microemboli (fig. 1). Cerebral Cerebrovasc Dis 2000;10:403–408 405 Downloaded by: Kings's College London 220.127.116.11 - 10/25/2017 6:53:36 PM TCD microembolus monitoring was technically possible in 83 (69.2%) of the 120 patients, 39 women and 44 men with an age of 73 B 11 years. Five (4.2%) patients had a suitable temporal bone window, but an extensive bilateral TCD examination revealed an occluded symptomatic MCA main stem. Twenty-eight (23.3%) patients had an insufficient temporal bone window which resulted in a signal-to-noise ratio which was insufficient for satisfactory long-time monitoring. This was significantly correlated with female sex (p ! 0.001) and high age (p = 0.007). TCD monitoring was impossible in an additional 4 (3.3%) patients because they were too restless. It was possible to reexamine 63 of the 83 initially examined patients which gave a total number of 146 TCD recordings. The first monitoring took place 42.8 B 18.3 h and the second 149.7 B 35.1 h following symptom start. During admission, 37 (44.6%) of the 83 patients had an echocardiographic examination. In 31 cases this examination was normal or revealed only minor structural cardiac abnormalities. In 6 cases, however, major structural abnormalities (possible cardiac embolic sources) were found. Twenty-two (26.5%) patients had more than one cerebral CT or cerebral MRI examination. An internal carotid artery stenosis (150%) or an internal carotid artery occlusion was found in 11 (13.3%) of the 83 patients. All of these but 1 were located ipsilateral to the symptoms. MCA stenosis (150%) was detected in the symptomatic hemisphere in 2 (2.4%) patients. At the initial TCD examination 65 (78.3%) of the 83 sucessfully examined patients were receiving an antiplatelet medication: 63 patients acetylsalicylic acid and 2 patients ticlopidine. Twelve (14.5%) patients were anticoagulated: 10 with warfarin and 2 with heparin. Two (2.4%) patients had a combined treatment with both acetylsalicylic acid and warfarin. Only 6 (7.2%) patients had neither antiplatelet nor anticoagulant treatment. Table 1. Clinical outcome parameters for 67 surviving stroke patients examined for cerebral microemboli In-hospital days SSS score at discharge BI score at discharge Patients with microemboli (n = 16) Patients without microemboli (n = 51) p 31.8B37.1 39.7B13.7 16.6B6.3 26.4B26.9 40.3B10.6 17.0B5.5 0.53 0.84 0.84 The maximum (best) scores of SSS and BI are 48 and 20 points, respectively. Table 2. Biochemical parameters for coagulation system activation (prothrombin fragment 1+2) and platelet activation (ß-thromboglobulin) in 38 stroke patients (2 with TIA) monitored for cerebral microemboli Prothrombin fragment 1+2, nmol/l ß-Thromboglobulin, ng/ml Patients with microemboli (n = 13) Patients without microemboli (n = 25) p 1.9B1.0 42.8B24.5 1.5B0.7 39.6B18.0 0.26 0.94 The normal range of prothrombin fragment 1+2 is 0.4–1.1 nmol/l, that of ß-thromboglobulin 10–40 ng/ml. Outcome Six (8.2%) of the 73 stroke patients, who had TCD microembolus monitoring, died in hospital. Three (50%) of these had cerebral microemboli which was not significant (p = 0.17) with regard to in-hospital deaths. MCA occlusion was, however, significantly (p = 0.01) associated with in-hospital death. The 67 surviving stroke patients who had TCD microembolus monitoring studies were grouped as to whether they had microemboli or not. There was no significant difference in these two groups with regard to the total number of in-hospital days, the SSS score, or the BI score at discharge (table 1). Hematological and Biochemical Parameters Assessment of the 83 patients who had TCD monitoring showed no correlation between the presence of microemboli and the following biochemical parameters: hemoglobin, platelet count, C-reactive protein, glucose, creatinine, total cholesterol, and triglycerides. We also found no significant association between the presence of mi- 406 Cerebrovasc Dis 2000;10:403–408 croemboli and the parameters for coagulation system and platelet activation (table 2). However, the values for prothrombin fragment 1+2 (1.7 B 0.8 nmol/l, normal range 0.4–1.1 nmol/l, n = 38) and ß-thromboglobulin (41.2 B 21.0 ng/ml, normal range 10–40 ng/ml, n = 38) were both above the normal range which suggests both activation of the coagulation system and platelet activation. Anticoagulated patients had a significantly lower value for prothrombin fragment 1+2 as compared with those treated with antiplatelets alone (0.9 B 0.7 vs. 1.8 B 0.8 nmol/l, p = 0.03, n = 38). Discussion In this study we found that more than 1 of 4 patients admitted to a stroke unit due to acute anterior cerebral ischemia had cerebral microemboli during the acute phase. In other comparable acute TCD studies, the proportion of patients with microemboli during the acute phase after stroke has shown considerable variation, ranging from 9.3 to 70.7% [14–19]. This large variation may have several explanations. The patient populations were not similar in the different studies with regard to age, sex, and stroke type. In some studies, for example, the mean age of the patients were 63 years or less which strongly Lund/Rygh/Stensrød/Sandset/Brucher/ Russell Downloaded by: Kings's College London 18.104.22.168 - 10/25/2017 6:53:36 PM microemboli were significantly more prevalent (p = 0.047) in cardioembolic stroke than in the other stroke subtypes. Three (30.0%) of the 10 TIA patients had microemboli. suggests a certain degree of selection with regard to age [16, 17, 19]. The patient population in our study had a mean age of 74 years and represented all patients admitted to a regional stroke unit. The fact that microembolus detection using TCD is a relatively new method entails a development with regard to more exact identification criteria for microemboli. These have improved with experience during the last years which may make the value of reported results dependent to some degree on when the study was carried out. Results from different studies may be easier to compare in the future, since we now have internationally accepted criteria [12, 13]. The reported incidence and prevalence of cerebral microemboli following acute ischemic stroke also depend on the duration of the TCD monitoring, the number of examinations carried out, the time interval from symptoms to examination, and on whether the examinations were performed uni- or bilateral. Ideally, TCD monitoring should be carried out as long as possible during the first days after admission. We found, however, in a pilot study that it was practically very difficult in many patients to continue monitoring for longer than 45 min due to restlessness and lack of cooperation. An additional problem, especially in older female patients, is the absence of an adequate temporal bone acustic window [20, 21]. This results in a very poor signal-to-noise ratio which prevents TCD examination or excludes monitoring over longer periods. Serial TCD studies have shown that the incidence of cerebral microemboli is highest closer to symptoms, whereas the total prevalence in a study population may increase with repeated examinations [16–18]. However, in this study we found a slight increase in the incidence at the second examination which was probably due to the fact that the second examination did not include patients who had minor symptoms, since many of these had been discharged. The presence of cerebral microemboli following acute ischemic stroke provides evidence which suggests that the pathophysiological mechanism for the stroke may be embolic. Traditionally, the clinical diagnosis of brain embolism is made by the combination of clinical features and the detection of a potential embolic source. Evidence of cerebral microemboli may make classification of embolic stroke more exact . This could have important clinical consequences, since it may facilitate more specific therapeutic decisions. In this study strokes were classified using the TOAST classificiation system. Cerebral microemboli were detected in more than half of the patients in whom the TOAST classification was cardioembolic stroke which indirectly suggests that this classification was correct. Bilateral TCD monitoring will provide even more evidence suggesting cardioembolic etiology. Only 1 of the 7 patients with large-artery atherosclerosis (150% stenosis of a large extra- or intracranial artery) had microemboli, but the small number of patients in this group excludes conclusions. Cerebral microemboli were present in 17.6% of the patients with lacunar stroke. This suggests that microinfarctions may not always only be due to small-vessel occlusion following local lipohyalinosis or microatheroma [16, 19, 23]. The TOAST classification could not be determined in 30 (41.1%) of the stroke patients which underlines the difficulties with present-day classification systems, especially when these are applied to unselected stroke populations where there are relatively high numbers of elderly patients. Five of the patients in this group had microemboli which may indirectly suggest an embolic pathogenesis. It is, therefore, possible that TCD microembolus monitoring may in the future lead to a reduction in the number of undetermined cases . We did not find a correlation between the presence of microemboli and the clinical outcome assessed by the SSS or BI scores at discharge. In a similar study Del Sette et al.  found no association between the presence of microemboli during the acute phase and the 30-day outcome measured as disability using the Oxford Disability Scale. A TCD examination during the acute phase of an ischemic stroke will, however, allow the detection of a relevant MCA occlusion which does provide important prognostic information. It is of interest that cerebral microemboli were detected in 26.5% of the patients despite the fact that over 90% had antiplatelet or anticoagulation treatment or both. There was no significant correlation between the presence of microemboli and the markers for coagulation system and platelet activation in the subgroup of 38 patients studied. These activation measurements suggest, however, that both activation of the coagulation system and platelet activation were still present despite treatment and that the latter was either insufficient or inappropriate to prevent subclinical microembolus formation in many of the patients. Larger patient populations should now be studied to determine not only the prognostic significance of cerebral microemboli in acute stroke, but also the possibility that their detection may be of help in monitoring therapy. Cerebral Microembolus Detection Cerebrovasc Dis 2000;10:403–408 Downloaded by: Kings's College London 22.214.171.124 - 10/25/2017 6:53:36 PM 407 In conclusion, this study has shown that cerebral microemboli are frequent in unselected acute ischemic stroke patients despite antiplatelet or anticoagulation treatment. TCD detection of microemboli gives valuable pathophysiological information and may, therefore, improve current ischemic stroke subtype classification. Acknowledgement This study was supported by The Norwegian Council on Cardiovascular Diseases. References 408 10 Adams HP, Bendixen BH, Kappelle LJ, Biller J, Love BB, Gordon DL, Marsh EE and the TOAST investigators: Classification of subtype of acute ischemic stroke: Definitions for use in a multicenter clinical trial. Stroke 1993;24:35– 41. 11 D’Olhaberriague L, Litvan I, Mitsias P, Mansbach HH: A reappraisal of reliability and validity studies in stroke. Stroke 1996;27:2331– 2336. 12 Consensus Committee of the 9th International Cerebral Hemodynamic Symposium: Basic identification criteria of Doppler microembolic signals. Stroke 1995;26:1123. 13 Ringelstein EB, Droste DW, Babikian VL, Evans DH, Grosset DG, Kaps M, Markus HS, Russell D, Siebler M: Consensus on microembolus detection by TCD: International Consensus Group on Microembolus Detection. Stroke 1998;29:725–729. 14 Grosset DG, Georgiadis D, Abdullah I, Bone I, Lees KR: Doppler emboli signals vary according to stroke subtype. Stroke 1994;25:382– 384. 15 Tong DC, Albers GW: Transcranial Dopplerdetected microemboli in patients with acute stroke. Stroke 1995;26:1588–1592. 16 Daffertshofer M, Ries S, Schminke U, Hennerici M: High-intensity transient signals in patients with cerebral ischemia. Stroke 1996;27: 1844–1849. Cerebrovasc Dis 2000;10:403–408 17 Sliwka U, Lingnau A, Stohlmann W-D, Schmidt P, Mull M, Diehl RR, Noth J: Prevalence and time course of microembolic signals in patients with acute stroke: A prospective study. Stroke 1997;28:358–363. 18 Del Sette M, Angeli S, Stara I, Finocchi C, Gandolfo C: Microembolic signals with serial transcranial Doppler monitoring in acute focal ischemic deficit: A local phenomenon? Stroke 1997;28:1311–1313. 19 Koennecke H-C, Mast H, Trocio SH, Sacco RL, Ma W, Mohr JP, Thompson JLP: Frequency and determinants of microembolic signals on transcranial Doppler in unselected patients with acute carotid territory ischemia. Cerebrovasc Dis 1998;8:107–112. 20 Marinoni M, Ginanneschi A, Forleo P, Amaducci L: Technical limits in transcranial Doppler recording: Inadequate acoustic windows. Ultrasound Med Biol 1997;23:1275– 1277. 21 Eden A: Ultrasonic intensity and skull penetration in transcranial Doppler sonography; in Oka M, von Reutern G-M, Furuhata H, Kodaira K (eds). Recent Advances in Neurosonology. Amsterdam, Elsevier, 1992. 22 Caplan LR: Clinical diagnosis of brain embolism. Cerebrovasc Dis 1995;5:79–88. 23 Fisher CM: Lacunes: Small, deep cerebral infarcts. Neurology 1965;15:774–784. Lund/Rygh/Stensrød/Sandset/Brucher/ Russell Downloaded by: Kings's College London 126.96.36.199 - 10/25/2017 6:53:36 PM 1 Bousser M-G: Aspirin or heparin immediately after a stroke? Lancet 1997;349:1564–1565. 2 Caplan LR: Prevention of strokes and recurrent strokes. J Neurol Neurosurg Psychiatry 1998;64:716. 3 Yamamoto H, Bogousslavsky J: Mechanisms of second and further strokes. J Neurol Neurosurg Psychiatry 1998;64:771–776. 4 Hennerici MG, Schwartz A: Acute stroke subtypes – is there a need for reclassification? Cerebrovasc Dis 1998;8(suppl 2):17–22. 5 Lees KR: Multifactorial approach to stroke investigation and prevention. Lancet 1998;352: 923–924. 6 Futrell N: Pathophysiology of acute ischemic stroke: New concepts in cerebral embolism. Cerebrovasc Dis 1998;8(suppl 1):2–5. 7 Russell D, Madden KP, Clark WM, Sandset PM, Zivin JA: Detection of arterial emboli using Doppler ultrasound in rabbits. Stroke 1991;22:253–258. 8 Spencer MP, Thomas GI, Nicholls SC, Sauvage LR: Detection of middle cerebral artery emboli during carotid endarterectomy using transcranial Doppler ultrasonography. Stroke 1990;21: 415–423. 9 Oslo City, Department of Finance and Statistics, 1998.