HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) All Versions of this Article: 2273020530v1 228/1/126    most recent Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Kirchhof, K. Articles by Sartor, K. Search for Related Content PubMed PubMed Citation Articles by Kirchhof, K. Articles by Sartor, K.

Differentiation of White, Mixed, and Red Thrombi: Value of CT in Estimation of the Prognosis of Thrombolysis—Phantom Study¹

¹ From the Department of Neurology, Division of Neuroradiology (K.K., C.M., K.S.), Department of Radiation Oncology (T.W.), and Institute of Pathology, Division of Neuropathology (S.Z.), University of Heidelberg Medical Center, Im Neuenheimer Feld 400, 69120 Heidelberg, Germany. Received May 8, 2002; revision requested July 11; final revision received October 17; accepted October 28. Address correspondence to T.W. (e-mail: thomas_welzel@med.uni-heidelberg.de).

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
PURPOSE: To test with serial computed tomographic (CT) scans whether white, mixed, and red thrombi could be differentiated according to their CT attenuation.

MATERIALS AND METHODS: Platelet-enriched plasma and whole blood were mixed to produce samples with hematocrit levels of 0, 0.005, 0.03, 0.15, and 0.35. A thrombin solution was added, and after 2 hours the retracted clots were transferred into polyethylene tubes with a length of 4 cm and an inner diameter of 3 mm. Ten probes of each sample were placed into a plastic box filled with a solution of gelatin, gadopentetate dimeglumine, and distilled water. Ten tubes filled with gelatin served as control. With this phantom, the CT numbers of white, mixed, and red thrombi were measured over 144 hours. CT was performed with a multisection scanner and a collimation of 0.5 mm. Statistical analyses were performed for differences between groups and over time.

RESULTS: The CT numbers of white, mixed, and red thrombi differed significantly (P < .05) for most time measurements, except for white and mixed thrombi, which had a low hematocrit level at 24 and 144 hours (P > .05).

CONCLUSION: With CT it appears feasible to differentiate thrombi according to their hematocrit level and thus estimate the effectiveness of thrombolysis.

© RSNA, 2003

Index terms: Arteries, thrombosis, 17.769 • Brain, infarction, 17.781 • Phantoms • Thrombolysis, 17.1265 • Thrombosis, CT, 17,12115, 17.12119 • Thrombosis, experimental studies, 17,12115, 17.12119, 17.1265

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Early vascular recanalization forms the basis for the successful treatment of focal cerebral ischemia. This treatment can be achieved by using systemic intravenous or local intraarterial thrombolysis and various techniques for mechanical recanalization. Although it is difficult to decide which of these approaches is best in each case, two findings might help to solve this problem: First, it is known from the results of autopsy studies (1–4) that thromboembolic stroke can be caused by white blood cell, mixed blood cell, and red blood cell clots. White cell clots were found in the cerebral arteries in about one-third of patients with an embolic occlusion in the distribution of the internal carotid artery. In patients with primary thrombosis of the internal carotid artery, the frequency of white cell clots in the cerebral arteries was even higher. White thrombi consist of varying amounts of cellular debris, fibrin, and platelets but only few red cells. Red thrombi contain erythrocytes and some fibrin (4,5).

Second, an intraarterial infusion of streptokinase rapidly resolved the red but not the white portions of coronary thrombi in patients with acute myocardial infarction (6). In a rabbit model of thromboembolic cerebral ischemia, intraarterial thrombolysis with recombinant tissue-type plasminogen activator improved cerebral perfusion and reduced the infarct volume in red but not white emboli (7). It was postulated that fibrin-rich white thrombi hinder fibrinolysis because they retract more than do red thrombi, which results in reduced permeability to the bulk flow of thrombolytic agents, increased fibrin content per unit of clot volume, and decreased plasminogen content (8–10).

Thus, the lysability of a thrombus appears to increase with increasing hematocrit levels and, therefore, hemoglobin. Hemoglobin determines the attenuation of a clot. Therefore, it should be possible to determine the hematocrit levels of thromboembolic material with computed tomography (CT) and to estimate the potential effectiveness of thrombolytic therapy. The aim of our phantom study was to test with serial CT scans whether white, mixed, and red thrombi could be differentiated according to their CT attenuation.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Preparation of the Phantom
To prepare thrombi with a homogeneous and reproducible composition, we obtained blood from five healthy male volunteers with a syringe (CPDA-1 Monovette; Sarstedt, Nümbrecht, Germany) that contained trinatriumcitrate as anticoagulant, as well as natriumphosphate, dextrosemonohydrate, and adenine. Informed consent was obtained from all volunteers prior to enrollment in the study, which was approved by the institutional review board. Blood cell counts were obtained (CELL-DYN 1700; Abbott Laboratories, Abbott Park, Ill). After centrifugation at 200 g for 10 minutes, a platelet-enriched plasma was separated. The concentration of thrombocytes was determined (CELL-DYN 1700; Abbott Laboratories), and the fibrinogen level was determined (STA Compact; Roche, Mannheim, Germany) according to the method of Clauss (11).

Platelet-enriched plasma and whole blood were mixed to produce samples with a hematocrit level of 0 (group 1), 0.005 (group 2), 0.03 (group 3), 0.15 (group 4), and 0.35 (group 5). For each sample, 10 x 4 mL was transferred into separate tubes, and 700 µL of a thrombin solution (6 mg/1 mL 0.9% saline solution; 50 NIH units/mg; Merck, Darmstadt, Germany) was added to each tube. After 2 hours, the retracted clots were transferred into polyethylene tubes with a length of 4 cm and an inner diameter of 3 mm. Ten tubes were filled with gelatin (Rheingold Gelatine; Ewalt Gelatine GmbH, Bad Sobernhain, Germany) and served as control. The ends of the tubes were soldered over a flame.

The probes were placed into a plastic box filled with a solution of 108 g of gelatin and 0.6 mol/L of gadopentetate dimeglumine (Magnevist; Schering, Berlin, Germany) in 2 L of distilled water. The gadopentetate dimeglumine was used to bring the T2 relaxation rate of the gelatin closer to the relaxation rate of brain tissue. Thus, we produced a phantom with 10 tubes for each of the five types of clots, for a total of 50 tubes, and 10 tubes with gelatin. The phantom was kept at room temperature throughout the whole experiment.

CT Procedures
CT was performed by one author (K.K.) with a multisection scanner (Volume Zoom; Siemens, Forchheim, Germany) 6, 12, 24, 48, 96, and 144 hours after the induction of thrombosis. Multisection helical CT was performed with the following technical parameters: 0.5-mm collimation, pitch of 1 (table increment of 1 mm per rotation), 0.5-mm section thickness, 200-mm field of view, 512 x 512 pixel matrix size, 120-kV tube voltage, 110-mA tube current, 0.5-second rotation time, and 15-second acquisition time.

Two authors (K.K., C.M.) measured the attenuation of each thrombus and the gelatin within regions of interest that were placed centrally in three sections of each probe with consensus. The diameters of the regions of interest were about 2 mm. The same measurements were performed within the gelatin outside the tubes to obtain control numbers. Thus, for each of the five types of thrombi, the gelatin inside the tubes and the gelatin outside the tubes, 30 measurements of the attenuation were performed at each time.

Histopathologic Examination of Thrombi
At the end of the experiment, the thrombi were fixed in 4% paraformaldehyde, embedded in paraffin, and cut into 4-µm sections. Routine staining was performed with Masson trichrome stain. The stained sections were evaluated by two authors (K.K., S.Z.). According to their fibrin and erythrocyte content, the thrombi were classified into white, mixed, and red categories. Thrombi in which the estimated proportion of erythrocytes or of fibrin accounted for less then 10% of the thrombus volume were classified as white or red thrombi, and the remainder were classified as mixed thrombi. The classification was the result of a consensus opinion after discussion between the two authors.

Statistical Analysis
We used a two-way analysis of variance with repeated measurements to test for differences in attenuation between the types of thrombi over time. Changes over time were also examined with the paired samples t test. The Mann-Whitney U test was used to look for artificial differences between the CT numbers of the gelatin inside and the gelatin outside the plastic tubes that might be caused by the walls of the tubes. Differences were regarded as significant if the P values were less than .05.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
The blood cell counts of the volunteers’ whole blood showed a median count of 4.26 x 10¹²/L ± 0.37 (SD) erythrocytes, 200 x 10⁹/L ± 16 thrombocytes, and 4.9 x 10⁹/L ± 0.9 leukocytes; a hematocrit level of 0.374 ± 0.026; and a mean corpuscular hemoglobin concentration of 34.8 g/dL ± 0.3. In the platelet-enriched plasma, the count was 0.02 x 10¹²/L ± 0.01 erythrocytes, 430 x 10⁹/L ± 25 thrombocytes, and 0.7 x 10⁹/L ± 0.3 leukocytes, with a fibrinogen level of 2.35 g/L ± 0.39. In the whole blood, the erythrocyte count was 4.09–4.93 x 10¹²/L, the thrombocyte count was 169–202 x 10⁹/L, the leukocyte count was 3.2–5.8 x 10⁹/L, the hematocrit level was 0.346–0.414, and the mean corpuscular hemoglobin concentration was 34.8– 35.5 g/dL. In the platelet-enriched plasma, the erythrocyte count was 0.01–0.02 x 10¹²/L, the thrombocyte count was 393–453 x 10⁹/L, the leukocyte count was 0.2–0.8 x 10⁹/L, and the fibrinogen level was 2.1–3.1 g/L.

Figure 1 demonstrates the histologic differences between the five groups of thrombi 6 days after the induction of thrombosis. The thrombi of the first group (Fig 1a) consisted of fibrin that enclosed numerous plasma-filled clefts. There were only a few scattered erythrocytes. Because of thrombocytolysis, no platelet aggregates were distinguished in the fibrin-rich thrombi. In the thrombi of the second group (Fig 1b), large fibrin strands and erythrocytes were distributed rather homogeneously. The thrombi of the third group (Fig 1c) consisted predominately of erythrocytes. The fibrin formed large lumps, which gave the thrombi an inhomogeneous appearance. In the fourth group (Fig 1d), fibrin was scarce and almost completely concealed by erythrocytes that were packed even more densely than in the third group. No histologic study was performed in the fifth group, as the thrombi were almost liquefied after 6 days. According to the histologic findings, thrombi of group 1 were classified as white; thrombi of groups 2 and 3, mixed; and those of groups 4 and 5, red.

View larger version (133K): [in this window] [in a new window] [Download PPT slide]   Figure 1a. Photomicrographs show representative histologic sections of the thrombi of groups 1-4. The thrombi are 6 days old and depict (a) white cell clots, (b, c) mixed cell clots, and (d) red cell clots. No histologic study was performed in the red thrombi of the fifth group because they were almost liquefied. The dark erythrocytes (short arrows and white asterisks) can be distinguished from the light gray fibrin (long arrows and black asterisk). Thrombi of group 1 show artificial plasma-filled clefts () that probably result from extensive clot retraction. (Masson Goldner stain; original magnification, x200.)

View larger version (128K): [in this window] [in a new window] [Download PPT slide]   Figure 1b. Photomicrographs show representative histologic sections of the thrombi of groups 1-4. The thrombi are 6 days old and depict (a) white cell clots, (b, c) mixed cell clots, and (d) red cell clots. No histologic study was performed in the red thrombi of the fifth group because they were almost liquefied. The dark erythrocytes (short arrows and white asterisks) can be distinguished from the light gray fibrin (long arrows and black asterisk). Thrombi of group 1 show artificial plasma-filled clefts () that probably result from extensive clot retraction. (Masson Goldner stain; original magnification, x200.)

View larger version (118K): [in this window] [in a new window] [Download PPT slide]   Figure 1c. Photomicrographs show representative histologic sections of the thrombi of groups 1-4. The thrombi are 6 days old and depict (a) white cell clots, (b, c) mixed cell clots, and (d) red cell clots. No histologic study was performed in the red thrombi of the fifth group because they were almost liquefied. The dark erythrocytes (short arrows and white asterisks) can be distinguished from the light gray fibrin (long arrows and black asterisk). Thrombi of group 1 show artificial plasma-filled clefts () that probably result from extensive clot retraction. (Masson Goldner stain; original magnification, x200.)

View larger version (113K): [in this window] [in a new window] [Download PPT slide]   Figure 1d. Photomicrographs show representative histologic sections of the thrombi of groups 1-4. The thrombi are 6 days old and depict (a) white cell clots, (b, c) mixed cell clots, and (d) red cell clots. No histologic study was performed in the red thrombi of the fifth group because they were almost liquefied. The dark erythrocytes (short arrows and white asterisks) can be distinguished from the light gray fibrin (long arrows and black asterisk). Thrombi of group 1 show artificial plasma-filled clefts () that probably result from extensive clot retraction. (Masson Goldner stain; original magnification, x200.)

The time course of the CT numbers of the five types of thrombi is presented in the Table and Figure 2. Two-way analysis of variance with repeated measurements showed significant differences in the CT numbers (F_4,145 = 978.99, P < .001). The CT numbers differed between all groups at almost all times (P < .05), except for groups 4 and 5, in which differences in attenuation were never observed, and groups 1 and 2 with a similar hematocrit level at 24 and 144 hours (P > .05). Thus, the CT numbers could be used to reliably distinguish white, mixed, and red thrombi at almost all time measurements.

View larger version (17K): [in this window] [in a new window] [Download PPT slide]   Figure 2. Graph shows the time course of the CT numbers for the five groups of thrombi 6-144 hours after the induction of thrombosis. The CT numbers are the mean plus or minus SD. The proportion of erythrocytes increases from group 1 to group 5, whereas that of fibrin decreases. The corresponding histologic sections are shown in Figure 1. The thrombi can be differentiated reliably according to their CT numbers. = data of group 1, = data of group 2, = data of group 3, = data of group 4, and = data of group 5.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Results of this phantom study support the hypothesis that CT is a suitable technique for estimating the proportion of erythrocytes within in vitro thrombi. These results are in good agreement with those of studies (13,14) in which the authors demonstrated that the attenuation coefficient of whole blood measured with CT increased linearly with the concentration of hemoglobin and with the hematocrit level. The CT numbers in these studies were similar for both unclotted and clotted samples of plasma or whole blood: Plasma had CT numbers of 24–26 HU, and platelet-fibrin thrombi had CT numbers around 20 HU. Unclotted and clotted blood with a hematocrit level near 0.9 had CT numbers of 78–86 HU. The authors did not state how old their samples were, but it seems appropriate to compare their numbers with those of the 6-hour measurement in our study, which revealed 24 HU ± 8 for white cell clots and 76 HU ± 9 for the red thrombi of group 5. Since plasma and retracted fibrin thrombi appear to have similar CT numbers, the plasma-filled clefts in the fibrin clots in our study probably had no decisive influence on the CT numbers of group 1 thrombi.

Because of the linear relation between the attenuation coefficient and the hematocrit level of clotted blood samples, it appears feasible to classify in vitro thrombi with the help of their CT numbers as white, mixed, and red. According to the measurements in our study, the CT numbers of pure fibrin thrombi did not exceed 24 HU ± 8, whereas red thrombi had CT numbers of 65 HU ± 9 or higher. However, the thrombi found in an autopsy study (4) not only consisted of fibrin and erythrocytes but also contained atheromatous and cellular debris. In the adjacent vessel wall, calcific depositions frequently appeared in close proximity to the thrombi (4). Since we can assume that the debris had lower CT numbers than did the erythrocytes and that the debris did not respond to thrombolysis, an attenuation-based estimation of the hematocrit level and the effectiveness of thrombolysis should still be possible.

Vascular calcifications can cause a spectral shift because of beam hardening, which results in an artificial decrease in the attenuation of vaso-occlusive material. In cases of severe calcification, the prognosis of thrombolysis may thus be underestimated. Because calcification was not found within thrombi (4), it is unlikely that one would measure a higher attenuation that could be used to predict a red thrombus that one would anticipate to have a better response to thrombolysis. The inner diameter of the tubes in our study was 3 mm, so their lumen diameter approximates the lumen diameter of the distal part of the internal carotid artery and of the origin of the middle cerebral artery. We therefore can assume that the spatial resolution of the CT sequence used in this study is suitable for attenuation measurements of thrombi in the distal part of the internal carotid artery and probably even in the trunk of the middle cerebral artery. Although it is unlikely that partial volume effects can be avoided entirely in clinical reality, the influence of partial volume effects on CT numbers can be minimized by using a collimation of 0.5 mm.

The difference in the attenuation between the gelatin in the plastic tubes and the gelatin outside of the tubes was caused by the wall of the polyethylene tubes. This is a known systemic error (14), which can be corrected as described in the Materials and Methods section. That the corrected CT numbers of the thrombi in our study correspond to the results of previous studies (13,14) indicates that the CT measurements were reliable despite the systemic error. Tubes that caused a spectral shift of a lesser degree could not be used because it was not possible to close their ends.

When deciding to prepare all thrombi at the same time, we had to accept that the thrombi had an age of nearly 6 hours when the phantom was completed. Retrospectively, this was a minor drawback of the experiment. As the thrombi retracted predominately during the first 3 hours, we can assume that the CT numbers that were measured at 6 hours also apply for thrombi that are 3 hours old. However, this does not imply that we cannot use reference values in patients within a therapeutic window of 1–3 hours. Mural thrombi are, at least in part, older than 3 hours, as most emboli probably are.

Practical application: The results of our phantom study suggest that the proportion of erythrocytes within a thrombus can be estimated by using its CT numbers. Because the lysability of thromboembolic material depends on its hematocrit level, measurement of its CT numbers may be useful to help predict the prognosis of thrombolytic therapy in patients with cerebral infarction.

	ACKNOWLEDGMENTS

 
We are indebted to Grit Welzel for her substantial help with the statistical analysis. We thank Nicole Windisch for her laboratory assistance.

	FOOTNOTES

 
Author contributions: Guarantors of integrity of entire study, K.K., K.S.; study concepts and design, K.K., T.W.; literature research, K.K., C.M.; experimental studies, K.K., C.M., S.Z.; data acquisition, K.K., C.M., S.Z.; data analysis/interpretation, all authors; statistical analysis, K.K., T.W.; manuscript preparation, K.K., T.W.; manuscript editing and revision/review, K.K., T.W., K.S.; manuscript definition of intellectual content and final version approval, all authors.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

1. Castaigne P, Lhermitte F, Gautier JC, Escourolle R, Derouesne C. Internal carotid artery occlusion: a study of 61 instances in 50 patients with post-mortem data. Brain 1970; 93:231-258.[Free Full Text]
2. Jorgensen L, Torvik A. Ischaemic cerebrovascular diseases in an autopsy series. II. Prevalence, location, pathogenesis, and clinical course of cerebral infarcts. J Neurol Sci 1969; 9:285-320.[CrossRef][Medline]
3. Torvik A, Jorgensen L. Thrombotic and embolic occlusions of the carotid arteries in an autopsy series. II. Cerebral lesions and clinical course. J Neurol Sci 1966; 3:410-432.[CrossRef]
4. Friedman M, Bovenkamp GJV. The pathogenesis of a coronary thrombus. Am J Pathol 1966; 48:19-44.[Medline]
5. Jorgensen L. Experimental platelet and coagulation thrombi: a histological study of arterial and venous thrombi of varying age in untreated and heparinized rabbits. APMIS 1964; 62:189-223.
6. Brown BG, Gallery CA, Badger RS, et al. Incomplete lysis of thrombus in the moderate underlying atherosclerotic lesion during intracoronary infusion: quantitative angiographic observations. Circulation 1986; 73:653-661.[Abstract/Free Full Text]
7. Kirchhof K, Sikinger M, Welzel T, Zoubaa S, Sartor K. The effect of intra-arterial recombinant tissue-type plasminogen activator on red and white autologous emboli in a rabbit model of focal cerebral ischemia. Thromb Haemost. (in press).
8. Blinc A, Keber D, Lahajnar G, et al. Magnetic resonance imaging of retracted and nonretracted blood clots during fibrinolysis in vitro. Haemostasis 1992; 22:195-201.[Medline]
9. Blinc A, Kennedy SD, Bryant RG, Marder VJ, Francis CW. Flow through clots determines the rate and pattern of fibrinolysis. Thromb Haemost 1994; 71:230-235.[Medline]
10. Carr ME, Hardin CL. Fibrin has larger pores when formed in the presence of erythrocytes. Heart Circ Physiol 1987; 22:H1069-H1073.
11. Clauss A. Gerinnunsphysiologische schnellmethode zur bestimmung des fibrinogens. Acta Haematol 1957; 17:237-246.[Medline]
12. Rogers WJ, Prichard JW, Hu YL, et al. Characterization of signal properties in atherosclerotic plaque components by intravascular MRI. Arterioscler Thromb Vasc Biol 2000; 20:1824-1830.[Abstract/Free Full Text]
13. New PFJ, Aronow S. Attenuation measurements of whole blood and blood fractions in computed tomography. Radiology 1976; 121:635-640.[Abstract]
14. Norman D, Price D, Boyd D, Fishman R, Newton TH. Quantitative aspects of computed tomography of the blood and cerebrospinal fluid. Radiology 1977; 123:335-338.[Abstract]

This article has been cited by other articles:

				T Kharitonova, M Thoren, N Ahmed, J M Wardlaw, R von Kummer, L Thomassen, N Wahlgren, and for the SITS investigators Disappearing hyperdense middle cerebral artery sign in ischaemic stroke patients treated with intravenous thrombolysis: clinical course and prognostic significance J. Neurol. Neurosurg. Psychiatry, March 1, 2009; 80(3): 273 - 278. [Abstract] [Full Text] [PDF]

				O. Ozdemir, A. Leung, M. Bussiere, V. Hachinski, and D. Pelz Hyperdense Internal Carotid Artery Sign: A CT Sign of Acute Ischemia Stroke, July 1, 2008; 39(7): 2011 - 2016. [Abstract] [Full Text] [PDF]

				H. S. Kim, D. H. Lee, C. G. Choi, S. J. Kim, and D. C. Suh Progression of middle cerebral artery susceptibility sign on t2*-weighted images: its effect on recanalization and clinical outcome after thrombolysis. Am. J. Roentgenol., December 1, 2006; 187(6): W650 - W657. [Abstract] [Full Text] [PDF]

				E. Y. Kim, J. H. Heo, S-K Lee, D. J. Kim, S-H Suh, J. Kim, and D. I. Kim Prediction of thrombolytic efficacy in acute ischemic stroke using thin-section noncontrast CT Neurology, November 28, 2006; 67(10): 1846 - 1848. [Abstract] [Full Text] [PDF]

				S. H. Thomas, L. H. Schwamm, and M. H. Lev Case records of the Massachusetts General Hospital. Case 16-2006. A 72-year-old woman admitted to the emergency department because of a sudden change in mental status. N. Engl. J. Med., May 25, 2006; 354(21): 2263 - 2271. [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) All Versions of this Article: 2273020530v1 228/1/126    most recent Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Kirchhof, K. Articles by Sartor, K. Search for Related Content PubMed PubMed Citation Articles by Kirchhof, K. Articles by Sartor, K.