The fate of circulating megakaryocytes during cardiopulmonary bypass

The fate of circulating megakaryocytes during cardiopulmonary bypass Megakaryocytes with intact cytoplasm normally leave the bone marrow, enter central venous blood, and are filtered in the lungs. During cardiopulmonary bypass, large megakaryocytes are not filtered by the lungs and may not be removed in the extracorporeal circuit by arterial line filters. In such circumstances, they could enter the systemic circulation and block smaller cerebral vessels, resulting in neurologic impairment. To investigate the fate of circulating megakaryocytes during cardiopulmonary bypass, central venous blood and oxygenated blood samples before and after arterial line filtration (40 ~m polyester screen filter) were obtained from 10 patients undergoing cardiopulmonary bypass. Megakaryocytes were isolated by whole blood filtration and identified by their characteristic structure after May-Grunwald-Giemsa staining. In preliminary studies, megakaryocyte identification was verified by immunolabeling. AU samples contained megakaryocytes with copious cytoplasm. Their frequencies in central venous blood and oxygenated blood before and after the arterial line filtration (corrected for hemodilution) were 23.4 ± 9.3 per milliliter (mean ± standard error of the mean, range 3.1 to 89.7 per milliliter), 21.0 ± 8.2 per milliliter (2.0 to 84.2 per milliliter) and 17.1 ± 7.4 per milliliter (3.1 to 80.4 per milliliter), respectively. Megakaryocytes with scant or no visible cytoplasm were also observed. The results confirm that circulating· megakaryocytes are a normal physiologic component. During cardiopulmonary bypass, megakaryocytes with copious cytoplasm (mean diameter 42.7 ~m, range 22 to 78 ~m) can pass through the extracorporeal circuit. In the absence of filtration by the lungs, these large ceUs have access to the systemic circulation. This study supports a possible role for circulating megakaryocytes in the development of cerebral dysfunction after cardiopulmonary bypass. (J THORAC CARDIOVASC SURG 1993;106:658-63)

M. J. Woods, FIMLS, MMedSci,* M. Greaves, MD, FRCP, MRCPath, G. H. Smith, MB, FRCS, and E. A. Trowbridge, PhD, MSc, Sheffield, United Kingdom

Circulating megakaryocytes are normal blood components.' Megakaryocytes with intact cytoplasm leave the bone marrow/ and gain access to the central venous blood supply.' Few megakaryocytes, largely devoid of cytoplasm, are found in central arterial and peripheral

From the Departments of Medical Physics and Clinical Engineering and Haematology, University of Sheffield, Royal Hallamshire Hospital, and the Department of Cardiac Surgery, Northern General Hospital, Sheffield, Ll.K, M. J. Woods is supported by a Medical Research Council Training Award. Received for publication April 30, 1992. Accepted for publication Nov. 23, 1992. Address for reprints: E. A. Trowbridge, PhD, MSc, Department of Medical Physics and Clinical Engineering, Royal Hallamshire Hospital, Glossop Road, Sheffield SIO 2JF, U.K. Copyright © 1993 by Mosby-Year Book, Inc. 0022-5223/93/$1.00 + .10

658

12/1/44676

venous blood.l-" The normal fate for circulating megakaryocytes is filtration in the pulmonary circulation, where their cytoplasm is removed, probably to form platelets.i During cardiopulmonary bypass (CPB), the filtration function of the lungs is absent. This procedure therefore provides an opportunity to study the fate of megakaryocytes when their passage in the blood is not arrested by the pulnionary microcirculation. If large megakaryocytes with intact cytoplasm enter the systemic circulation during CPB, they theoretically could cause blockage of smaller blood vessels. This process may be important in the development of postoperative cerebral dysfunction, which occurs in as many as 80% of patients after CPB.6-8 Several causes of this phenomenon have been pro posed,9-12 but a link with the presence of circulating megakaryocytes has not previously been investigated. In this study, we used a whole blood filtration technique-' to isolate megakaryocytes from blood obtained at

The Journal of Thoracic and Cardiovascular Surgery Volume 106, Number 4

different sites within the extracorporeal circuit during CPB. The potential for megakaryocytes to enter the systemic circulation of patients undergoing CPB was then assessed.

Patients and methods Patients. Ten patients were entered into the study (eight male, two female; mean age 60.5 years, range 42 to 70 years). Cor~nary artery bypass grafting was performed on six patients, aortic valve replacement was performed on two patients, mitral valve replacement was performed on one patient, and coronary artery bypass grafting plus mitral valve repair was performed on one patient. Preoperative assessment included a full blood cell count and coagulation screen. On the basis of these test results, no patient was considered to have significant hematologic disease. All surgical procedures were carried out with the support of total CPB employing 2 L Plasma-Lyte acellular priming fluid (Baxter Healthcare Ltd., Thetford, Norfolk, England) and nonpulsatile flow with a minimum patient temperature of 28° C. Depending on the type of operation, venous drainage was achieved either by cannulation of the right atrium or by separate caval cannulation with 7 mm plastic cannulas. All patients received the arterial return from the CPB circuit through a cannula in the ascending aorta. Mean flow rate (from information in case notes of eight patients) at 28° C was 2.1 Lzmin (range 1.9 to 2.5 L/min). Blood was oxygenated with a Shiley M-2000 membrane oxygenator (Shiley Europe, Staines, England) in six patients, a Bard William Harvey HF-5700 membrane oxygenator (Bard Ltd., Crawley, England) in two patients, and a Shiley S-IOOA bubble oxygenator (Shiley Europe) in two patients. In all cases, a polyester screen filter (Pall Biomedical Ltd., Portsmouth, England) with a pore size of 40 ~m was fitted to the arterial line of the extracorporeal circuit. The average duration of CPB (from information in case notes of eight patients) was 96 minutes (range 49 to 152 minutes). Control subjects. Megakaryocytes were evaluated in peripheral venous blood of 28 healthy volunteers. The group consisted of 12 male and 16 female volunteers with a mean age of 34.8 years (range 23 to 59 years). Blood sampling. Hospital ethical committee approval and informed consent of patients were sought for the blood sampling protocol. Approximately 10 to 15 minutes after the beginning of CPB, blood was withdrawn from three sampling sites within the extracorporeal circuit. Central venous blood was obtained just before oxygenation. Oxygenated blood was taken before passage through the arterial line filter and also after flowing through the same filter but before returning to the systemic circulation. Blood was withdrawn by dry syringe and immediately transferred to tubes containing sufficient tripotassium ethylenediaminetetraacetic acid to anticoagulate a 4 ml sample volume. At the time of sampling, all patients were moderately hypo.thermic (31 ° to 33° C). Blood was removed from an antecubital vein of each control subject with a standard venepuncture technique. Samples were anticoagulated with ethylenediaminetetraacetic acid. Megakaryocyte isolation. A total of 3 or 4 ml of well-mixed blood was filtered at 37° C under unit gravity through Nuclepore polycarbonate membranes (Costar UK Ltd., High Wycombe, Buckinghamshire, England) with a 5 ~m pore

Woods et al. 659

diameter. This was followed by a small aliquot of prewarmed saline solution, which facilitated the passage of most blood cells, except megakaryocytes, through the membranes. With the same technique, megakaryocytes were isolated from 2 ml of each peripheral venous blood sample. Megakaryocyte identification. After blood filtration, membranes were routinely air dried, fixed in methanol, and stained by the May-Grunwald-Giemsa (MGG) method. Megakaryocytes were initially identified by their large size during microscopic examination of MGG-stained membranes with a lowpower objective (XIO). Further analysis was completed at high-power magnification (xiOO objective), with particular attention given to nuclear configuration and density of nuclear staining. Only megakaryocytes with a longest diameter of 20 JLm or more (by comparison with 5 JLm membrane pores) were recorded and classified into four types according to the amount of granular eytoplasm present: type 1, naked nucleus, typically lobed and densely stained but devoid of visible cytoplasm; type 2, nucleus with scanty cytoplasm forming a thin rim around the nuclear perimeter or present as a small portion (or portions) attached to part of the nucleus; type 3, nucleus with moderate amount of cytoplasm (cytoplasm/nucleus area ratio < 1.0); type 4, nucleus possessing copious cytoplasm (cytoplasm/nucleus area ratio ~ 1.0). To verify megakaryocyte identity, a single central venous blood sample was filtered as described and the membrane was subjected to immunolabeling by an alkaline phosphatase (AP) / anti-AP technique. With this procedure, megakaryocytes were recognized by the presence of the platelet membrane glycoprotein Ilia with a monoclonal antibody (CD61; Dako Ltd., High Wycombe, Buckingharnshire, England). Megakaryocyte counting. All membranes were systematically scanned to allow the entire area of each filter to be examined for the presence of megakaryocytes. Megakaryocyte counts were expressed as number per milliliter of whole blood. In the patient group, counts were corrected for blood dilution during CPB by the following formula: MKa = MKbX (HCTp/HCTc)/V where MKa is the megakaryocyte number per I ml undiluted whole blood, MKb is the number of megakaryocytes counted on membrane, BCTp is the preoperative hematocrit value, BCTc is the hematocrit value of sample obtained during CPB, and V is the volume (rnl) of diluted whole blood filtered. Hematocrit values of samples obtained before operation and during CPB are shown for each patient in Table I. Statistics. A Wilcoxon signed-rank test for paired data was used to compare megakaryocyte counts, for all megakaryocyte morphologic types, on both sides of the oxygenator and on both sides of the arterial filter. Statistical significance was assessed at the 5% level. Results

Patients. Most nonmegakaryocytes passed through 5 JLm polycarbonate membranes whereas large megakary-

ocytes were retained. Representatives of the four morphologic classes of megakaryocyte recognized in M GG-stained membranes are illustrated in Fig. 1. Megakaryocytes were found in all blood samples filtered. Equivalent type 2,3, and 4 megakaryocytes were identi-

660

The Journal of Thoracic and Cardiovascular Surgery October 1993

Woods et al.

Fig. 1. Examples of type 1 (A), type 2 (B), type 3 (C), and type 4 (D) megakaryocytes isolated from blood obtained during CPB. The regular, round structures are 5 Jim membranepores. Table I. Hematocrit values for 10 patients in the CPB

study group Hematocrit value Patient

Before operation

During CPB"

1 2 3 4 5 6 7 8 9 10

0.464 0.458 0.381 0.387 0.413 0.404 0.400 0.408 0.413 0.417

0.257 0.257 0.176 0.179 0.266 0.225 0.256 0.244 0.220 0.245

·Results given for central venous blood samples obtained 10 to 15 minutes after the start of bypass.

fied in central venous blood by the demonstration of glycoprotein 111a with the AP/ anti-AP technique. Individual total megakaryocyte counts are displayed in Fig. 2. The mean total counts ± standard error of the mean (SEM) and ranges for central venous blood and

oxygenated blood before arterial line filtration and after filtration were as follows: 34.0 ± 10.9 per milliliter (6.2 to 105.3 per milliliter), 31.6 ± 9.5 per milliliter (5.0 to 99.3 per milliliter), and 31.0 ± 9.8 per milliliter (7.7 to 110.1 per milliliter), respectively. The distribution of megakaryocyte morphologic classes at each of these three sites is shown in Fig. 3. The number of megakaryocytes with copious cytoplasm (type 4) was significantly reduced after blood had passe? through the arterial line filter (p < 0.025). All other megakaryocyte count comparisons across the oxygenator or arterial filter failed to demonstrate a statistically significant difference (p > 0.05). Control subjects. Megakaryocytes were isolated from all peripheral venous blood samples analyzed. The mean count ± SEM was 4.1 ± 1.0 per milliliter (range 0.5 to 19.5 per milliliter). Of the megakaryocytes, 83.9% were type 1 cells, 13.9% were type 2, 1.3% were type 3, and 0.9% were type 4. Discussion The whole blood filtration method used here to separate megakaryocytes from other blood cells by size was a modification of that reported by Ohashi. 14 An advantage

The Journal of Thoracic and Cardiovascular Surgery Volume 106. Number 4

Woods et al. 661

120 •

Venous Arterial Pre-filter Ill! Arterial Post·fllter

o

100 80

60 40

20

o

1

2

3

456

7

8

9

10

Patients

Fig. 2. Total megakaryocyte countspermilliliter for eachpatientat the threesitesfromwhichblood waswithdrawn during CPR

35 30

• Venous 0 Arterial Pre-filter IllI Arterial Post-filter

25

20 15

10

5

o

Type 1

Type 2

Type 3

Type 4

Fig. 3. Distribution of megakaryocyte morphologic types at the three sampling sites (mean ± SEM, n = 10). of this procedure is that blood is subjected to minimal manipulation to maximize megakaryocyte yield. Techniques such as sedimentation, cell lysis, and centrifugation, used by other workers to remove nonmegakaryocytes,4, 15-17 may cause disruption of cell cytoplasm or loss of entire megakaryocytes. Efficient isolation of megakaryocytes from marrow aspirate and blood has previously been demonstrated with the method described in this study." Megakaryocytes isolated from blood samples obtained during CPB were identified by their characteristic structure after MGG staining (Fig. I). All samples contained megakaryocytes. Further evidence for the presence of these cells in central venous blood was provided by the

demonstration of glycoprotein l l la. The use of the APjanti-AP technique, incorporating several washing procedures, for immunolabeling of megakaryocytes subjects these cells to extra manipulation. In addition, this method may not delineate type I megakaryocytes, which lack visible cytoplasm on light microscopy. MGG staining was therefore adopted routinely for the recognition of mature megakaryocytes on membranes. It was proposed by Levine, Hazzard, and Lamberg'" that application of certain morphologic criteria, including large cell size and characteristic nuclear configuration, allows detection of virtually all normal megakaryocytes without cytochemicalor immunolabeling procedures. Total megakaryocyte counts were variable between

662

Woods et al.

patients (Fig. 2), but mean total counts were similar for the three blood sampling sites. When these figures were analyzed further by morphologic class (Fig. 3), type 4 megakaryocytes were most prevalent. The frequency of type 4 megakaryocytes in central venous blood was 2304 ± 9.3/ml (mean ± SEM, range3.1 to 89.7/ml), in oxygenated blood before arterial line filtration, 21.0 ± 8.2/ml (2.0 to 84.2/ml), and after filtration, 17.1 ± 7A/ml (3.1 to 80A/ml). With the assumption of steady-state flow conditions, the frequency of these cells was significantly reduced by passage across the arterial filter (p < 0.025). From a practical viewpoint, however, 73% of type 4 megakaryocytes entering the extracorporeal circuit successfully negotiated the oxygenator and filtration units. When whole blood filtration was applied to peripheral venous blood taken from healthy volunteers, megakaryocytes were isolated from all samples. The vast majority of these cells were devoid of visible cytoplasm. Type 4 megakaryocytes were rarely observed in blood from peripheral veins. The origin of circulating megakaryocytes can be explained by consideration of their normal physiology. Electron microscopic evidence has demonstrated that megakaryocytes with intact cytoplasm leave the bone marrow? their site of formation, and pass through the marrow-blood barrier to enter the blood. This is consistent with the presence of type 4 megakaryocytes in central venous blood.v" Transport of megakaryocytes in the blood is halted in the lungs, where the majority shed their cytoplasm. Few megakaryocytes, largely devoid of cytoplasm, are found in central arterial and peripheral venous blood. This is in agreement with observations in this study as well as in previous reports.v 4 This information implicates the lungs as an important site of platelet production.I?: 20 A mechanism has been proposed by which megakaryocyte cytoplasm is physically fragmented in the pulmonary microcirculation to produce platelets. 5,21 Theoretic calculations to determine the number of megakaryocytes in central venous blood necessary to maintain a normal platelet count in steadystate thrombopoiesis indicate a number similar in order of magnitude to the number isolated experimentally. 19 With similar calculations, it can be shown that for an equivalent cardiac output and CPB of l-hour duration, approximately 7 to 8 million megakaryocytes will enter the extracorporeal circulation. It is possible that a certain proportion of megakaryocytes isolated during CPB originated from disrupted marrow during sternotomy. Cardiotomy suction may have provided a transport route to the blood for these megakaryocytes. This could explain the relatively high

The Journal of Thoracicand Cardiovascular Surgery October 1993

megakaryocyte counts observed in two patients (Fig. 2). The results of this study have shown that, during CPB, significant numbers of megakaryocytes with copious cytoplasm passed through both the oxygenator and arterial line filter of the extracorporeal circuit. These cells, normally filtered in the lungs, therefore had access to the systemic circulation. Clinically, this may have relevance to the cerebral dysfunction found in a significant proportion of patients after CPB. 6-8 In this setting, neurologic abnormalities are often asymptomatic or mildly debilitating, but overall they still contribute to postoperative morbidity and mortality.? The exact cause of the phenomenon is still to be elucidated, although blockage of cerebral vessels by embolic material from a variety of sources has been implicated. 11, 22-25 During CPB, relatively large megakaryocytes with intact cytoplasm (mean diameter 42.7 1Lm, range 22 to 78 1Lm, size measured by image analysis as previously described-") could block smaller cerebral vessels after passing into the systemic circulation through the extracorporeal circuit. In addition, release of megakaryocyte cytoplasm could provide a chemical stimulus for vasoconstriction. Vessel obstruction and vasoconstriction may both be important in the development of neurologic sequelae.i" Our investigation supports the concept that megakaryocytes are normal components of blood. During CPB, megakaryocytes with copious cytoplasm have the potential to enter the systemic circulation in the absence of filtration by the pulmonary capillary bed. The possibility that these megakaryocytes contribute to cerebral dysfunction after CPB warrants further investigation. We thank the perfusionists of the Department of Cardiology, Northern General Hospital,for their assistance in obtaining bloodsamplesand Susan Doy for typing the manuscript. REFERENCES I. Trowbridge EA. The circulating megakaryocyte, platelet volume heterogeneity and thrombopoiesis. In: Martin J, Trowbridge A, eds. Platelet heterogeneity, biology and pathology. London: Springer-Verlag, 1990:155-83. 2. Tavassoli M, AokiM. Migrationof entire megakaryocytes throughthe marrow-blood barrier. BrJ Haematol1981;48: 25-9. 3. Tinggaard Pedersen N. Occurrenceof megakaryocytes in variousvessels and their retention in the pulmonary capillaries in man. Scand J Haematol 1978;21 :369-75. 4. Shoff PK, Kirwin KS, Levine RF. Megakaryocytes in circulating blood [Abstract]. Circulation I987;76(suppl): 1339. 5. TrowbridgeEA. Pulmonaryplateletproduction: a physical analogue of mitosis? Blood Cells 1988;13:451-8. 6. Shaw PJ, Bates D, Cartlidge NEF, Heaviside D, Julian

The Journal of Thoracic and Cardiovascular Surgery Volume 106. Number 4

DG, Shaw DA. Early neurological complications of coronary artery bypass surgery. BMJ 1985;291:1384-7. 7. Shaw PJ, Bates D, Cartlidge NEF, eta!. Early intellectual dysfunction following coronary bypass surgery. Q J Med 1986;225:59-68. 8. Smith PLC, Treasure T, Newman SP, et a!. Cerebral consequences of cardiopulmonary bypass. Lancet 1986;1: 823-5. 9. Barash PG. Cardiopulmonary bypass and postoperative neurologic dysfunction. Am Heart J 1980;99:675-7. 10. Allen CMC. Cabbages and CABG. BMJ 1988;297:14856. II. Shaw PJ, Bates D, Cartlidge NEF, et a!. An analysis of factors predisposing to neurological injury in patients undergoing coronary bypass operations. Q J Med 1989; 267:633-46. 12. Campbell DE, Raskin SA. Cerebral dysfunction after cardiopulmonary bypass: aetiology, manifestations and interventions. Perfusion 1990;5:251-60. 13. Woods MJ, Wagner BE,GreavesM, Trowbridge EA. The isolation of circulating megakaryocytes [Abstract]. Br J HaematoI1990;76(Suppll):13. 14. Ohashi T. Megakaryocytes in peripheral blood-studies on normal subjects and patients with various blood diseases. Acta Haematol Jpn 1988;51:1122-9. 15. Scheinin TM, Koivuniemi AP. Megakaryocytes in the pulmonary circulation. Blood 1963;22:82-7. 16. Kaufman RM, Airo R, Pollack S, Crosby WHo Circulating megakaryocytes and platelet release in the lung. Blood 1965;26:720-31. 17. Dago C, Karpas CM, Pincus L, Tytun A, Oppenheim A.

Woods et al.

18. 19.

20.

21.

22.

23.

24. 25.

26.

663

Circulating megakaryocytes: a quantitative observation in normal subjects by age and sex. Acta CytoI1971;15:41O-3. Levine RF, Hazzard KC, Lamberg JD. The significance of megakaryocyte size. Blood 1982;60:1122-31. Trowbridge EA, Martin JF, Slater DN, et a!. The origin of platelet count and volume. Clin Phys Physiol Meas 1984; 5:145-70. Slater DN, Trowbridge EA, Martin JF. The megakaryocyte in thrombocytopenia: a microscopic study which supports the theory that platelets are produced in the pulmonary circulation: Thromb Res 1983;31:163-76. Trowbridge EA, Harley PJ. A stochastic model ofpulmonary platelet production. IMA J Math Appl Med Bioi 1988;5:45-63. Hill JD,J)sborn JJ, Swank RL, Aguilar MJ, de Lanerolle P, Gerbode F. Experience using a new Dacron wool filter during extracorporeal circulation. Arch Surg 1970;101: 649-52. Aguilar MJ, Gerbode F, Hill JD. Neuropathologic complications of cardiac surgery. J THORAC CARDIOVASC SURG 1971;61:676-85. Aberg T, Kihlgren M. Cerebral protection during openheart surgery. Thorax 1977;32:525-33. Blauth C, Arnold J, Kohner EM, Taylor KM. Retinal microembolism during cardiopulmonary bypass demonstrated by fluorescein angiography. Lancet 1986;2:837-9. Woods MJ, Greaves M, Trowbridge EA. Megakaryocyte ploidy in thrombocytosis: improved microdensitometric measurements with a new image analysis system. Eur J Haematol 1990;44:220-6.